AISI DSMHoles Report Final R2

TheJohnsHopkinsUniversity
DepartmentofCivilEngineering
LatrobeHall210
Baltimore,MD21218
ResearchReport
DIRECTSTRENGTHDESIGNOF
COLDFORMEDSTEELMEMBERS
WITHPERFORATIONS
by
CristopherD.Moen
GraduateResearchAssistant
B.W.Schafer
AssociateProfessor
Submittedto:
AmericanIronandSteelInstitute
CommitteeonSpecificationsfortheDesignofColdFormedSteelStructuralMembers

1140ConnecticutAve,Suite705

Washington,DC20036
March2009

Abstract
Coldformedsteel(CFS)structuralmembersarecommonlymanufacturedwithholes
toaccommodateplumbing,electrical,andheatingconduitsinthewallsandceilingsof
buildings.Currentdesignmethodsavailabletoengineersforpredictingthestrengthof
CFS members with holes are prescriptive and limited to specific perforation locations,
spacings,andsizes.TheDirectStrengthMethod(DSM),arelativelynewdesignmethod
forCFSmembersvalidatedformemberswithoutholes,predictstheultimatestrengthof
ageneralCFScolumnorbeamwiththeelasticbucklingpropertiesofthemembercross
section(e.g.,platebuckling)andtheEulerbucklingload(e.g.,flexuralbuckling).This
research project, sponsored by the American Iron and Steel Institute, extends the
appealinggeneralityofDSMtocoldformedsteelbeamsandcolumnswithperforations.
The elastic buckling properties of rectangular plates and coldformed steel beams
andcolumns,includingthepresenceofholes,arestudiedwiththinshellfiniteelement
eigenbuckling analysis. Buckled mode shapes unique to members with holes are
categorized. Parameter studies demonstrate that critical elastic buckling loads either
decreaseorincreasewiththepresenceofholes,dependingonthemembergeometryand
hole size, spacing, and location. Simplified alternatives to FE elastic buckling analysis
formemberswithholesaredevelopedwithclassicalplatestabilityequationsandfreely
availablefinitestripanalysissoftware.
ii
Experimentsoncoldformedsteelcolumnswithholesareconductedtoobservethe
interactionbetweenelasticbuckling,loaddeformationresponse,andultimatestrength.
The experimental results are used to validate an ABAQUS nonlinear finite element
protocol,whichisimplementedtosimulateloadingtocollapseofseveralhundredcold
formed steel beams and columns with holes. The results from these simulations,
supplemented with existing beam and column data, guide the development of design
equationsrelatingelasticbucklingandultimatestrengthforcoldformedsteelmembers
withholes.Theseequationsandthesimplifiedelasticbucklingpredictionmethodsare
presented as a proposed design procedure for an upcoming revision to the American
Iron and Steel Institutes North American Specification for the Design of ColdFormed
SteelStructuralMembers.
iii
SummaryofProgress
TheprimarygoalofthisAISIfundedresearchistoextendtheDirectStrengthMethodto
coldformedsteelmemberswithholes.
Researchbegins
September2005
ProgressReport#1
February2006
Accomplishments:
Evaluated the ABAQUS S9R5, S4, and S4R thin shell elements for accuracy and
versatilityinthinwalledmodelingproblems
Studiedtheinfluenceofelementaspectratioandelementquantitywhenmodeling
roundedcornersinABAQUS
DevelopedcustomMATLABtoolsformeshingholes,plates,andcoldformedsteel
membersinABAQUS
Determinedtheinfluenceofaslottedholeontheelasticbucklingofastructuralstud
channelandclassifiedlocal,distortional,andglobalbucklingmodes
Investigated the influence of hole size on the elastic buckling of a structural stud
channel
Performed a preliminary comparison of existing experimental data on coldformed
steelcolumnswithholestoDSMpredictions
Conducted a study on the influence of the hole width to plate width ratio on the
elasticbucklingbehaviorofasimplysupportedrectangularplate
Papersfromthisresearch:
Moen, C.D., Schafer, B.W. (2006) Impact of Holes on the Elastic Buckling of Cold
Formed Steel Columns with Application to the Direct Strength Method. Eighteenth
InternationalSpecialtyConferenceonColdFormedSteelStructures,Orlando,FL.
Moen,C.D.,Schafer,B.W.(2006)StabilityofColdFormedSteelColumnsWithHoles.
StabilityandDuctilityofSteelStructuresConference,Lisbon,Portugal.
ProgressReport#2
August2006
Accomplishments:
Evaluatedtheinfluenceofslottedholespacingontheelasticbucklingofplates(with
DeterminedtheimpactofflangeholesontheelasticbucklingofanSSMAstructural
stud
iv
Conducted a preliminary investigation into the nonlinear solution algorithms

availableinABAQUS
Compared the ultimate strength and loaddisplacement response of a rectangular
plate and an SSMA structural stud column with and without a slotted hole using
nonlinearfiniteelementmodelsinABAQUS
Calculatedtheeffectivewidthofarectangularplatewithandwithoutaslottedhole
usingnonlinearfiniteelementmodelsinABAQUS
ProgressReport#3
February2007
Accomplishments:
Conductedanexperimentalstudytoevaluatetheinfluenceofaslottedwebholeson
thecompressivestrength,ductility,andfailuremodesofshortandintermediatelength
Csectionchannelcolumns
Studiedtheinfluenceofslottedwebholesontheelasticbucklingbehaviorofcold
formed steel Csection beams and identified unique hole modes similar to those
observedincompressionmembers
Demonstrated that the Direct Strength Method is a viable predictor of ultimate
strengthforbeamswithholes
Moen, C.D., Schafer, B.W. (2008). Experiments on coldformed steel columns with
holes.ThinWalledStructures,46(10),11641182.
Moen,C.D.,Schafer,B.W.(2008).Observingandquantifyingtheelasticbucklingandt
estedresponseofcoldformedsteelcolumnswithholes.FifthInternationalConference
onThinWalledStructures,Brisbane,Australia.
ProgressReport#4
July2007
Accomplishments:
Completed experimental program including tensile coupon tests, elastic buckling

studyofspecimens,andDSMstrengthcomparison
Developed nonlinear finite element approach including a prediction method for
residual stresses and completed preliminary finite element studies of the 24 column
specimens
ProposedDSMapproachforcolumnswithholes,comparedtheoptionsagainstthe
columndatabase,conductedpreliminarysimulationstoexplorecolumnstrengthcurves
Moen,C.D.,Igusa,T.,Schafer,B.W.(2008).Predictionofresidualstressesandstrainsin
coldformedsteelmembers.ThinWalledStructures,46(11),12741289.
Moen,C.D.,Igusa,T.,Schafer,B.W.(2008).Amechanicsbasedpredictionmethodforr
esidualstressesandinitialplasticstrainsincoldformedsteelstructuralmembers.Fifth
ConferenceonCoupledInstabilitiesinMetalStructures,Sydney,Australia.
ProgressReport#5
February2008
Accomplishments:
Developed critical elastic buckling stress equations for stiffened and unstiffened
elementswithholesunderuniaxialcompression.
Implemented finite strip approximation method for predicting the critical elastic
localanddistortionalbucklingloadofcoldformedsteelmemberswithholes.
DerivedandtestedamethodforpredictingtheEulerbucklingloadsofcoldformed
steel columns and beams with holes, including flexural and flexuraltorsional
bucklingofcolumnsandlateraltorsionalbucklingofbeams.
Papers from this research:

Moen,C.D.,Schafer,B.W.(2008).Simplifiedmethodsforpredictingelasticbucklingof
coldformedsteelstructuralmemberswithholes.19thInternationalSpecialtyConference
onColdFormedSteelStructures,St.Louis,Missouri.
FinalReport
Fall2008
Accomplishments:
Created simulated column and beam experiment database with nonlinear finite
elementanalysisinABAQUSandCUFSMelasticbucklingapproximatemethods
Used the simulation results and existing column and beam experiment results to
develop and validate proposed DSM Holes design equations for CFS columns and
beams
vi
TableofContents
Chapter 1 Introduction............................................................................................................................. 1
Chapter 2 Thin-shell finite element modeling in ABAQUS ............................................................... 8
2.1 Comparison of ABAQUS thin-shell elements ............................................................................ 9
2.2 Modeling holes in ABAQUS ....................................................................................................... 16
2.3 Modeling Rounded Corners in ABAQUS ................................................................................. 18
2.4 Summary of modeling guidelines.............................................................................................. 21
Chapter 3 Elastic buckling of cold-formed steel cross-sectional elements with holes .................. 22
3.1 Plate and hole dimensions .......................................................................................................... 23
3.2 Finite element modeling assumptions ...................................................................................... 25
3.3 Stiffened element in uniaxial compression ............................................................................... 25
3.4 Stiffened element in bending ...................................................................................................... 43
3.5 Unstiffened element in uniaxial compression .......................................................................... 58
Chapter 4 Elastic buckling of cold-formed steel members with holes ............................................ 66
4.1 Finite element modeling assumptions ...................................................................................... 67
4.2 Elastic buckling of columns with holes ..................................................................................... 67
4.3 Elastic buckling of beams with holes....................................................................................... 125
Chapter 5 Experiments on cold-formed steel columns with holes ................................................ 159
5.1 Acknowledgements ................................................................................................................... 160
5.2 Testing Program ......................................................................................................................... 160
5.3 Elastic buckling calculations ..................................................................................................... 196
5.4 Experiment results ..................................................................................................................... 205
Chapter 6 Predicting residual stresses and plastic strains in cold-formed steel members ......... 223
6.1 Stress-strain coordinate system and notation ........................................................................ 226
6.2 Prediction method assumptions .............................................................................................. 227
6.3 Derivation of the residual stress prediction method ............................................................. 230
6.4 Derivation of effective plastic strain prediction method ...................................................... 240
6.5 Employing the prediction method in practice: quantifying the coil radius influence ...... 244
6.6 Comparison of prediction method to measured residual stresses ...................................... 248
6.7 Discussion ................................................................................................................................... 254
6.8 Acknowledgements ................................................................................................................... 257
Chapter 7 Nonlinear finite element modeling of cold-formed steel structural members ........... 258
7.1 Preliminary nonlinear FE studies ............................................................................................ 259
7.2 Nonlinear finite element modeling of columns with holes .................................................. 288
Chapter 8 The Direct Strength Method for cold-formed steel members with holes ................... 313
8.1 DSM for columns with holes .................................................................................................... 314
8.2 DSM for laterally braced beams with holes ............................................................................ 374
Chapter 9 Conclusions and proposed future work .......................................................................... 409
9.1 Conclusions ................................................................................................................................. 409
9.2 Future work ................................................................................................................................ 411
References .................................................................................................................................................. 415
Appendix A
ABAQUS input file generator in Matlab ................................................................ 419
Appendix B
ABAQUS element-based elastic buckling results .................................................. 425
Appendix C
Derivation of elastic buckling coefficients for unstiffened elements .................. 435
Appendix D
Elastic buckling prediction method of cross-sectional elements with holes...... 441
Appendix E
Derivation of global critical elastic buckling load for a column with holes....... 445
Appendix F
Column experiment results ...................................................................................... 449
Appendix G
Residual stresses backstress for kinematic hardening implementation ........... 474
Appendix H
Experiment true stress-strain curves ....................................................................... 476
Appendix I Column experiment nonlinear FE simulation results ............................................... 497
vii
Appendix
Appendix
Appendix
Appendix
J Contact simulation in ABAQUS ................................................................................... 521

K
Simulated column experiments database ............................................................... 524
L
Simulated beam experiment database .................................................................... 528
M
Comparison of tested strengths to predicted strengths with AISI S100-07 ....... 531
viii
ListofFigures
Figure 1.1 Perforations are provided in structural studs to accommodate utilities in the walls of
buildings................................................................................................................................... 2
Figure 1.2 Hole patterns in storage rack columns................................................................................... 2
Figure 1.3 Column elastic buckling curve generated with CUFSM ..................................................... 4
Figure 1.4 DSM global buckling failure design curve and equations .................................................. 4
Figure 1.5 DSM local buckling failure design curve and equations ..................................................... 5
Figure 1.6 DSM distortional buckling failure design curve and equations ......................................... 5
Figure 2.1 ABAQUS S4\S4R shell element with four nodes and a linear shape function, ABAQUS
S9R5 shell element with nine nodes and a quadratic shape function ............................ 10
Figure 2.2 Buckled shape of a stiffened plate ........................................................................................ 10
Figure 2.3 Accuracy of ABAQUS S9R5, S4, and S4R elements for a stiffened element with varying
aspect ratios, 8:1 finite element aspect ratio for the S9R5 element, 4:1 element aspect
ratio for the S4 and S4R elements ....................................................................................... 13
Figure 2.4 Accuracy of S4, S4R, and S9R5 elements as a function of the number of elements
provided per buckled half-wavelength, stiffened element, square waves (k=4).......... 14
Figure 2.5 Buckled shape of an unstiffened element, m=1 shown ...................................................... 15
Figure 2.6 Accuracy of S9R5 elements as the number of finite elements provided along an
unstiffened element varies, L/h=4 ...................................................................................... 15
Figure 2.7. Finite element mesh and plate dimensions: slotted, rectangular, and circular holes .. 17
Figure 2.8 Hole discretization using S9R5 elements ............................................................................. 18
Figure 2.9 The critical elastic buckling stress converges to a constant magnitude when the S9R5
element aspect ratio a/b is between 0.5 and 2 and element corner angles are skewed
................................................................................................................................................. 18
Figure 2.10 ABAQUS S9R5 initial curvature limit requires at least five elements to model corner
................................................................................................................................................. 19
Figure 2.11 SSMA 600S162-68 C-section corner modeled with a) one S9R5 element, b) three S9R5
elements.................................................................................................................................. 20
Figure 2.12 The number of S9R5 corner elements has a minimal influence on the critical elastic
buckling loads of an SSMA 600162-68 C-section column with L=48 in. ........................ 21
Figure 3.1 Stiffened and unstiffened elements in a lipped C-section ................................................. 23
Figure 3.2 Element and hole dimension definitions ............................................................................. 24
Figure 3.3 Definition of unstiffened strip A and B for a plate with holes. ................................. 24
Figure 3.4 Definition of neutral axis location for stiffened elements in bending.............................. 24
Figure 3.5 ABAQUS boundary conditions and loading conditions for a stiffened element in
uniaxial compression ............................................................................................................ 26
Figure 3.6 Influence of a slotted hole on the elastic buckling stress of a simply supported
rectangular plate with varying length ............................................................................... 27
Figure 3.7 Comparison of buckled shape and displacement contours for a rectangular plate with
hhole/h=0.66 and L/Lhole =3, (a) with slotted hole and (b) without hole. Notice the
change in length and quantity of buckled cells with the addition of a slotted hole. ... 28
Figure 3.8 Buckled shape of a simply supported plate (a) with a slotted hole and (b) without a
hole. L=15Lhole , hhole/h=0.66. The slotted hole dampens buckling but does not
significantly change the natural half-wavelength of the plate........................................ 28
Figure 3.9 (a) Slotted hole causes local buckling (hhole/h=0.26), compared to (b) buckled cells at
the natural half-wavelength of the plate ............................................................................ 29
Figure 3.10 Definition of center-to-center dimension for the slotted holes ....................................... 30
ix
Figure 3.11 Influence of slotted hole spacing on the elastic buckling load of a long simply
supported rectangular plate ................................................................................................ 30
Figure 3.12 Comparison of buckled shapes for a long stiffened element (L=24 Lhole ) with a slotted
hole spacing of S/Lhole=4 and hhole/h=0.66, 0.44, and 0.26. .................................................. 31
Figure 3.13 Variation in fcr with increasing hhole/h for a stiffened element correspond to buckling
mode shapes (see Figure 3.12 for examples of plate buckling and unstiffened strip
buckling mode shapes)......................................................................................................... 33
Figure 3.14 Unstiffened strip elastic buckling stress conversion from the net to the gross section
................................................................................................................................................. 35
Figure 3.15 Accuracy of stiffened element prediction method as a function of hole spacing S to
plate width h (a) without and (b) with the dimensional limits in Eq. (3.8) and Eq.(3.9)
................................................................................................................................................. 38
Figure 3.16 Accuracy of stiffened element prediction method as a function of hole spacing S to
length of hole Lhole (a) without and (b) with the dimensional limits in Eq. (3.8) and
Eq.(3.9) .................................................................................................................................... 38
Figure 3.17 Accuracy of the stiffened element prediction method as a function of hole width hhole
to plate width h (a) without and (b) with the dimensional limits in Eq. (3.8) and
Eq.(3.9) .................................................................................................................................... 38
Figure 3.18 For plates where the unstiffened strip is narrow compared to the plate width, plate
buckling occurs between the holes. .................................................................................... 39
Figure 3.19 Plate buckling and unstiffened strip buckling may both exist for a plate with holes.
These modes are predicted conservatively as unstiffened strip buckling. ................... 40
Figure 3.20 Accuracy of prediction method for stiffened elements with square or circular holes as
a function of hole width hhole to plate width h. .................................................................. 40
Figure 3.21 Accuracy of the stiffened element elastic buckling prediction method as a function of
unstiffened strip width hstrip versus plate width h for offset holes (a) without and (b)
with the dimensional limits in Eq. (3.8) and Eq.(3.9) ....................................................... 41
Figure 3.22 Holes at the edge of a wide stiffened plate reduce the axial stiffness (and critical
elastic buckling stress) but do not change the buckled shape. ....................................... 42
Figure 3.23 Accuracy of the stiffened element elastic buckling prediction method as a function of
hole offset hole versus plate width h for offset holes (a) without and (b) with the
dimensional limits in Eq. (3.8), Eq.(3.9), and Eq. (3.11) .................................................... 43
Figure 3.24 Boundary and loading conditions for a stiffened element in bending .......................... 43
Figure 3.25 Stiffened plates loaded with a linear bending stress gradient exhibit buckling of the
unstiffened strip adjacent to the hole in the compression region of the plate. ............. 45
Figure 3.26 Influence of slotted holes on critical elastic buckling stress fcr of stiffened elements in
bending as a function of (a) hole size relative to plate width and (b) hole spacing as a
function of hole length. ........................................................................................................ 46
Figure 3.27 Hole location influence on critical elastic buckling stress fcr for a stiffened plate in
bending (Y=0.50h) (Buckled mode shapes corresponding to A, B, C, and D are
provided in Figure 3.28.) ...................................................................................................... 47
Figure 3.28 The buckled mode shape changes as slotted holes move from the compression region
to the tension region of a stiffened element in bending (hhole/h=0.20). ........................... 48
Figure 3.29 Hole location influence on critical elastic buckling stress fcr for a stiffened plate in
bending (Y=0.75h) ................................................................................................................. 49
Figure 3.30 Derivation of stress ratio for unstiffened strip A. ................................................... 51
Figure 3.31 Derivation B and conversion of the compressive stress at the edge of unstiffened
strip B to the stress fcrB at the edge of the plate ............................................................. 52
Figure 3.32 Derivation of fcrh for the case when hA+hholeY (when the hole is located partially in
the compressed region and partially in the tension region of the plate) ....................... 53
Figure 3.33 Derivation of fcrh for the case when hA+hhole<Y (hole lies completely in the compressed
region of the plate). ............................................................................................................... 54
Figure 3.34 Influence of Lhole/yA on the accuracy of the prediction method for stiffened elements in
bending (a) without and (b) with the dimensional limits defined in Eq. (3.9), Eq.
(3.24), Eq. (3.25), and Eq. (3.26)............................................................................................56
Figure 3.35 Influence of hA/Y on the accuracy of the prediction method for stiffened elements in
(3.24), Eq. (3.25), and Eq. (3.26)............................................................................................56
Figure 3.36 Influence of S/h on the accuracy of the prediction method for stiffened elements in
(3.24), Eq. (3.25), and Eq. (3.26)............................................................................................57
Figure 3.37 Influence of S/Lhole on the accuracy of the prediction method for stiffened elements in
(3.24), Eq. (3.25), and Eq. (3.26)............................................................................................57
Figure 3.38 Influence of h/hhole on the accuracy of the prediction method for stiffened elements in
(3.24), Eq. (3.25), and Eq. (3.26)............................................................................................58
Figure 3.39 ABAQUS boundary and loading conditions for unstiffened plate loaded uniaxially.
................................................................................................................................................. 58
Figure 3.40 The presence of holes causes a decrease in critical elastic buckling load for
unstiffened plates in uniaxial compression. ...................................................................... 60
Figure 3.41 Buckled shapes of unstiffened plates with holes. ............................................................. 60
Figure 3.42 The critical elastic buckling stress of a stiffened plate decreases as holes are shifted
toward the simply supported edge (+hole) ....................................................................... 62
Figure 3.43 The critical elastic buckling stress for stiffened elements with (a) transversely offset
holes and (b) centered holes (from Section 3.5.2) decreases as a function of hole
length Lhole to hA ..................................................................................................................... 62
Figure 3.44 (a) Comparison of ABAQUS and empirical plate buckling coefficients for an
unstiffened element with holes and (b) ABAQUS to predicted elastic buckling stress
for an unstiffened element ................................................................................................... 64
Figure 4.1 C-section and hole dimension notation ............................................................................... 68
Figure 4.2 Columns are modeled with pinned warping-free boundary conditions and
compressed from both ends ................................................................................................ 68
Figure 4.3(a) SSMA 250S162-33 web plate and structural stud, and (b) SSMA 400S162-33 web
plate and structural stud ...................................................................................................... 70
Figure 4.4. Effect of a slotted hole on the elastic buckling load of simply supported plates and
structural studs ...................................................................................................................... 72
Figure 4.5 The presence of a hole creates unique local buckling modes where unstiffened strip
buckling adjacent to the hole occurs symmetrically (LH) or asymmetrically (LH2)
increase the distortional tendency of the flanges.............................................................. 74
Figure 4.6 SSMA slotted hole location and local buckling LH mode, L=48 in.,
x/L=0.06,0.125,0.25,0.375,0.50. Note the distortional tendencies of the flanges at the
hole. ......................................................................................................................................... 74
Figure 4.7 Influence of a slotted hole on the (a) distortional (D) and (b) global flexural-torsional
(GFT) modes of a cold-formed steel column ..................................................................... 75
Figure 4.8 Influence of SSMA slotted hole location on Pcr for a 362S162-33 C-section (refer to
Figure 4.5, Figure 4.6, and Figure 4.7 for buckled shape summaries) ........................... 76
Figure 4.9 Connection detail for structural stud to exterior wall requires a screw or bolt hole
placed in the stud flange (Western States Clay Products Association 2004) ................ 77
Figure 4.10 Influence of flange hole diameter on the local (L), distortional (D), and global (GFT)
elastic buckling loads of an SSMA 362S162-33 structural stud....................................... 78
xi
Figure 4.11 Local (L) buckling is dominated by flange and web deformation near the holes as
bhole/b exceeds 0.70 .................................................................................................................. 78
Figure 4.12 Experimental program boundary conditions as implemented in ABAQUS ................ 80
Figure 4.13 Influence of fixed-fixed boundary conditions versus warping free boundary
conditions on Pcrd for column experiments(L/H<4 ) as a function of (a) column length
to fundamental distortional half-wavelength calculated with CUFSM and (b) column
length to member length. .....................................................................................................87
Figure 4.14 Influence of fixed-fixed boundary conditions versus warping free boundary
conditions on Pcrl for column experiments ( L/H<4) as a function of (a) hole width
relative to column width and (b) hole length relative to column length ...................... 87
Figure 4.15 Influence of weak-axis pinned boundary conditions versus warping free boundary
conditions on (a) Pcrl as a function of hole length to column length and (b) Pcrd as a
function of column length to member length. .................................................................. 88
Figure 4.16 Rules for modeling a column net cross-section in CUFSM ............................................. 90
Figure 4.17 Local elastic buckling curve of net cross-section when (a) hole length is less than Lcrh
and (b) when hole length is greater than Lcrh..................................................................... 91
Figure 4.18 Comparison of CUFSM and ABAQUS predictions of unstiffened strip buckling. ...... 93
Figure 4.19 ABAQUS results verify CUFSM local buckling predictions for an SSMA 362S162-33
column with evenly spaced web holes. ............................................................................. 93
Figure 4.20 CUFSM and ABAQUS local buckling mode shapes are consistent when considering a
slotted flange hole. ................................................................................................................ 94
Figure 4.21 ABAQUS results verify CUFSM predictions for an SSMA 362S162-33 cross section
with evenly spaced flange holes. ........................................................................................ 95
Figure 4.22 ABAQUS predicts local plate buckling with distortional buckling interaction which is
not detected in CUFSM. ....................................................................................................... 96
Figure 4.23 ABAQUS results are slightly lower than CUFSM predictions, CUFSM predicts
correctly that plate local buckling controls over unstiffened strip buckling. ............... 96
Figure 4.24 Predicted Pcrh (CUFSM, buckling of the net cross-section) and Pcr (CUFSM, buckling
of the gross cross section, no hole) are compared relative to the ABAQUS Pcrl with
experiment boundary conditions as a function of (a) hole width to flat web width and
(b) hole length to column length ......................................................................................... 98
Figure 4.25 Predicted Pcrl (CUFSM approximate method) is compared relative to the ABAQUS
Pcrl with experiment boundary conditions as a function of (a) hole width to flat web
width and (b) hole length to column length...................................................................... 98
Figure 4.26 CUFSM approximate method for calculating Pcrd for a column with holes. ............... 100
Figure 4.27 Modified cross section to be used in CUFSM to predict Pcrd for a column with holes.
............................................................................................................................................... 101
Figure 4.28 ABAQUS boundary conditions and imposed rotations for web plate ........................ 102
Figure 4.29 Plate deformation from imposed edge rotations, hhole/h=0.50 ..................................... 103
Figure 4.30 Transverse rotational stiffness of the plate is significantly reduced in the vicinity of
the slotted hole .................................................................................................................... 104
Figure 4.31 Comparison of CUFSM and ABAQUS distortional buckling mode shapes. .............. 107
Figure 4.32 CUFSM distortional buckling prediction method is conservative when considering
an SSMA 262S162-68 column with uniformly spaced holes. ........................................ 107
Figure 4.33 Warping-fixed boundary condition amplification of Pcrd .............................................. 109
Figure 4.34 Accuracy of the CUFSM approximate method for predicting Pcrd improves as column
length increases relative to the fundamental distortional half-wavelength for
warping-fixed columns ...................................................................................................... 110
Figure 4.35 A weighted thickness cross section can be input directly into a program that solves
the classical cubic stability equation for columns (e.g. CUTWP). ................................ 113
xii
Figure 4.36 Weak-axis flexural and flexural-torsional global buckling modes for an SSMA
1200S162-68 column with evenly spaced circular holes. ............................................... 114
Figure 4.37 Variation in net section properties as circular hole diameter increases. ..................... 115
Figure 4.38 Comparison of weighted thickness and weighted properties cross-sectional area.
............................................................................................................................................... 116
Figure 4.39 Comparison of weighted thickness and weighted properties strong axis moment
of inertia. .............................................................................................................................. 117
Figure 4.40 Comparison of weighted thickness and weighted properties weak axis moment
of inertia. .............................................................................................................................. 117
Figure 4.41 ABAQUS boundary conditions for warping free and applied unit twist at x=0 in. and
warping free but rotation restrained at x=100 in. ........................................................... 119
Figure 4.42 Angle of twist decreases linearly in the SSMA 1200S162-68 column with warping free
end conditions. .................................................................................................................... 119
Figure 4.43 The weighted properties approximation for Javg matches closely with the ABAQUS
prediction for the SSMA 12S00162-68 column with holes ............................................. 120
Figure 4.44 ABAQUS boundary conditions for warping free and applied unit twist at x=0 in. and
warping fixed and rotation restrained at x=100 in. ........................................................ 121
Figure 4.45 Angle of twist is nonlinear along the SSMA 1200S162-68 column with warping fixed
end conditions at x=100 in. ................................................................................................ 122
Figure 4.46 Comparison of weighted thickness and weighted properties approximations to
the ABAQUS derived warping torsion constant Cw,avg. ................................................ 123
Figure 4.47 Comparison of weighted thickness and weighted properties prediction methods
for the SSMA 1200S162-68 weak-axis flexural buckling mode. Predictions using net
section properties are also plotted as a conservative benchmark. ............................... 124
for the SSMA 1200S162-68 flexural-torsional column buckling mode. Predictions
using net section properties are also plotted as a conservative benchmark. .............. 125
Figure 4.49 Cross section of beam specimen showing aluminum strap angles connected to Cflanges ................................................................................................................................... 127
Figure 4.50 C-section and hole dimension notation ........................................................................... 127
Figure 4.51 Experiment test setup with hole spacing, location of lateral bracing, spacing of
aluminum angle straps, and load points ......................................................................... 129
Figure 4.52 Finite element model boundary conditions for beam eigenbuckling analyses .......... 130
Figure 4.53 Channel and hole meshing details and modeling of aluminum angle straps ............ 131
Figure 4.54 ABAQUS meshing details for C-section rounded corners ............................................ 132
Figure 4.55 Modeling of the beam concentrated loads in ABAQUS ................................................ 133
Figure 4.56 Local buckling modes for specimen 2B,20,1&2(H) with and without holes .............. 135
Figure 4.57 Local buckling modes for specimen 3B,14,1&2(H) with and without holes .............. 135
Figure 4.58 Local buckling modes for specimen 6B,18,1&2(H) with and without holes ............. 136
Figure 4.59 Local buckling modes for specimen BP-40(H) with and without holes ..................... 136
Figure 4.60 Local buckling modes for specimen 12B,16,1&2(H) with and without holes ............ 137
Figure 4.61 Distortional buckling modes for specimen 2B,20,1&2(H) with and without holes .. 138
Figure 4.64 Distortional buckling modes for specimen BP5-40(H) with and without holes ......... 140
Figure4.65Distortionalbucklingmodesforspecimen12B,16,1&2(H)withandwithoutholes 140
Figure 4.66 Elastic buckling curve for 12 deep specimen with modal participation summarized,
note that selected L and D are mixed local-distortional modes ................................... 141
Figure 4.67 Possible global buckling mode occurs about the compression flange lateral brace
point ...................................................................................................................................... 142
xiii
Figure 4.68 Influence of holes on beam specimen Mcrl (Channel 1 and Channel 2 plotted)
considering (a) all local buckling modes and (b) the lowest local buckling mode .... 147
Figure 4.69 Influence of holes on beam specimen Mcrd (Channel 1 and Channel 2 plotted) as a
function of hole depth to flat web depth considering (a) all distortional buckling
modes and (b) the lowest distortional buckling mode .................................................. 148
Figure 4.70 Influence of holes on beam specimen Mcrd (Channel 1 and Channel 2 plotted) as a
function of web depth to flange width considering (a) all distortional buckling modes
and (b) the lowest distortional buckling mode ............................................................... 148
Figure 4.71 Influence of test boundary conditions on Mcrl................................................................ 149
Figure 4.72 Influence of test boundary conditions on (a) Mcrd and (b) on the distortional halfwavelength........................................................................................................................... 151
Figure 4.73 Boost in Mcrd from the angle restraints increases as the fundamental distortional halfwavelength increases relative to the restraint spacing Sbrace .......................................... 151
Figure 4.74 Guidelines for restraining beam net cross-sections in the CUFSM local buckling
approximate method .......................................................................................................... 154
Figure 4.75 Comparison of ABAQUS to predicted Mcrl for C-sections with holes in the beam
database as a function of (a) web depth and (b) hole width relative to flat web depth
............................................................................................................................................... 155
Figure 4.76 Comparison of mechanics-based and weighted-average prediction methods to
ABAQUS results for the distortional buckling load Mcrd of C-sections with holes in
the elastic buckling database ............................................................................................. 156
Figure 4.77 ABAQUS boundary conditions and applied loading for an SSMA 1200S162-68 beam
with holes (hhole/h=0.50 shown) ........................................................................................ 157
for the SSMA 1200S162-68 lateral-torsional beam buckling mode. Predictions using
net section properties are also plotted as a conservative benchmark. ......................... 158
Figure 5.1 Column testing parameters and naming convention....................................................... 161
Figure 5.2 Tested lengths of cold-formed steel columns with holes as a function of (a) column
length L and and (b) L versus out-to-out column width H ........................................... 163
Figure 5.3 Typical column specimens with slotted holes .................................................................. 164
Figure 5.4 Column test setup and instrumentation ............................................................................ 165
Figure 5.5 Novotechnik position transducer with ball-jointed magnetic tip .................................. 166
Figure 5.6 Central Machinery metal band saw used to rough cut column specimens .................. 167
Figure 5.7 362S162-33 short column specimen with bismuth end diaphragms .............................. 168
Figure 5.8 600S162-33 short column specimen oriented in CNC machine ...................................... 169
Figure 5.9 An end mill is used to prepare the column specimens .................................................... 169
Figure 5.10 The intermediate length specimens were end-milled in a manual milling machine. 170
Figure 5.11 The specimens are clamped at the webs only to avoid distortion of the cross-section
............................................................................................................................................... 170
Figure 5.12 Setup procedure for measuring specimen cross section dimensions .......................... 172
Figure 5.13 Procedure for measuring specimen cross-section dimensions ..................................... 173
Figure 5.14 Procedure for measuring flange-lip and flange-web angles ......................................... 174
Figure 5.15 Specimen measurement nomenclature ............................................................................ 175
Figure 5.16 Base metal and zinc thickness definitions ....................................................................... 177
Figure 5.17 Removal of tensile coupon zinc coating as a function of time ...................................... 179
Figure 5.18 A height gauge is used to measure specimen length ..................................................... 180
Figure 5.19 Lengths are measured at the four corners of the C-section column............................. 181
Figure 5.20 A dial gauge and precision stand are used to measure initial web imperfections .... 184
Figure 5.21 Web imperfection measurement grid and coordinate system ...................................... 184
Figure 5.22 Column specimen alignment schematic .......................................................................... 186
xiv
Figure 5.23 Influence of platen bending stiffness on end moments for a fixed-fixed eccentric
column .................................................................................................................................. 188
Figure 5.24 Column specimen weak axis out-of-straightness schematic ......................................... 189
Figure 5.25 Digital calipers are used to measure the distance from the column web to platen edge
............................................................................................................................................... 190
Figure 5.26 Tensile coupons are first rough cut with a metal ban saw ............................................ 192
Figure 5.27 Tensile coupon dimensions as entered in the CNC milling machine computer ........ 192
Figure 5.28 A custom jig allows three tensile coupons to be milled at once in the CNC machine
............................................................................................................................................... 193
Figure 5.29 ATS machine used to test tensile coupons....................................................................... 194
Figure 5.30 Gradually yielding stress-strain curve with 0.2% strain offset method ...................... 195
Figure 5.31 Sharp-yielding stress strain curve using an autographic method for determining Fy
............................................................................................................................................... 195
Figure 5.32 (a) Local and distortional elastic buckled mode shapes for (a) short (L=48 in.) 362S16233 specimens and (b) intermediate length (L=48 in.) 362S162-33 specimens. ............. 198
Figure 5.33 Local and distortional elastic buckled mode shapes for (a) short (L=48 in.) 600S162-33
specimens and (b) intermediate length (L=48 in.) 600S162-33 specimens. .................. 198
Figure 5.34 Local (L) and distortional (D) DSM strength predictions are similar in magnitude for
both 362S162-33 and 600S162-33 cross-sections, indicating that L-D modal interaction
will occur during the tested response of the columns. .................................................. 201
Figure 5.35 Comparison of global mode shapes for intermediate length 362S162-33 and 600S16233 specimens. ....................................................................................................................... 205
Figure 5.36 Load-displacement progression for short column specimen 362-2-24-NH ................ 207
Figure 5.37 Load-displacement progression for short column specimen 362-2-24-H ..................... 208
Figure 5.38 Load-displacement curve for a 362S162-33 short column with, without a slotted hole
............................................................................................................................................... 208
Figure 5.39 Comparison of load-deformation response and lateral flange displacements for
specimen 362-2-24-NH ....................................................................................................... 210
Figure 5.40 Influence of a slotted hole on 362S162-33 short column lateral flange displacement 210
Figure 5.41 Load-displacement progression for short column specimen 600-1-24-NH ................ 212
Figure 5.42 Load-displacement progression for short column specimen 600-1-24-H .................... 212
Figure 5.43 Comparison of load-displacement response for short 600S162-33 column specimens
with and without holes ...................................................................................................... 213
Figure 5.44 Load-displacement progression, intermediate length column specimen 362-3-48-NH
............................................................................................................................................... 215
Figure 5.45 Load-displacement progression for intermediate length column specimen 362-3-48-H
............................................................................................................................................... 215
Figure 5.46 Load-displacement curve, 362S162-33 intermediate column with and without a hole
............................................................................................................................................... 216
Figure 5.47 362S162-33 long column mid-height flange displacements show the global torsional
failure mode ......................................................................................................................... 216
............................................................................................................................................... 218
............................................................................................................................................... 218
Figure 5.50 Load-displacement comparison of intermediate length 600S162-33 specimens with
and without holes ............................................................................................................... 219
Figure 5.51 Short 600S162-33 column flange-lip corner lifts off platen during post-peak portion of
test ......................................................................................................................................... 222
Figure 6.1 Cold-formed steel roll-forming: (left) Sheet coil enters roll-forming line, (right) steel
sheet is cold-formed into C-shape cross-section (photos courtesy of Bradbury Group).
............................................................................................................................................... 224
xv
Figure 6.2 Forming a bend: plastic bending and elastic springback of thin sheets results in a
nonlinear through-thickness residual stress distribution. ............................................. 225
Figure 6.3 Stress-strain coordinate system as related to the coiling and cold-forming processes.
............................................................................................................................................... 226
Figure 6.4 Roll-forming setup with sheet coil fed from the (a) top of the coil and (b) bottom of
coil. The orientation of the coil with reference to the roll-forming bed influences the
direction of the coiling residual stresses. ......................................................................... 229
Figure 6.5 Coiling of the steel sheet may result in residual curvature which results in bending
residual stresses as the sheet is flattened. ........................................................................ 231
Figure 6.6 Longitudinal residual stress distribution from coiling. ................................................... 232
Figure 6.7 Predicted longitudinal residual stress distribution from coiling, uncoiling, and
flattening of a steel sheet. ................................................................................................... 233
Figure 6.8 Cold-forming of a steel sheet. .............................................................................................. 235
Figure 6.9 Fully plastic transverse stress state from cold-forming. .................................................. 235
Figure 6.10 Force couple (Fpt) applied to simulate the elastic springback of the steel sheet after
the imposed radial deformation is removed. .................................................................. 236
Figure 6.11 Cold-forming of a steel sheet occurs as plastic bending and elastic springback,
resulting in a self-equilibrating transverse residual stress. ........................................... 237
Figure 6.12 Plastic bending and elastic springback from cold-forming in the transverse direction
result in longitudinal residual stresses because of the plane strain conditions.......... 238
Figure 6.13 Flowchart summarizing the prediction method for residual stresses in roll-formed
members. .............................................................................................................................. 239
Figure 6.14 Plastic strain distribution from sheet coiling with a radius less than elastic-plastic
threshold rep.......................................................................................................................... 241
Figure 6.15 Effective plastic strain in a cold-formed steel member from sheet coiling when the
radius rx is less than the elastic-plastic threshold rep....................................................... 242
Figure 6.16 Effective von Mises true plastic strain at the location of cold-forming of a steel sheet.
............................................................................................................................................... 243
Figure 6.17 Flowchart summarizing the prediction method for effective plastic strains in rollformed members ................................................................................................................. 244
Figure 6.18 Coil coordinate system and notation. ............................................................................... 245
Figure 6.19 Influence of sheet thickness and yield stress on through-thickness longitudinal
residual stresses (z-direction, solid lines are predictions for mean coil radius, dashed
lines for mean radius +/- one standard deviation). ....................................................... 247
Figure 6.20 The mean-squared error of the predicted and measured bending residual stresses for
de M. Batista and Rodrigues (De Batista and Rodrigues 1992), Specimen CP1 is
minimized when rx=1.60rinner. ............................................................................................ 251
Figure 6.21 (a) Histogram and (b) scattergram of bending residual stress prediction error (flat
cross-sectional elements) for 18 roll-formed specimens. ............................................... 254
Figure 6.22 (a) Histogram and (b) scattergram of bending residual stress prediction error (corner
cross-sectional elements) for 18 roll-formed specimens. ............................................... 254
Figure 6.23 Definition of apparent yield stress, effective residual stress, and effective plastic
strain as related to a uniaxial tensile coupon test. .......................................................... 256
Figure 7.1 True stress-strain curve derived from a tensile coupon test (Yu 2005)........................... 260
Figure 7.2 Simply supported boundary conditions with equation constraint coupling at loaded
edges ..................................................................................................................................... 261
Figure 7.3 Application of (a) uniform load and (b) uniform displacement to a stiffened element
............................................................................................................................................... 262
Figure 7.4 Type 1 imperfection (Schafer and Pekz 1998) ................................................................. 263
Figure 7.5 Modified Riks method load-displacement solutions and failure modes ...................... 265
Figure 7.6 Correlation between initial imperfection shape and fold line locations at failure ....... 265
xvi
Figure 7.7 Artificial damping load-displacement solutions and failure modes ............................. 268
Figure 7.8 Stiffened element boundary conditions with rigid body coupling at loaded edges ... 269
Figure 7.9 Initial geometric imperfection field used for the stiffened element with and without a
hole ........................................................................................................................................ 270
Figure 7.10 Deformation at ultimate load of a stiffened element with a hole loaded in
compression. The common failure mechanism is material yielding adjacent to the
hole followed by plate folding. ......................................................................................... 271
Figure 7.11 Load-displacement curve for the RIKS1 and RIKS2 models showing direction
reversal along load path ..................................................................................................... 272
Figure 7.12 RIKS1 and RIKS2 models experience convergence problems and return along the
loading path, the RIKS3 model successfully predicts a peak load and finds a postpeak load path ..................................................................................................................... 274
Figure 7.13 STATIC1 and STATIC2 load-displacement curves demonstrate convergence
difficulties near the peak load. .......................................................................................... 275
Figure 7.14 STAB1 and STAB2 load-displacement curves demonstrate a highly nonlinear postpeak equilibrium path ........................................................................................................ 277
Figure 7.15 The STAB1 and STAB2 models (artificial damping, displacement control) exhibit a
sharp drop in load as folding of the plate initiates near the hole. The STAB3 model
(artificial damping, load control) finds the compressive load at which a complete loss
of stiffness occurs. ............................................................................................................... 279
Figure 7.16 Comparison of ultimate limit state and elastic buckling plate behavior, initial
imperfections are not considered in these results .......................................................... 280
Figure 7.17 Load-displacement sensitivity to imperfection magnitude for a plate without a hole
............................................................................................................................................... 282
Figure 7.18 Load-displacement sensitivity to imperfection magnitude for a plate with a slotted
hole ........................................................................................................................................ 282
Figure 7.19 Calculation of effective width at a cross-section along a stiffened element ............ 284
Figure 7.20 Definition of longitudinal (S11) membrane stress .......................................................... 284
Figure 7.21 (a) longitudinal membrane stresses and (b) effective width of a stiffened element at
failure .................................................................................................................................... 285
Figure 7.22 (a) longitudinal membrane stresses and (b) effective width of a stiffened element with
a slotted hole at failure ....................................................................................................... 286
Figure 7.23 Effective width comparison for a plate with and without a slotted hole .................... 286
Figure 7.24 Through the thickness variation of effective width of a plate without a hole ............ 287
Figure 7.25 Through the thickness variation of effective width of a plate with a slotted hole ..... 287
Figure 7.26 Through thickness variation in longitudinal (S11) stresses in a plate at failure......... 288
Figure 7.27 ABAQUS boundary conditions simulating column experiments ................................ 290
Figure 7.28 ABAQUS plastic strain curve for specimen 362-1-24-NH assuming (a) plasticity
initiates at the proportional limit and (b) plasticity initiates at 0.2% offset yield stress
............................................................................................................................................... 293
Figure 7.29 ABAQUS plastic strain curve for specimen 600-1-24-NH assuming (a) plasticity
initiates at the proportional limit and (b) plasticity initiates at the beginning of the
yield plateau (refer to Appendix H for the details on the development of this curve).
............................................................................................................................................... 293
Figure 7.30 Influence of ABAQUS material model on the load-deformation response of specimen
600-1-24-NH (work this figure with Figure 7.29) ........................................................... 294
Figure 7.31 Slotted holes are filled with S9R5 elements to obtain no hole imperfection shapes .. 295
Figure 7.32 L and D imperfection magnitudes described with a CDF (Schafer and Pekz 1998) 297
Figure 7.33 Method for measuring distortional imperfection magnitudes from experiments ..... 297
Figure 7.34 Definition of out-of-straightness imperfections implemented in ABAQUS ............... 298
Figure 7.35 ABAQUS element local coordinate system for use with residual stress definitions . 300
Figure 7.36 Transverse residual stress distribution applied at the corners of the cross-section... 300
xvii
Figure 7.37 Longitudinal residual stress distribution applied at the corners of the cross-section
............................................................................................................................................... 300
Figure 7.38 Equivalent plastic strain distribution at the corners of the cross-section .................... 301
Figure 7.39 Influence of section points on the unbalanced moment (accuracy) of the transverse
residual stress distribution as implemented in ABAQUS ............................................. 302
Figure 7.40 Load-displacement response of specimen 362-1-24-NH ............................................... 306
Figure 7.41 Load-displacement response of specimen 362-1-24-H ................................................... 306
Figure 7.48 Influence of residual stresses (RS) and plastic strains (PS) on the FE loaddisplacement response of specimen 600-1-24-NH .......................................................... 311
Figure 7.49 Influence of residual stresses (RS) and plastic strains (PS) on the FE loaddisplacement response of specimen 362-1-24-NH. ......................................................... 312
Figure 8.1 ABAQUS simulated column experiments boundary conditions and application of
loading .................................................................................................................................. 315
Figure 8.2 SSMA 800S250-97 structural stud with web holes considered in the DSM distortional
buckling study ..................................................................................................................... 318
Figure 8.3 SSMA 800S250-97 structural stud failure mode transition from distortional buckling to
yielding at the net section .................................................................................................. 320
Figure 8.4 Comparison of simulated column strengths (Anet/Ag=1.0) to (a) the existing DSM
distortional buckling design curve and to (b) the proposed DSM distortional
buckling curve for columns with holes ............................................................................ 321
global buckling design curve and to (b) the proposed DSM distortional buckling
curve for columns with holes ............................................................................................ 325
Figure 8.12 Comparison of column test-to-prediction ratios for columns (Anet/Ag=1.0) failing by
local-global buckling interaction as a function of (a) local slenderness (b) global
slenderness ........................................................................................................................... 328
xviii
Figure 8.13 Comparison of column test-to-prediction ratios for columns failing by local-global
buckling interaction with Pne calculated (a) without the influence of holes (b) and
with the influence of holes ................................................................................................. 330
Figure 8.14 Comparison of column test-to-prediction ratios for columns failing by local-global
buckling interaction as a function of Pynet/Pne where Pne is calculated (a) without the
influence of holes (b) and with the influence of holes ................................................... 331
Figure 8.15 SSMA 350S162-68 column failure mode changes from distortional-flexural torsional
buckling failure to weak axis flexure as hole size increases (L=34 in.) ........................ 331
Figure 8.16 SSMA 800S250-43 (L=74 in.) column web local buckling changes to unstiffened strip
buckling at peak load as hole size increases.................................................................... 332
Figure 8.17 Comparison of DSM local buckling design curve options when Pynet=0.80 Pyg and (a)
Pcre=100Pyg, (b) Pcre=5Pyg, and (c) Pcre=Pyg .......................................................................... 334
Figure 8.18 Comparison of simulated test strengths to predictions for columns with local
buckling-controlled failures as a function of local slenderness (tested strength is
normalized by Pne)............................................................................................................... 345
Figure 8.19 Comparison of simulated test strengths to predictions for columns with local
normalized by Pyg) .............................................................................................................. 346
Figure 8.20 Comparison of simulated test strengths to predictions for columns with distortional
buckling-controlled failures as a function of distortional slenderness ........................ 347
Figure 8.21 Comparison of simulated test strengths to predictions for columns with global
buckling-controlled failures (i.e., no local interaction) as a function of global
slenderness ........................................................................................................................... 348
Figure 8.22 Test-to-predicted ratios for local buckling-controlled simulated column failures as a
function of local slenderness ............................................................................................. 349
Figure 8.23 Test-to-predicted ratios for distortional buckling-controlled simulated column
failures as a function of distortional slenderness ........................................................... 350
Figure 8.24 Test-to-predicted ratios for simulated global buckling-controlled column failures (i.e.,
no local buckling interaction) as a function of global slenderness ............................... 351
Figure 8.25 Test-to-predicted ratios for simulated local buckling-controlled column failures as a
function of net cross-sectional area to gross cross-sectional area ................................. 352
Figure 8.26 Test-to-predicted ratios for simulated distortional buckling-controlled column
failures as a function of net cross-sectional area to gross cross-sectional area ........... 353
Figure 8.27 Test-to-predicted ratios for simulated global buckling-controlled column failures as a
Figure 8.28 Test-to-predicted ratios for simulated local buckling-controlled column failures as a
function of column length, L, to flat web width, h ......................................................... 355
Figure 8.29 Test-to-predicted ratios for simulated distortional buckling-controlled column
failures as a function of column length, L, to flat web width, h.................................... 356
Figure 8.30 Test-to-predicted ratios for simulated global buckling-controlled column failures as a
function of column length, L, to web width, h ................................................................ 357
Figure 8.31 Comparison of experimental test strengths to predictions for columns with local
normalized by Pne)............................................................................................................... 360
Figure 8.32 Comparison of experimental test strengths to predictions for columns with local
normalized by Py) ................................................................................................................ 361
Figure 8.33 Comparison of experimental test strengths to predictions for columns with
distortional buckling-controlled failures as a function of distortional slenderness .. 362
Figure 8.34 Comparison of experimental test strengths to predictions for columns with global
buckling-controlled failures as a function of global slenderness ................................. 363
xix
Figure 8.35 Test-to-predicted ratios for experiment local buckling-controlled column failures as a
function of local slenderness ............................................................................................. 364
Figure 8.36 Test-to-predicted ratios for experiment distortional buckling-controlled column
failures as a function of distortional slenderness .......................................................... 365
Figure 8.37 Test-to-predicted ratios for experiment global buckling-controlled column failures as
a function of global slenderness ........................................................................................ 366
failures as a function of net cross-sectional area to gross cross-sectional area ........... 368
a function of net cross-sectional area to gross cross-sectional area .............................. 369
function of column length, L, to flat web width, h ......................................................... 370
failures as a function of column length, L, to flat web width, h.................................... 371
a function of column length, L, to web width, h ............................................................. 372
Figure 8.44 ABAQUS simulated beam experiments boundary conditions and application of
loading .................................................................................................................................. 375
Figure 8.45 SSMA 800S162-43 beam with web holes considered in the DSM local buckling study
............................................................................................................................................... 378
Figure 8.46 Comparison of simulated beam strengths (Inet/Ig=1.0, no holes) to (a) the existing DSM
local buckling design curve and to (b) the proposed DSM local buckling curve for
beams with holes ................................................................................................................. 380
Figure 8.47 Comparison of simulated beam strengths (Inet/Ig=0.95) to (a) the existing DSM local
buckling design curve and to (b) the proposed DSM local buckling curve for beams
with holes ............................................................................................................................. 380
with holes ............................................................................................................................. 381
with holes ............................................................................................................................. 381
Figure 8.50 SSMA 550S162-54 structural stud failure mode transition from distortional buckling
to yielding at the net section .............................................................................................. 383
Figure 8.51 Comparison of simulated beam strengths (Inet/Ig=1.0) to (a) the existing DSM
buckling curve for beams with holes ............................................................................... 384
Figure 8.54 Comparison of simulated test strengths to predictions for laterally braced beams with
local buckling-controlled failures as a function of local slenderness ........................... 394
Figure 8.55 Comparison of simulated test strengths to predictions for laterally braced beams with
distortional buckling-controlled failures as a function of distortional slenderness .. 395
Figure 8.56 Test-to-predicted ratios for local buckling-controlled simulated laterally braced beam
failures as a function of local slenderness........................................................................ 396
xx
Figure 8.57 Test-to-predicted ratios for distortional buckling-controlled simulated laterally

braced beam failures as a function of distortional slenderness .................................... 397
Figure 8.58 Test-to-predicted ratios for simulated local buckling-controlled laterally braced beam
failures as a function of net cross-sectional moment of inertia to gross cross-sectional
moment of inertia ................................................................................................................ 398
Figure 8.59 Test-to-predicted ratios for simulated distortional buckling-controlled laterally
braced beam failures as a function of net cross-sectional moment of inertia to gross
cross-sectional moment of inertia ..................................................................................... 399
Figure 8.60 Test-to-predicted ratios for simulated local buckling-controlled laterally braced beam
failures as a function of column length, L, to beam depth, H ....................................... 400
Figure 8.61 Test-to-predicted ratios for simulated distortional buckling-controlled laterally
braced beam failures as a function of column length, L, to H ....................................... 401
Figure 8.62 Comparison of experimental test strengths to predictions for laterally braced beams
with local buckling-controlled failures as a function of local slenderness .................. 404
Figure 8.63 Comparison of experimental test strengths to predictions for laterally braced beams
with distortional buckling-controlled failures as a function of distortional slenderness
............................................................................................................................................... 405
Figure 8.64 Test-to-predicted ratios for experimental local buckling-controlled laterally braced
beam failures as a function of net cross-sectional moment of inertia to gross crosssectional moment of inertia ............................................................................................... 406
Figure 8.65 Test-to-predicted ratios for experimental distortional buckling-controlled laterally
braced beam failures as a function of net cross-sectional moment of inertia to gross
cross-sectional moment of inertia ..................................................................................... 407
Figure C.1 CUFSM finite strip modeling definition for an unstiffened element in compression 435
Figure C.2 The plate buckling coefficient kA for an unstiffened element in compression (the
multiple curves represent 0A1 with a step of 0.1, 11 curves total) ..................... 436
Figure C.3 The fitted curve for kA is a conservative predictor when Lhole/yA2 ............................. 437
Figure C.4 CUFSM finite strip modeling definition for an unstiffened element with compression
on the free edge, tension on the simply-supported edge .............................................. 438
Figure C.5 Variation in plate buckling coefficient kB for an unstiffened element with B ranging
from 0 to 10 .......................................................................................................................... 438
Figure C.6 Curve fit to minimum kB for B ranging from 0 to 10 ..................................................... 439
Figure C.7 Family of curves used to simulate boost in kB when Lhole/hB2, B ranges from 0 to 10
............................................................................................................................................... 440
Figure E.1 Long column with two holes spaced symmetrically about the longitudinal midline.445
xxi
ListofTables
Table 3.1 Plate widths corresponding to SSMA structural stud designations ..................................................... 26
Table 3.2 Parameter ranges in stiffened element verification study. .................................................................... 36
Table 3.3 Parameter range for stiffened element verification study with offset holes. ...................................... 41
Table 3.4 Parameter ranges considered for stiffened elements in bending with holes. ...................................... 44
Table 3.5 Study parameter limits for stiffened element in bending (Y/h=0.50) with offset holes .................... 46
Table 3.6 Parameter range for study of regularly-spaced holes on unstiffened elements. ................................ 59
Table 3.7 Parameter range considered for unstiffened element study with offset holes ................................... 61
Table 4.1 SSMA structural stud and plate dimensions ........................................................................................... 69
Table 4.2 Summary of column experimental data .................................................................................................. 79
Table 4.3 Fixed-fixed column experiment dimensions and material properties................................................. 82
Table 4.4 Fixed-fixed column experiment elastic buckling properties ................................................................. 83
Table 4.5 Weak-axis pinned column experiment dimensions and material properties ..................................... 84
Table 4.6 Weak-axis pinned column experiment elastic buckling properties ..................................................... 84
Table 4.7 Parameter ranges for fixed-fixed and weak-axis pinned column specimens with holes .................. 85
Table 4.8 DSM prequalification limits for C-sections ............................................................................................. 85
Table 4.9 DSM prequalification limits for beam C-sections................................................................................. 143
Table 4.10 Parameter ranges for beam specimens with holes ............................................................................. 143
Table 4.11 Beam experiment cross-section dimensions, material properties, and tested strengths ............... 144
Table 4.12 Beam experiment elastic buckling results ........................................................................................... 145
Table 5.1 FSM local and distortional buckling half-wavelengths for nominal 362S162-33 and 600S162-33
cross-sections.................................................................................................................................................... 162
Table 5.2 Voltage conversion factors for column test instrumentation .............................................................. 166
Table 5.3 Summary of measured cross section dimensions................................................................................. 175
Table 5.4 Summary of measured lip-flange and flange-web cross section angles ............................................ 176
Table 5.5 Specimen bare steel and zinc coating thicknesses ................................................................................ 178
Table 5.6 Measured column specimen length ....................................................................................................... 181
Table 5.7 Specimen end flatness .............................................................................................................................. 182
Table 5.8 Measured slotted hole dimensions and locations ................................................................................ 183
Table 5.9 Initial web imperfections (deviations from the average elevation of the web) ................................ 185
Table 5.10 Specimen gross centroid and offset from applied load during tests ............................................... 187
Table 5.11 Summary of out-of-straightness calculations ..................................................................................... 190
Table 5.12 Voltage conversion factors for tensile coupon testing ....................................................................... 194
Table 5.13 Summary of column specimen steel yield stress ................................................................................ 195
Table 5.14 Column specimen steel yield stress statistics...................................................................................... 196
Table 5.15 Critical elastic buckling loads, influence of holes on elastic buckling ............................................. 200
Table 5.16 Specimen ultimate strength results ...................................................................................................... 206
Table 6.1 Statistics of the residual stresses in roll-formed members .................................................................. 249
Table 6.2 Radial location in the coil that minimizes the sum of the mean square prediction error for rollformed members .............................................................................................................................................. 252
Table 7.1 Summary of nonlinear finite element models and associated solution controls .............................. 271
Table 7.2 Local and distortional imperfection magnitudes ................................................................................. 297
Table 7.3 Out-of-straightness imperfection magnitudes ...................................................................................... 298
Table 7.4 Comparison of nonlinear FE simulation peak loads to experiments ................................................. 303
Table 8.1 DSM test-to-predicted statistics for column simulations .................................................................... 344
Table 8.2 DSM test-to-predicted ratio statistics for column experiments .......................................................... 359
Table 8.3 DSM test-to-predicted ratio statistics for column experiments (stub columns only) ....................... 359
Table 8.4 DSM test-to-predicted statistics for laterally braced beam simulations ............................................ 393
Table 8.5 DSM test-to-predicted ratio statistics for column experiments .......................................................... 403
xxii
Chapter 1
Introduction
The goal of this research work is to develop a general design method for cold
formedsteelstructuralmemberswithholes.Coldformedsteelbeamsandcolumnsare
typically manufactured with perforations. For example, in low and midrise
construction, holes are prepunched in structural studs to accommodate the passage of
utilities in the walls and ceilings of buildings as shown in Figure 1.1. In coldformed
steelstoragerackcolumns,perforationpatternsareprovidedtoallowforvariableshelf
configurations as shown in Figure 1.2. (Members with discrete holes, for example C
sectionswithpunchedholesasshowninFigure1.1,aretheresearchfocusinthisthesis,
althoughmanyofthetoolsandmethodsdevelopedherecanbeextendedtoperforation
patternsinstoragerackswithadditionalresearcheffort.)Existingdesignproceduresfor
coldformed steel members with holes are limited to certain hole sizes, shapes, and
configurations. These limitations can hamper an engineers design flexibility and
decrease the reliability of coldformed steel components where holes exceed these
prescriptivelimits.

Figure1.1Perforationsareprovidedinstructuralstudstoaccommodateutilitiesinthewallsofbuildings
Figure1.2Holepatternsinstoragerackcolumns
ThebasicframeworkofthedesignproceduredevelopedinthisthesisistheDirect
Strength Method (DSM) (AISIS100 2007, Appendix 1). DSM is relatively new and
represents a major advancement in coldformed steel design because it provides
engineersandcoldformedsteelmanufacturerswiththetoolstopredictthestrengthofa
memberwithanygeneralcrosssection.Coldformedsteelmembersaremanufactured
from thin sheet steel, and therefore member resistance is influenced by crosssection
instabilities(e.g.,platebucklinganddistortionofopencrosssections)inadditiontothe
globalbucklinginfluenceconsideredinthickerhotrolledsteelsections.DSMexplicitly
defines the relationship between elastic buckling and loaddeformation response with
empiricalequationstopredictultimatestrength.
To calculate the capacity of a coldformed steel member with DSM, the elastic
buckling properties of a general coldformed steel crosssection are obtained from an

elasticbucklingcurve.Thecurvecanbegeneratedwithsoftwareemployingthefinite
stripmethodtoperformaseriesofeigenbucklinganalysesoverarangeofbuckledhalf
wavelengths. In this research work the freely available program CUFSM is utilized
(Schafer and dny 2006). An example of an elastic buckling curve is provided in
Figure1.3foracoldformedsteelCsectioncolumnandhighlightsthethreecategoriesof
elastic buckling considered in DSM local buckling, distortional buckling, and global
buckling. Local buckling occurs as plate buckling of individual slender elements in a
crosssection. Distortional buckling exists only for open crosssections such as a C
section,wherethecompressedflangesbuckleinwardoroutwardalongthelengthofa
member. Global buckling, also known as Euler buckling, defines buckling of the full
member at long halfwavelengths including both flexural and flexuraltorsional effects
(andlateraltorsionaleffectsinbeams).
The critical elastic buckling loads associated with local, distortional, and global
bucklingPcrl,Pcrd,andPcreforcolumns(Mcrl,Mcrd,andMcreforbeams),canbeobtained
directlyfromtheelasticbucklingcurve.Thecriticalelasticbucklingloadsarethenused
topredicttheultimatestrengthwiththreeempiricaldesigncurvespresentedinFigure
1.4 to Figure 1.6 for coldformed steel columns. (The current DSM column design
equations for members without holes are also provided in these figures.) The local,
distortional, and global slenderness of a member (l, d, c) are calculated from the
criticalelasticbucklingloads,definingamemberssensitivitytoeachtypeofbucklingat
failure (high slenderness corresponds to high sensitivity, low slenderness to low

sensitivity). The nominal resistances (Pnl, Pnd, and Pne) are obtained by inserting the
slenderness magnitudes into the DSM design equations. The minimum of the local,
distortional,andglobalnominalstrengthsistakenasthestrengthofthemember.
20
18
P (kips)
loadcrfactor
Distortional
buckling
Local buckling
16
14
12
10
PPcrd
Global
buckling
crd
6
4
Pcre
Pcrl P
crl
2
0
10
Pcre
10
10
10
half-wavelength
half-wavelength (in.)
Figure1.3ColumnelasticbucklingcurvegeneratedwithCUFSM
1.5
Flexural, Torsional, or Torsional-Flexural Buckling
The nominal axial strength, Pne, for flexural or torsional- flexural buckling is
2
for c 1.5 Pne = 0.658 c Py
1
Pne /Py
0.877
for c > 1.5 Pne = 2 Py = 0.877 Pcre

c
where
Py
= AgFy
Py Pcre
Pcre= Critical elastic global column buckling load

Ag
0.5
0.5
= gross area of the column
1.5
2.5
3.5
Global slenderness, c =(Py /P cre)0.5
Figure1.4DSMglobalbucklingfailuredesigncurveandequations
1.5
Local Buckling
The nominal axial strength, Pnl, for local buckling is
for l 0.776 Pnl = Pne
P
for l > 0.776 Pnl = 1 0.15 crl
Pne
where
0.4
P
crl
Pne
0.4
Pne
Pne Pcrl
Pn /Pne
Pcrl = Critical elastic local column buckling load
Pne =
Nominal axial strength for global buckling
0.5
Global
failure
0
0.5
Local buckling interacts with

global buckling at failure
1.5
2.5
3.5
local slenderness, =(Pne/P cr )0.5

l
l
Figure1.5DSMlocalbucklingfailuredesigncurveandequations
1.5
Distortional Buckling
The nominal axial strength, Pnd, for distortional buckling is
for d 0.561 Pnd = Py
for d > 0.561 Pnd = 1 0.25 crd

Py
1
Pnd /Py
where
0.6
Pcrd
P
y
0.6
Py
= Py Pcrd
Pcrd = Critical elastic distortional column buckling load

Py = Column yield strength
0.5
0.5
1.5
2.5
3.5
Distortional slenderness, d=(Py /Pcrd)0.5
Figure1.6DSMdistortionalbucklingfailuredesigncurveandequations
This research aims to extend the appealing generality of DSM to coldformed steel
memberswithperforations.Theprimaryresearchgoalsaretostudyandquantifythe
influenceofholesontheelasticbucklingofcoldformedsteelbeamsandcolumnsand
thentodevelopmodificationstotheexistingDSMdesignequationswhichrelateelastic
bucklingtoultimatestrength.Theresearchplanisimplementedinthreephases:
PhaseI(Chapters24)
1. Studytheinfluenceofholesontheelasticbucklingofthinplates,andthenoncold
formedsteelbeamsandcolumns.
2. Evaluate the viability of DSM for members with holes by comparing existing
experimentsonmemberswithholestothecurrentDSMspecification.
PhaseII(Chapters57)
1.Conductexperimentsoncoldformedsteelcolumnstoobservetheinfluenceofholes
onultimatestrengthandpostbucklingresponse.
2.Defineandvalidateanonlinearfiniteelementmodelingprotocolthroughparameter
studiesonthinplatesandcomparisontoexperimentalresults.
PhaseIII(Chapters78)
1. Formalize the relationship between elastic buckling and ultimate strength for
memberswithholesusingnonlinearfiniteelementsimulationsandexistingdata.
2.ModifythecurrentDSMspecificationtoaccountformemberswithholes
Phase I research is primarily focused on elastic buckling. Chapter 2 describes
preliminary thin shell finite element eigenbuckling studies which are used to evaluate
the accuracy of different shell element types in ABAQUS and to define finite element
meshingguidelines.Chapter3extendsthiselasticbucklingresearchwitheigenbuckling
analysesoftypicalcrosssectionalelementsconsideredincoldformedsteeldesign.For
example, a stiffened element is a simplysupported plate used to model the web of a

coldformedsteelCsectionandanunstiffenedelementisaplatesimplysupportedon
threeedgesandfreeonthefourthedgetosimulatethebehaviorofthefreelegofacold
formed steel hat section. Chapter 4 examines the elastic buckling of full coldformed
steelbeamsandcolumnswithholesanddevelopsusefulsimplifiedmethodstopredict
elastic buckling, including the influence of holes, without finite element analysis. The
elasticbucklingpropertiesofexistingbeamandcolumnexperimentsarealsocalculated
andmergedwiththetestedstrengthsintoadatabase.Thisdatabaseisemployednear
the end of the project to validate the proposed modifications to the DSM design
equationsformemberswithholes.
PhaseIImarksashiftfromelasticbucklingtothestudyoftheinfluenceofholeson
loaddeformationresponseandultimatestrength.Chapter5describesanexperimental
program on short and intermediate length coldformed steel columns with holes.
Chapter 6 initiates the development of a nonlinear finite element protocol with a
significant effort to define the residual stresses and initial plastic strains from the
manufacturing process. The capabilities of the commercial program ABAQUS
(ABAQUS2007a)areexploredatthebeginningofChapter7withpreliminarynonlinear
finite element simulation studies on rectangular plates with holes. The experimental
results from Chapter 5 are then employed in Phase III to fully develop and verify the
modeling protocol. The research culminates in Chapter 8 with the development of a
database of simulated tests which are used in combination with existing experimental
datatovalidatetheDSMdesignmethodforcoldformedsteelmemberswithholes.
Chapter 2
Thin-shell finite element modeling
in ABAQUS
A set of ABAQUS modeling guidelines is formalized in this chapter to provide a
consistentmethodologyforthefiniteelementstudiesconductedinthisthesisresearch.
Finiteelementeigenbucklinganalysisisavaluabletoolforstudyingtheelasticbuckling
properties of thinwalled structures. The accuracy of an analysis is influenced by
decisions made while implementing the finite element model, including the choice of
finiteelementtypeandthemeshinggeometryanddensity.Studiesarepresentedhere
which compare finite element eigenbuckling predictions of plate buckling problems to
known theoretical solutions. The eigenbuckling analyses are performed with the
commercial finite element program ABAQUS (ABAQUS 2007a). The accuracy of
ABAQUSthinshellelementsareevaluated,andfiniteelementconvergencestudiesare
presentedtoidentifylimitsonelementaspectratio.Rulesformodelingroundedcorners
andmeshingaroundholesarealsoprovidedwithsupportingelasticbucklingstudies.
2.1 ComparisonofABAQUSthinshellelements
ThreeABAQUSfiniteelementscommonlyemployedintheelasticbucklinganalysis
ofthinwalledstructuresaretheS9R5,S4,andS4RelementsasshowninFigure2.1.The
S4 and the S4R finite elements are four node general purpose shell elements valid for
both thick and thin shell problems (ABAQUS 2007a). Both elements employ linear
shape functions to interpolate deformation between nodes. The S9R5 element is a
doublycurved thin shell element with nine nodes derived with shear flexible Mindlin
strain definitions and Kirchoff constraints (classical plate theory with no transverse
sheardeformation)enforcedaspenaltyfunctions(Schafer1997).Thiselementemploys
quadraticshapefunctions(resultingfromtheincreaseinthenumberofnodesfrom4to
9)whichprovidetwoimportantbenefitswhenmodelingthinwalledstructures:(1)the
abilitytodefineinitiallycurvedgeometriesand(2)theabilitytoapproximateahalfsine
wave with just one element. The 5 in S9R5 denotes that each element node has 5
degreesoffreedom(threetranslational,tworotational)insteadof6(threetranslational,
three rotational). The rotation of a node about the axis normal to the element mid
surfaceis removedfromtheelementformulationtoimprovecomputationalefficiency.
The R in the S9R5 (and S4R) designation denotes that the calculation of the element
stiffnessisnotexact;thenumberofGaussianintegrationpointsisreducedtoimprove
computational efficiency and to avoid shear locking. This reduced integration
approach underestimates element stiffness and sometimes results in artificial element
deformation modes with zero strain across the element, commonly referred to as
hourglassmodes(Schafer1997).Theaccuracyofeigenbucklingfiniteelementmodels
arecomparedhereforeachoftheseABAQUSelementtypesagainsttheexactsolutions
fortwocommonplatebucklingproblems.
S9R5
S4/S4R
Figure2.1ABAQUSS4\S4Rshellelementwithfournodesandalinearshapefunction,ABAQUSS9R5shell
elementwithninenodesandaquadraticshapefunction
2.1.1 Modeling accuracy for a stiffened element
Elastic buckling analyses of a stiffened element were performed in ABAQUS to
compare the accuracy of the ABAQUS S9R5, S4, and S4R elements against a known
solution. A stiffened element is a common term used in thinwalled structures to
describe a crosssectional element restrained on both edges (see Figure 3.1) which is
approximated as a thin simplysupported plate (with sides free to wave) and loaded
uniaxiallyasshowninFigure2.2.
fcr
L
h
Figure2.2Buckledshapeofastiffenedplate
10
Thetheoreticalbucklingstressforastiffenedelementis:
2
2E t
f cr = k
,
12(1 2 ) h
(2.1)
wherehisthewidthoftheplate,Eisthemodulusofelasticityoftheplatematerial, is
thePoissonsratio,andtisthethicknessoftheplate.
Thebucklingcoefficientkis:
2
mh n 2 L
,
k =
+
mh
L
(2.2)
whereListhelengthoftheplateandmandnarethenumberofhalfwavelengthsinthe
longitudinal and transverse directions, respectively (Chajes 1974). In Figure 2.2, m=4
andn=1.
Plate buckling coefficients (k) are approximated in ABAQUS by performing
eigenbucklinganalysesofstiffenedelementswithABAQUSS4,S4R,andS9R5elements.
The element aspect ratio is set at 8:1 for the S9R5 element and 4:1 for the S4R and S4
elementstoensureaconsistentcomparisonbetweenfiniteelementmodels(i.e.,similar
numbersofnodesandcomputationaldemand).Theseparticularelementaspectratios
were also chosen because they are expected to be towards the upper limit of what is
requiredtodiscretizethegeometryofcoldformedsteelmembers(especiallyatrounded
cornerswheretheelementaspectratiocanbequitehigh).Theplatethicknessissetto
t=0.0346 in. E=29500 ksi and =0.30 for all finite element models. The ABAQUS
boundaryandloadingconditionsareimplementedasshowninFigure3.1.
11
Figure 2.3 compares the theoretical k from Eq. (2.1) to the ABAQUS buckling
coefficientsforvaryingplateaspectratios(L/h).TheS9R5elementperformsaccurately
overtherangeofelementaspectratiosconsidered,withamaximumerrorof1.3percent.
TheS4andS4Relementsarenotasaccurate,withmaximumerrorsof11.4percentand
9.7percent,respectively.TheaccuracyoftheplatemodelswithS4andS4Relements
increase with increasing plate aspect ratio, which indirectly implies that solution
accuracy increases as the number of elements per halfwave increase (in the loaded
direction).Thishypothesisisconsistentwiththeelementformulations,sincetheS9R5
element uses a quadratic shape function to estimate displacements (and can therefore
capture the halfsine wave of a buckled plate with as little as one element) and the S4
and S4R elements use linear shape functions (requiring at least three elements to
coarsely estimate the shape of a half sine wave). The S4R element is observed to be
slightlylessstiffthantheS4elementinFigure2.3,whichishypothesizedtooccurasa
resultofthereducedintegrationstiffnessapproximation.
Comparing the number of elements required to model a buckled halfwave is a
moreusefulindicatorofmeshdensityandmodelaccuracythanjusttheelementaspect
ratio alone. Figure 2.4 verifies this supposition by demonstrating the improvement in
modeling accuracy for a stiffened element as the number of finite elements per square
halfwaveincrease.TheS4elementexperiencesmembranelockingwhenthenumberof
elementsperhalfwaveislessthan2,resultinginexceedinglyunconservativevaluesfor
k. The S4R avoids this membrane locking with a reduced integration scheme that
assumes the membrane stiffness is constant in the element (ABAQUS 2007a).
12
Regardless, the accuracy of the S4R element degrades when less than 5 elements per
halfwavelengthareusedandneitherfournodeelement(i.e.,theS4ortheS4R)isableto
capturethesinusoidalshapeofthebuckledhalfwavewithlessthanthreeelementsper
buckled halfwave. The S9R5 accurately predicts the shape of the buckled halfwave
and the buckling coefficient k with just one element. k is within 2.1 percent of the
theoretical value for one element per halfwave and reduces to 0.1 percent for two
elementsperhalfwave.
6
S4
S4R
S9R5
Theory
plate buckling coefficient, k
5.5
S4\S4R (4:1)
S9R5 (8:1)
5
L/h=2 shown
4.5
3.5
0.5
1.5
2
2.5
plate aspect ratio, L/h
3.5
Figure2.3AccuracyofABAQUSS9R5,S4,andS4Relementsforastiffenedelementwithvaryingaspect
ratios,8:1finiteelementaspectratiofortheS9R5element,4:1elementaspectratiofortheS4andS4R
elements
13
7
S4
S4R
S9R5
6.5
buckling coeff., k
6
5.5
5
4.5
4
3.5
3
6
8
10
12
14
16
number of elements per buckled half-wave
18
20
Figure2.4AccuracyofS4,S4R,andS9R5elementsasafunctionofthenumberofelementsprovidedper
buckledhalfwavelength,stiffenedelement,squarewaves(k=4)
2.1.2 Modeling accuracy for an unstiffened element
An unstiffened element is another common crosssection component considered in
theelasticbucklingofthinwalledcrosssections(seeFigure3.1),thebehaviorofwhich
isconservativelyapproximatedasaplatesimplysupportedonthreesidesandfreeon
thefourthsideparalleltothedirectionofauniaxiallyappliedstress.Thebuckledshape
ofanunstiffenedelementisdepictedinFigure2.5.
fcr
L
h
14
Figure2.5Buckledshapeofanunstiffenedelement,m=1shown
Thetheoreticalbucklingcoefficient k ofanunstiffenedelementcanbecalculatedwith
thenumericalsolutionofthefollowingequations(Timoshenko1961):
2
m 2 2
m 2 2
tanh(h ) ,
(
)
tanh
h
=
+
L2
L2
(2.3)
m 2 2 m 2 1 2 2
m 2 2 m 2 1 2 2
k ,and = 2 +
k .
= 2 +
Lh
L
Lh
L
(2.4)
Figure 2.6 compares the theoretical to predicted k versus the number of S9R5
elementsprovidedalongthelengthLofanunstiffenedelement.Theplatedimensions
areheldconstantatL/h=4,whiletheelementaspectratioisvariedfrom1:1to64:1.The
S9R5elementproducesanerrorof4.3percentwithanelementaspectratioof16:1and
anerrorof1.0percentwithanelementaspectratioof8:1.
0.8
0.7
buckling coefficient, k
0.6
8:1
0.5
4:1
1:1
2:1
16:1
0.4
32:1
0.3
element aspect
ratio (typ.)
0.2
0.1
0
k=0.486 when
L/h=4
64:1
0
10
20
30
40
50
number of S9R5 elements along length of plate
60
70
Figure2.6AccuracyofS9R5elementsasthenumberoffiniteelementsprovidedalonganunstiffened
elementvaries,L/h=4
15
2.2 ModelingholesinABAQUS
The ability to incorporate holes into the geometry of an ABAQUS finite element
modelisakeyprerequisitetostudyingtheinfluenceofholesonthestructuralbehavior
ofcoldformedsteelstructuralmembers.Toclearthishurdle,customMatlabcodewas
writtenbytheauthorwhichgeneratesafiniteelementmeshofaplatecontainingahole
(Mathworks2007).Thecodediscretizesthegeometryaroundaholebycreatinglayers
ofS9R5elementsasshowninFigure2.7foraslottedhole,acircularhole,andasquare
hole. (See Appendix A for a description of the custom mesh generation program.
Additional Matlab tools were developed to integrate the hole mesh geometry into an
existing finite element model.) The discretization results in S9R5 elements with
opposite edges which are not initially parallel. The initial geometry of 9 node
quadrilateralelementswithoutparalleledgescanbedefinedwithoutlossofaccuracyas
longasthemidlinenodesremaincenteredbetweenthecornernodes(Cook1989),which
is an advantage over the S4 and S4R elements. ABAQUS recommends that the angle
betweenisoparametriclines(i.e.,corneranglesofanelement)shouldnotbelessthan45
degrees or greater than 135 degrees to ensure accurate numerical integration of the
elementstiffnessmatrix(ABAQUS2007a).Thislimitcoincideswiththeminimumand
maximum S9R5 corner angles for the elements at the bisection of the 90 degree plate
cornersasshowninFigure2.8.
ThisstudyestablishesABAQUSS9R5finiteelementmeshguidelinesforplateswith
holesbystudyingtheconvergenceoftheelasticstabilitysolutionaselementaspectratio
16
varies. Figure 2.7 provides the typical mesh layout and summarizes the plate
dimensions considered in this study. The plate is modeled as a stiffened element in
ABAQUS, simply supported on four sides and loaded uniaxially in compression (see
Figure 3.5 for the ABAQUS implementation of the boundary and loading conditions).
Theplatethicknesst=0.0346in.,E=29500ksi,and=0.30forallfiniteelementmodels.
L=3.4 in.
L=6.0 in.
h=3.4 in.
hhole=1.5 in.
Lhole=4.0 in.
Figure2.7.Finiteelementmeshandplatedimensions:slotted,rectangular,andcircularholes
Theconvergenceoftheelasticbucklingsolutionfortheplateswithholesisstudied
by varying the S9R5 element aspect ratio (a:b) at the bisection of the plate corners as
showninFigure2.8,whereaandbaredefinedas
a=
h hhole
h
, b = hole .
N elem
N layers 2
(2.5)
Theaspectratioisvariedbyincreasingthenumberoffiniteelementlayersaroundthe
hole(Nlayers)whilethemaintainingthenumberofedgeelements(Nelem)constant(i.e.,the
meshdensityincreasesbuttheelementcorneranglesremainconstant).
Nelem =10
45
Nlayers =4
135
Element aspect ratio is a:b

Nlayers number of element layers around hole
Nelem number of edge elements
17
Figure2.8HolediscretizationusingS9R5elements
Figure 2.9 demonstrates that the critical elastic buckling stress for the lowest
bucklingmode(ahalfsinewaveinthiscase)forallholetypesconvergestoaconstant
magnitude when a:b is between 0.5 and 2. This result is employed as a modeling
guidelinefortheresearchworkinthisthesiswiththeexpressionsforaandbinEq.(2.5):
0.5
N elem
N layers
1 2 .
2 hhole
(2.6)
1.2
circular hole
square hole
SSMA slotted hole
fcr,hole/fcr,no hole
0.8
0.6
0.4
0.2
0.5
1.5
2
2.5
S9R5 element aspect ratio, a/b
3.5
Figure2.9ThecriticalelasticbucklingstressconvergestoaconstantmagnitudewhentheS9R5element
aspectratioa/bisbetween0.5and2andelementcorneranglesareskewed
2.3 ModelingRoundedCornersinABAQUS
TheS9R5elementcanbedefinedwithaninitialcurvedgeometryinABAQUSwhich
makesitconvenientformodelingroundedcornersofacoldformedsteelcrosssection.
ABAQUSrecommendsthattheinitialelementcurvatureshouldbelessthan10degrees,
18
where curvature of an S9R5 element is defined as the angle subtended by the nodal
normal and the average element normal as shown in Figure 2.10. The derivation in
Figure2.10demonstratesthatthiscurvaturerecommendationismetwhenfiveormore
S9R5 elements form the 90 degree corner. This limit is unfavorable from a modeling
perspectivebecausetheelementaspectratioincreasesasthenumberofelementsaround
thecornerincrease,anotherpotentialsourceofaccuracydegradation.Also,forafinite
elementmodelwithfour90degreecorners(e.g.,acoldformedsteellippedCsection),
increasing the number of elements at a corner can result in a considerable increase in
computational demand if the corner elements comprise a large proportion of the total
numberofelementsinacrosssection.
Average S9R5
element normal
S = r ,
S
S
Node normal (typ.)
ABAQUS recommends 10 degrees

to limit initial S9R5 element curvature
r
18
18
, S 90 =
N elem
r
2
S 90
, N elem 4.5
2S
Figure2.10ABAQUSS9R5initialcurvaturelimitrequiresatleastfiveelementstomodelcorner
AparameterstudywasconductedtoevaluatetheinfluenceofthenumberofS9R5
elements making up a 90 degree corner on the critical elastic buckling loads for local
buckling(Pcrl),distortionalbuckling(Pcrd),andglobalbuckling(Pcre)ofanSSMA600S162
68Csectioncolumn.Thenumberofcornerelementswerevariedfrom1to5,withthe
associated S9R5 aspect ratio a:b varying from 5 to 22. The column length was held
constant at L=48 inches for all models to accommodate multiple local and distortional
halfwaves.Thecolumnswereloadeduniaxiallyandmodeledwithwarpingfreeends
19
(CUFSMstyleboundaryconditions)asshowninFigure4.2.E=29500ksiand=0.30for
allfiniteelementmodels.Pyisthesquashloadofthecolumncalculatedwiththesteel
yieldstressFy=50ksi.
Figure 2.11 provides the typical mesh geometry of the column and compares a C
sectioncornermodeledasasmoothsurfacewithoneS9R5elementandwiththreeS9R5
elements.Figure 2.12 demonstrates that the number of S9R5 corner elements has a
minimal influence on the elastic buckling behavior of the column, with a slight
decreasingtrend(lessthan1%)incriticalelasticbucklingloadwithincreasingelement
quantity.Meshrefinementatthecornersdoesnotinfluencesolutionaccuracybecause
elastic buckling deformation occurs primarily within the more flexible crosssectional
elements.Ifthesimulationofsharpfoldingofthecornersisrequired,suchasinthecase
of nonlinear finite element modeling to collapse, additional corner elements may be
warrantedtoaccuratelycapturelocalizeddeformationgradients.
Figure2.11SSMA600S16268Csectioncornermodeledwitha)oneS9R5element,b)threeS9R5elements
20
1
0.9
0.8
0.7
Pcr/Py
0.6
0.5
0.4
Global - flexural torsional
Global - weak axis flexure
Distortional
Local
0.3
0.2
0.1
0
2
3
4
Number of corner S9R5 elements
Figure2.12ThenumberofS9R5cornerelementshasaminimalinfluenceonthecriticalelasticbuckling
loadsofanSSMA60016268CsectioncolumnwithL=48in.
2.4 Summaryofmodelingguidelines
TheS9R5elementwillbeimplementedinthisresearchworkbasedonitsversatility
anddemonstratedaccuracy.TheresultsoftheABAQUSstudiesinthischapterformthe
basis of the ABAQUS modeling guidelines below which will be implemented for both
eigenbucklingandnonlinearfiniteelementstudiesinthisthesis:
A minimum of two S9R5 elements per halfwavelength shall be provided in

stiffenedelementsinthedirectionnormaltotheappliedload(e.g.,flangesand
webofalippedCsection)
The S9R5 element aspect ratio shall be less than or equal to 8:1 in unstiffened
elements(e.g.,flangelipinaCsection)
TheS9R5elementaspectratioshallbebetween0.5and2.0whenmodelingholes
withthediscretizationschemedescribedinSection2.2(wheretheelementsides
arenotperpendicular)
Forbothstiffenedandunstiffenedelements,atleasttwoS9R5elementsshallbe
providedinthedirectionperpendiculartotheapplicationofload
Rounded corners shall be modeled with at least two S9R5 elements, and the
elementaspectratiooftheseelementsshallbelessthanorequalto16:1.
21
Chapter 3
Elastic buckling of cold-formed steel
cross-sectional elements with holes
Asimplifiedmethodfordeterminingtheelasticbucklingpropertiesofathinwalled
crosssection is to evaluate the contribution of each element in the crosssection
separately.Thiselementbyelementevaluationisthebasisoftheeffectivewidthdesign
method for coldformed steel beams and columns and can also be employed as a
conservative predictor of the local critical elastic buckling load (Pcrl) when designing
coldformedsteelmemberswiththeDirectStrengthMethod(AISIS1002007,Appendix
1). The two common crosssection element types in an open thinwalled cross section
are stiffened and unstiffened elements, examples of which are provided in Figure 3.1.
The boundary conditions of a stiffened element are conservatively approximated as a
simplysupportedplate.Theunstiffenedelementistreatedasaplatesimplysupported
onthreesidesandfreeonthefourthedgeparalleltotheapplicationofload.
22
Flange, stiffened
element
Lip,
unstiffened
element
Web,
stiffened
element
Figure3.1StiffenedandunstiffenedelementsinalippedCsection
Theinfluenceofholesontheelasticbucklingbehaviorofstiffenedandunstiffened
elements is evaluated in this chapter using thin shell finite element eigenbuckling
analysis.Thepresenceofholescanmodifythebuckledmodeshapeofanelementand
eitherincreaseordecreaseitscriticalelasticbucklingstress.Holespacingandholesize
relative to element size are studied for both stiffened and unstiffened elements, and
approximate methods for predicting element critical elastic buckling stress are
developedandpresentedforuseindesign.Theresearchresultspresentedherewillbe
usedasaframeworkfortheelasticbucklingstudiesoffullcoldformedsteelstructural
memberswithholesinChapter4.
3.1 Plateandholedimensions
Plateandholedimensionnomenclatureusedthroughoutthischapterissummarized
inFigure3.2.Thestripsofplatebetweenaholeandtheplateedgeswillbereferredtoas
unstiffened strip A and unstiffened strip B, where the widths of these unstiffened
strips are hA and hB respectively as shown in Figure 3.3. For stiffened elements in
bending,theneutralaxislocationisdefinedasYinFigure3.4andismeasuredfromthe
compressededgeoftheplate.
23

S/2
Detail A
Plate with holes

C
L Hole
hhole
Lhole
+hole
C
L Plate
rhole
Slotted hole
Detail A
Figure3.2Elementandholedimensiondefinitions
Unstiffened strip A
hA
hB
Unstiffened strip B
Detail A
Figure3.3DefinitionofunstiffenedstripAandBforaplatewithholes.
Compressed edge
Neutral axis
Y
Tension edge
Detail A
Figure3.4Definitionofneutralaxislocationforstiffenedelementsinbending.
24
3.2 Finiteelementmodelingassumptions
The elastic buckling behavior of stiffened and unstiffened elements with holes are
obtained with eigenbuckling analyses of plates in ABAQUS (ABAQUS 2007a). All

members are modeled with ABAQUS S9R5 reduced integration ninenode thin shell
elements.Thetypicalfiniteelementaspectratiois1:1andthemaximumaspectratiois
limitedto8:1(refertoChapter2foradiscussiononABAQUSthinshellfiniteelement
types and finite element aspect ratio limits). Element meshing was performed with a
Matlab (Mathworks 2007) program written by the author (refer to Appendix A for a
description of the program). The plate models are loaded from each end with stress
distributions applied as consistent nodal loads in ABAQUS. Converting a stress
distribution to consistent nodal loads for the S9R5 element requires a different
procedure than that followed for a 4node finite element (Schafer 1997). Coldformed
steel material properties are assumed as E=29500 ksi and =0.3 in all finite element
models.
3.3 Stiffenedelementinuniaxialcompression
3.3.1 Boundary and loading conditions
Thestiffenedelementismodeledwithsimplysupportedboundaryconditionsand
loadeduniaxiallywithauniformcompressivestressasshowninFigure3.5.
25
perimeter
supported in 2 (v = 0)
longitudinal midline
supported in 1 (u = 0)
2
transverse midline
supported in 3 (w = 0)
Figure3.5ABAQUSboundaryconditionsandloadingconditionsforastiffenedelementinuniaxial
compression
3.3.2 Influence of a single slotted hole
This study explores the influence of a single slotted hole on the elastic buckling
stress of a stiffened element. The plate length L is varied from three to twentyfour
timestheslottedholelength,Lhole,andthewidthoftheplatesarechosentoequaltheflat
web widths of four common Steel Stud Manufacturers Association (SSMA) structural
studs listed in Table 3.1 (SSMA 2001). The slotted hole has dimensions of hhole=1.5 in.,
Lhole=4in.,andrhole=0.75in.Holesarealwayscenteredtransverselybetweentheunloaded
edgesoftheplateinthisstudy.Theplatethickness,t,is0.0346in.
Table3.1PlatewidthscorrespondingtoSSMAstructuralstuddesignations
SSMA
Designation
(in)
250S162-33
2.28
0.66
362S162-33
3.40
0.44
600S162-33
5.78
0.26
800S162-33
7.78
0.19
hhole/h
The results of this study are presented in Figure 3.6 and demonstrate that as the
length of a stiffened element increases relative to the length of the hole, the critical
elasticbucklingstress,fcr,convergestoaconstantmagnitudewhichiseitherequaltoor
lowerthanthebucklingstressofaplatewithoutahole.TheconvergenceoccursasL/Lhole
26
exceeds 5, suggesting that the influence of the hole is independent of the plate end
conditionsbeyondthislength.
1.2
0.8
hhole/h=0.66
hhole/h=0.44
0.6
hhole/h=0.19
hhole/h=0.26
0.4
0.2
10
15
L/Lhole
20
25
Figure3.6Influenceofaslottedholeontheelasticbucklingstressofasimplysupportedrectangularplate
withvaryinglength
Whentheholeiswiderelativetothewidthoftheplate(hhole/h=0.66)andL/Lholeis
small(seeFigure3.6),theelasticbucklingstressoftheplatewiththeholeisasmuchas7
percenthigherthanforaplatewithoutahole.Thisincreaseinstressisexplainedbythe
buckledmodeshapesinFigure3.7.TheplatewiththeholeinFigure3.7ahasahigher
elasticbucklingstressthantheplatewithouttheholeinFigure3.7bbecausethenatural
patternofbuckledwavesismodifiedbythehole.Thebuckledcellsadjacenttothehole
areshorterandthereforestiffer.Thethinstripsattheholedampenbucklinginthiscase
becausetheyhaveanaxialstiffnesshigherthanthebuckledcellsawayfromthehole.
27
(a)
(b)
Figure3.7Comparisonofbuckledshapeanddisplacementcontoursforarectangularplatewithhhole/h=0.66
andL/Lhole=3,(a)withslottedholeand(b)withouthole.Noticethechangeinlengthandquantityofbuckled
cellswiththeadditionofaslottedhole.
As the plate length increases past L/Lhole=5 for the smallest plate width
(hhole/h=0.66),thebucklingstressconvergestothatofaplatewithoutahole.Figure3.8
demonstrates that for these long, slender stiffened elements the slotted hole dampens
bucklingneartheholebutdoesnotappreciablychangethenaturalhalfwavelengthof
thebuckledcellsaswasobservedfortheshorterplatesinFigure3.7.
(a)
(b)
Figure3.8Buckledshapeofasimplysupportedplate(a)withaslottedholeand(b)withoutahole.
L=15Lhole,hhole/h=0.66.Theslottedholedampensbucklingbutdoesnotsignificantlychangethenaturalhalf
wavelengthoftheplate.
For plates with hhole/h less than 0.66, the slotted hole causes a decrease in the
elastic buckling stress which converges to a constant magnitude as the plate length
exceedsL/Lhole=5.Figure3.9ademonstratesthatlocalbucklingneartheholecontrolsthe
28
elasticbucklingstressofthesewiderplates.Thedeformationattheholeresultsfromthe
localizedreductionintransverseplatebendingstiffness.
AsplatelengthdecreasesbelowL/Lhole<5andhhole/h=0.19,theinfluenceofthehole
onthecriticalelasticbucklingstressfluctuatesasshowninFigure3.6.Whenthelowest
elastic buckling mode shape results in an odd number of halfwaves, the hole falls
within the central halfwave and the critical elastic buckling stress decreases. For an
evennumberofhalfwaves,theholeislocatedatthetransitionbetweentwohalfsine
waves (because the hole is centeredat the midlength of theplate), forcing the buckled
cellstoshortenandincreasingthecriticalelasticbucklingstress.
(a)
(b)
Figure3.9(a)Slottedholecauseslocalbuckling(hhole/h=0.26),comparedto(b)buckledcellsatthenatural
halfwavelengthoftheplate
3.3.3 Influence of slotted hole spacing
Thepreviousstudydemonstratedthattheelasticbucklingbehaviorofastiffened
element with a single hole is sensitive to the size of the hole relative to the size of the
plate. The focus now shifts to the influence of multiple slotted holes on the elastic
bucklingstressofalongfixedlengthstiffenedelement.Inthisstudy,slottedholesare
addedonebyonetoastiffenedelement(whereL=24Lhole)suchthatthecentertocenter
spacingSvariesasshowninFigure3.10.
29
Lhole
S/2
hhole
Figure3.10Definitionofcentertocenterdimensionfortheslottedholes
Asholespacingdecreases,theelasticbucklingstressinFigure3.11eitherincreases
ordecreasesdependingontheratioofholewidthtoplatewidth.Whentherearemany
large holes (hhole/h=0.66, S/ Lhole < 4), buckling is dampened at the holes and the buckled
cells shorten their lengths to form between adjacent holes (see Figure 3.12 for buckled
shape).Thedecreaseinbuckledhalfwavelengthcausesanincreaseinelasticbuckling
stressoftheplate.
1.2
fcr,holes /fcr,no holes
0.8
0.6
hhole/h=0.66
0.9
hhole/h=0.44
0.85
0.4
0.75
0.2
hhole/h=0.19
0.8
hhole/h=0.26
2
3
4
S/Lhole
10
15
S/Lhole
20
25
Figure3.11Influenceofslottedholespacingontheelasticbucklingloadofalongsimplysupported
rectangularplate
30
Whentheholesaresmallerrelativetotheplatewidth(hhole/h<0.44)andarespaced
closely together (S/ Lhole < 4), the local buckling influence of adjacent holes combine to
sharply decrease the elastic buckling stress. The inset of Figure 3.11 highlights this
reduction in elastic buckling stress for hhole/h=0.19 and hhole/h=0.26, and Figure 3.12
provides a summary of the associated buckled shapes. When hole spacing increases
beyondS/Lhole=5,theelasticbucklingstressesapproachconstantmagnitudesforallplate
widths considered, which is consistent with the trends presented in Figure 3.6. This
observationisimportantfromadesignperspectivebecauseitservesasarationalbasis
forsettingholespacinglimitsincoldformedsteelmembers.
Buckling is dampened at the holes, halfwaves form between holes
hhole/h=0.66, S/Lhole=4
Buckling of the unstiffened strips adjacent to
the hole is dominant here
Buckled half-waves form along the length of
the plate
Figure3.12Comparisonofbuckledshapesforalongstiffenedelement(L=24Lhole)withaslottedhole
spacingofS/Lhole=4andhhole/h=0.66,0.44,and0.26.
Figure3.12highlightsthetwotypesofbucklingmodesthatcanoccurinstiffened
elements,platebucklingandunstiffenedstripbuckling.Theinfluenceofthesebuckling
31
modes on fcr is reflected in Figure 3.13. The maximum decrease in fcr occurs for a
relatively small hole when compared to the plate width (hhole/h=0.30) and lies at the
transition between plate buckling, where axial stiffness of the buckled cells is reduced
with the presence of holes, and unstiffened strip buckling. Unstiffened strip buckling
occursbetweenhhole/h=0.30andhhole/h=0.55resultinginarelativeincreaseinfcrasthestrips
adjacent to the holes increase the axial stiffness of the plate. As hhole/h increases past
hhole/h=0.55theunstiffenedstripadjacenttotheholebecomesnarrowandstiff,resulting
inplatebucklingawayfromtheholesandanfcrsimilartoaplatewithoutahole.(An
increaseincriticalelasticbucklingstressforlargeholesdoesnotnecessarilycorrespond
toanincreaseinultimatestrengthbecausethestrengthoftheplatewillbelimitedbythe
strength of the net crosssection.) This is another important observation that will be
usedwhendevelopinganelasticbucklingpredictionmethodforstiffenedelementswith
holesinSection3.3.4.
32
fcr,holes /fcr,no holes
1.5
Plate buckling
0.5
0.2
Unstiffened strip
buckling
0.4
Plate buckling
away from the hole
0.6
0.8
hhole/h
Figure3.13Variationinfcrwithincreasinghhole/hforastiffenedelementcorrespondtobucklingmodeshapes
(seeFigure3.12forexamplesofplatebucklingandunstiffenedstripbucklingmodeshapes)
3.3.4 Approximate prediction method for use in design
Approximations for the critical elastic buckling stress of stiffened elements (e.g.
columnweborflangeofalippedCsection)withholesunderuniaxialcompressionare
developed in this section considering two elastic buckling states, buckling of the plate
withoutholeinfluenceandbucklingoftheunstiffenedstripsadjacenttothehole.The
proposed prediction method is validated with thin shell finite element eigenbuckling
analyses for a variety of hole shapes, sizes, and spacings. Mandatory dimensional
tolerances on the prediction method are explicitly defined, and optional dimensional
limits,markedwithanasterisk(*),areprovidedtoavoidexcessiveconservatism.
33
3.3.4.1
Definitionsandassumptions
Figure 3.2 defines the plate and hole dimension notation used in the element
predictionmethod,includingtheholespacingS,platewidthh,andholelengthandhole
width,Lholeandhhole. holeistheoffsetdistanceofaholemeasuredfromthecenterlineof
the plate. The elastic buckling prediction method for a stiffened element is developed
assuming a long plate loaded uniaxially and simplysupported on all four sides with
evenly spaced holes. A summary of all prediction method equations is provided in
AppendixD.
3.3.4.2
PredictionEquations
Theelasticbucklingstressofastiffenedelementwithholesisapproximatedas
f crl = min[ f cr , f crh ].
(3.1)
Thecriticalelasticbucklingstressforplatebuckling(withoutholeinfluence)is
2
2E t
f cr = k
,
12(1 2 ) h
(3.2)
wherekiscommonlytakenequalto4whenconsideringlongrectangularplates(L/h>4).
When elastic buckling of the stiffened element is governed by the buckling of an
unstiffenedstripadjacenttothehole,thecriticalelasticbucklingstressofthegoverning
unstiffenedstripis:
f crh ,net = min [ f crA , f crB ]
34
(3.3)
f cri
2E t
= ki
12(1 2 ) hi
andi=AorB
(3.4)
TheplatebucklingcoefficientkiforunstiffenedstripsAandBareapproximatedby(Yu
andSchafer2007):
for Lhole hi 1 ,
for Lhole hi < 1 ,
k i = 0.425 +
(Lhole
0.2
,
0.95
hi ) 0.6
(3.5)
ki = 0.925 ,andi=AorB.
(3.6)
Eq.(3.5)accountsforthelengthoftheunstiffenedstrip.Asholelengthshortensrelative
totheunstiffenedstripwidth,kiincreases.ThisisanimprovementoverAISIS100which
conservativelyassumesthelowerboundk=0.425regardlessofholelength.WhenLhole/hiis
less than 1, k may be conservatively assumed equal to 0.925 via Eq. (3.6) or calculated
directly by solving the classical stability equations for an unstiffened element
(Timoshenko1961).
Anet = (h hhole ) t
Ag = ht
P2
P1 + P2 = Pcr = f crh ,net Anet
hhole
Pcr = f crh Ag
P1
fcrh,net
fcrh,net is the critical elastic buckling

stress of the wider unstiffened strip
f crh = f crh ,net
Anet
h
= f crh ,net 1 hole
Ag
h
Figure3.14Unstiffenedstripelasticbucklingstressconversionfromthenettothegrosssection
35
Tocomparethebucklingstressfromtheunstiffenedstrip(fcrh,net)tothatoftheentireplate
(fcr) equilibrium between the net and gross section must be considered, as shown in
Figure3.14andprovidedinthefollowing:
3.3.4.3
3.3.4.3.1
f crh = f crh ,net (1 hhole h ) .
(3.7)
Verificationandequationlimits
Holescenteredtransverselyinplate
ThinshellfiniteelementeigenbucklinganalysisinABAQUS,asdescribedinSection
3.2, is employed here to verify the accuracy of the approximate prediction method in
Section3.3.4.2.Theboundaryandloadingconditionsassumedforthestiffenedelement
aredescribedinFigure3.5.Thelengthoftheslottedhole,Lhole,widthoftheplateh,the
shape of hole (slotted, circular, square), the hole spacing S, length of the plate L, and
platethicknesstarevariedintheanalyses.Theplateandholedimensionsaswellasthe
ABAQUScriticalelasticbucklingstress,fcrl,forthe145modelsconsidered,areprovided
in Appendix B(the eigenbuckling results from the studies in Section 3.3.2 and Section
3.3.3 are included in the 145 models). The parametric ranges in this study are
summarizedforeachholetypeinTable3.2.
Table3.2Parameterrangesinstiffenedelementverificationstudy.
Hole type
Slotted
Circular
Square
Min
Max
Min
Max
Min
Max
hhole/h
0.10
0.70
0.10
0.70
0.10
0.70
S/Lhole
1.7
24.0
13.3
13.3
13.3
13.3
36
S/h
1.2
42.2
1.3
9.3
1.3
9.3
h/t
21
434
62
434
62
434
# of models
131
7
7
The results of the ABAQUS eigenbuckling analyses are compared to the stiffened
elementpredictionmethodinFigure3.15andFigure3.16.Figure3.15demonstratesthat
asholespacingSbecomessmallrelativetotheplatewidthh,thepredictionmethodis
not alwaysaccurate. As hole spacing decreases, holes begin to coincide with thelocal
bucklinghalfwavelengths(whichhavealengthofh)andtheinfluenceoftheindividual
holes act cumulatively to decrease the axial stiffness of the plate. A similar loss in
stiffness is observed in Figure 3.16 as hole spacing decreases relative to hole length.
Fromtheseobservations,thefollowinglimitsareimposedonthepredictionmethod:
S
1.5 ,
h
(3.8)
S
2 .
Lhole
(3.9)
IftheparameterlimitinEq.(3.9)issubstitutedintoEq.(3.8),athirddimensionallimitis
automaticallyimposed:
Lhole
0.75
h
(3.10)
Eq.(3.10)preventstheholelengthfrombeingtoolongrelativetothehalfwavelengthof
theplate.ThemeanandstandarddeviationoftheABAQUStopredictedratioforthe
stiffened elements within the limits of Eq. (3.8) and Eq. (3.9) are 1.02 and 0.04
respectively,demonstratingthatthepredictionmethodisviable.
37
1.5
Plate buckling controls
Unstiffened strip controls
fcrl, ABAQUS/fcrl, predicted
1.5
0.5
10
20
30
40
0.5
50
S/h
10
20
30
40
50
S/h
Figure3.15AccuracyofstiffenedelementpredictionmethodasafunctionofholespacingStoplatewidthh
(a)withoutand(b)withthedimensionallimitsinEq.(3.8)andEq.(3.9)
1.5
1.5

0.5
10
15
20
0.5
25
S/Lhole
10
15
20
25
S/Lhole
Figure3.16AccuracyofstiffenedelementpredictionmethodasafunctionofholespacingStolengthof
holeLhole(a)withoutand(b)withthedimensionallimitsinEq.(3.8)andEq.(3.9)
1.5
1.5

0.5
0.2
0.4
0.6
0.8
0.5
hhole/h
0.2
0.4
0.6
0.8
hhole/h
Figure3.17Accuracyofthestiffenedelementpredictionmethodasafunctionofholewidthhholetoplate
widthh(a)withoutand(b)withthedimensionallimitsinEq.(3.8)andEq.(3.9)
38
AsholewidthincreasesrelativetoplatewidthinFigure3.17,thecontrollingbuckled
state transitions from buckling of the unstiffened strip adjacent to plate buckling. The
stripsofwebmaterialadjacenttotheholeshaveahigheraxialstiffnessthanthesections
oftheplatewithoutholes,causingplatebucklingtooccurbetweentheholesasshown
inFigure3.18.
Strips of plate adjacent to the

hole are stiffer than plate
between holes
Figure3.18Forplateswheretheunstiffenedstripisnarrowcomparedtotheplatewidth,platebuckling
occursbetweentheholes.
As the hole width becomes small relative to plate width, the unstiffened strip
buckledstateispredictedbythesimplifiedmethodforslottedholes,althoughtheactual
behaviorisacombinationofplatebucklingandlocalbucklingattheholes,asshownin
Figure3.19.Theassumptionofunstiffenedstripbucklingwhentheslottedholewidthis
small relative to plate width is conservative, with a maximum ABAQUS to predicted
ratioof1.16whenhhole/hisintherangeof0.30.Figure3.19alsodemonstratesthatplate
buckling dominates over unstiffened strip buckling for stiffened elements with square
andcircularholes.Thepredictionmethodidentifiesthiselasticbucklingbehaviorand
accurately predicts fcrl as shown in Figure 3.20, where stiffened element results
containingjustsquareorjustcircularholesareplotted.
39

Plate buckling dominates over
unstiffened strip buckling for
square (and circular holes)
Plate buckling and

unstiffened strip buckling
are both present when
0.20hhole/h0.60
Figure3.19Platebucklingandunstiffenedstripbucklingmaybothexistforaplatewithholes.These
modesarepredictedconservativelyasunstiffenedstripbuckling.
1.5
fcr , ABAQUS/fcr , predicted

l
l
0.5
0.2
0.4
0.6
0.8
hhole/h
Figure3.20Accuracyofpredictionmethodforstiffenedelementswithsquareorcircularholesasafunction
ofholewidthhholetoplatewidthh.
3.3.4.3.2
Offsetholes
43 additional ABAQUS eigenbuckling analyses were performed to evaluate the
accuracy of the simplified prediction method in Section 3.3.4.2, but now with
transversely offset holes. For these models, the hole offset from the centerline of the
40
plate, hole, and the plate width, h, were varied. All plate models in this study have
regularly spaced slotted holes (S=20 in.) and constant plate length, L, of 100 in. The
boundary and loading conditions assumed for the stiffened element are described in
Figure 3.5. The model dimensions and critical elastic buckling stress, fcrl, for the 43
modelsconsidered,aresummarizedinAppendixB.hstripisthewidestunstiffenedstrip,
eitherhAandhB.TheparametricrangesforthisstudyaresummarizedinTable3.3.
Table3.3Parameterrangeforstiffenedelementverificationstudywithoffsetholes.
hhole/h
S/Lhole
S/h
h/t
hole/h
Min
0.10
5.0
1.3
62
0.000
Max
0.70
5.0
9.3
434
0.375
Hole type
Slotted
2.5
2.5
43
3
1.5
0.5
# of models
1.5
0.5
0.2
0.4
0.6
0.8
0.2
hstrip/h
0.4
0.6
0.8
hstrip/h
Figure3.21Accuracyofthestiffenedelementelasticbucklingpredictionmethodasafunctionof
unstiffenedstripwidthhstripversusplatewidthhforoffsetholes(a)withoutand(b)withthedimensional
limitsinEq.(3.8)andEq.(3.9)
The ABAQUS critical elastic buckling stress results are compared to the prediction
methodinFigure3.21,anddemonstratethatthepredictionmethodisconservativeand
thattheaccuracyofthemethodimprovesashstripdecreasesrelativetotheplatewidthh
andholelengthLhole.Theunstiffenedstripbuckledstateispredictedtocontrolformost
of the plate models, primarily because the shift in hole location results in a wider
41
unstiffened strip with less axial stiffness than that provided by the plate material
betweenholes.Whentheplateisrelativelywidecomparedtothewidthoftheholeand
theholeisshiftedneartheedgeoftheplateasshowninFigure3.22,thepredictionscan
be very conservative. The wide unstiffened strip is not a good approximation of the
actualbehavioroftheplateinthiscase.Predictionaccuracyvarieswithholeoffset, hole,
as shown in Figure 3.23a, and is most conservative as the hole offset becomes large
relativetotheplatewidthh.Toavoidoverlyconservativeresults,thefollowinglimiton
holeoffsetholeisproposedforstiffenedelements:
hole
h
0.15 *
(3.11)
ThemeanandstandarddeviationoftheABAQUStopredictedratioforthedatawithin
thedimensionallimitsofEq.(3.8),Eq.(3.9),andEq.(3.11)are1.14and0.15respectively
(alsoseeFigure3.23b).
The prediction method conservatively assumes

unstiffened strip buckling of the wider strip adjacent to the
hole, although plate buckling is observed.
hstrip
Figure3.22Holesattheedgeofawidestiffenedplatereducetheaxialstiffness(andcriticalelasticbuckling
stress)butdonotchangethebuckledshape.
42

1.5
1.5
0.5
0.5
2.5
fcr , ABAQUS/fcr , predicted
l
l
2.5
0.1
0.2
0.3
0.4
hole/h
0.5
0.1
0.2
0.3
0.4
hole/h
0.5
Figure3.23Accuracyofthestiffenedelementelasticbucklingpredictionmethodasafunctionofholeoffset
holeversusplatewidthhforoffsetholes(a)withoutand(b)withthedimensionallimitsinEq.(3.8),Eq.(3.9),
andEq.(3.11)
3.4 Stiffenedelementinbending
Thestiffenedelementismodeledwithsimplysupportedboundaryconditionsand
loaded with a bending compressive stress distribution as shown in Figure 3.24. The
location of the neutral axis about which bending occurs, Y, is measured from the
compressededgeoftheplate.
Restrain point in 3
(w=0)
Neutral axis
Restrain transverse
midline in 1 (u=0)
Restrain plate
perimeter in 2 (v=0)
Restrain point in 3
(w=0)
2
1
Figure3.24Boundaryandloadingconditionsforastiffenedelementinbending
43
3.4.2 Influence of transversely-centered slotted holes
Shell finite element eigenbuckling models of stiffened elements with regularly
spaced slotted holes are evaluated in this study. The bending stress distribution is
symmetric about the transverse centerline of the plate (Y=0.50h) for all models. The
slotted holes are centered transversely in the plate (hole=0). The plate and hole
dimensionsandthecriticalelasticbucklingstress,fcrl,forthe28modelsconsidered,are
summarizedinAppendixB.Theparametricrangesforthisstudyaresummarizedin
Table3.4.
Table3.4Parameterrangesconsideredforstiffenedelementsinbendingwithholes.
Hole type
Slotted
hhole/h
S/Lhole
S/h
h/t
Y/h
Min
0.10
1.67
1.33
61.93
0.50
Max
0.70
5.00
9.33
433.53
0.50
# of models
28
Figure 3.25 highlights the influence of hole width to plate width on stiffened
elements in bending. As hhole/h increases, the buckling mode transitions from plate
buckling (similar to a plate without a hole) to buckling of the compressed unstiffened
stripadjacenttothehole.Thebuckledhalfwavelengthofaplateinbendingisbetween
0.25h to 0.50h, which results in a shortened halfwavelength of the unstiffened strip
(often less than the length of the hole) when compared to the equivalent unstiffened
stripbucklingmodeforstiffenedelementsinuniaxialcompression(SeeSection3.3).
44
Unstiffened strip half-wavelength can

be less than the length of the hole.
hhole/h=0.10
hhole/h=0.30
hhole/h=0.50
Unstiffened strip buckling becomes more

predominant as the hole size increases
relative to plate width.
Figure3.25Stiffenedplatesloadedwithalinearbendingstressgradientexhibitbucklingoftheunstiffened
stripadjacenttotheholeinthecompressionregionoftheplate.
The maximum reduction in critical elastic buckling stress occurs in the range of
hhole/h=0.30 as shown in Figure 3.26a. This result is consistent with the elastic buckling
results for stiffened plates under axial compression (See Figure 3.13). The elastic
buckling behavior of stiffened elements in bending are different than in pure
compression though as hhole/h exceeds 0.50. Unstiffened strip buckling continues to
dominate for plate bending (with an associated reduction in fcr) while plate buckling
away from the hole controls for uniaxially compressed plates (with minimal influence
on fcr even for very large holes). This distinction between compression (columns) and
bending (beams) elastic buckling behavior of stiffened elements is important when
consideringhowtoapproximateelasticbucklingbehavior.fcrdecreasesasholespacing
becomes small relative to hole length as shown in Figure 3.26b, identifying S/Lhole as
another important parameter when predicting elastic buckling of stiffened elements in
bending(asitisforstiffenedelementsincompression,SeeSection3.3).
45
1.5
1.5
0.5
0.2
0.4
0.6
0.8
0.5
hhole/h
10
S/Lhole
Figure3.26Influenceofslottedholesoncriticalelasticbucklingstressfcrofstiffenedelementsinbendingas
afunctionof(a)holesizerelativetoplatewidthand(b)holespacingasafunctionofholelength.
3.4.3 Influence of offset slotted holes

3.4.3.1
NeutralaxislocationatY=0.50h
Shell finite element eigenbuckling models of stiffened elements with regularly
spacedoffsetslottedholesareevaluatedinthisstudy.Thebendingstressdistributionis
symmetricaboutthetransversecenterlineoftheplate(Y=0.50h)forallmodels.Thehole
offset,hole,rangesfrom0.375hto+0.375h,whereapositiveshiftmovestheholesintothe
compression region of the plate. The plate and hole dimensions and the critical elastic
bucklingstress,fcr,forthe92modelsconsidered,aresummarizedinAppendixB.The
parameterrangeconsideredinthisstudyisprovidedinTable3.5.
Table3.5Studyparameterlimitsforstiffenedelementinbending(Y/h=0.50)withoffsetholes
hhole/h
S/Lhole
S/h
h/t
Y/h
hole/h
Min
0.10
5.00
1.33
61.93
0.50
-0.375
Max
0.70
5.00
9.33
433.53
0.50
0.375
Hole type
Slotted
# of models
92
The presence of holes in the compression region of a stiffened element in bending
(Y=h/2) decreases the critical elastic buckling stress when compared to a plate without
holesasshowninFigure3.27.DependinguponthewidthoftheunstiffenedstripA
46
inthecompressedregionoftheplateandtheunstiffenedstripBinthetensileregion
of the plate (see Figure 3.3 for definitions) relative to hole depth h, unstiffened strip
buckling may occur above the hole, below the hole, or above and below the hole. fcr
varieswiththetransversepositionoftheholesintheplate(characterizedasthewidthof
unstiffened strip A, hA) in Figure 3.27. The trends in fcr can be related to the elastic
buckling modes in Figure 3.28. If the holes are located in the tensile region of the
stiffenedelement,thebuckledmodeshape(andfcr)areunchangedwhencomparedtoa
stiffenedelementwithoutholes.Therelationshipbetweenthesebuckledmodeshapes
and trends in fcr will be used in Section 3.4.4 when developing an approximate elastic
bucklingpredictionmethodforstiffenedelementsinbending.
0
0.1
hhole/h=0.10
hhole/h=0.20
B
0.2
0.3
hA /h
hhole/h=0.40
Compression
0.4
hhole/h=0.30
hhole/h=0.50
Neutral axis
hhole/h=0.60
hhole/h=0.70
0.5
0.6
D
0.7
Tension
0.8
0.9
1
0.5
1
1.5
Figure3.27Holelocationinfluenceoncriticalelasticbucklingstressfcrforastiffenedplateinbending
(Y=0.50h)(BuckledmodeshapescorrespondingtoA,B,C,andDareprovidedinFigure3.28.)
47
hA
Unstiffened strip
buckling (below hole)
B
Unstiffened strip
buckling (below and
above hole)
D
Unstiffened strip
buckling (above hole)
Plate buckling (no hole

influence)
Figure3.28Thebuckledmodeshapechangesasslottedholesmovefromthecompressionregiontothe
tensionregionofastiffenedelementinbending(hhole/h=0.20).
3.4.3.2
NeutralaxislocationatY=0.75h
TheneutralaxisintheshellfiniteelementeigenbucklingmodelsfromSection3.4.3.2
isnowmodifiedtoY=0.75h.ThetrendsinfcrinFigure3.29aresimilartothoseobserved
inFigure3.27(Y=0.50h).Elasticbucklingoftheunstiffenedstripbelowtheholesoccurs
when the hole is close to the compressed edge. The mode shape transitions to
unstiffenedstripbucklingabovetheholesastheholeoffsetincreasestowardthetensile
region of the plate. The plate and hole dimensions and the critical elastic buckling
stress,fcr,forthe92modelsconsideredhere,aresummarizedinAppendixB.
48
0
hhole/h=0.10
0.1
hhole/h=0.20
0.2
hhole/h=0.30
0.3
hhole/h=0.40
hhole/h=0.50
0.4
hhole/h=0.60
hA /h
Compression
hhole/h=0.70
0.5
0.6
Neutral axis
0.7
0.8
Tension
0.9
1
0.5
1
1.5
Figure3.29Holelocationinfluenceoncriticalelasticbucklingstressfcrforastiffenedplateinbending
(Y=0.75h)
Intheprevioussectionuniqueelasticbucklingmodeswereidentifiedforastiffened
elementinbendingwithholes.Bucklingoftheunstiffenedstripbetweentheholeand
the compressed edge of the plate (unstiffened strip A) or between the hole and the
tension edge of the plate (unstiffened strip B) may occur depending upon the
transverse location of the hole in the plate, the width of the hole (hhole) relative to the
depth of the plate (h), and the location of the plate neutral axis (Y). If the hole is
completelycontainedwithinthetensionregionoftheplatethentheholehasaminimal
influence on elastic buckling and the critical elastic buckling stress, fcr, remains
unchanged.Theseobservationscanbeusedtodefineanapproximationforthecritical
elasticbucklingstressofastiffenedelementwithholesinbending:
49
f crl = min[ f cr , f crh ] .
(3.12)
The critical elastic buckling stress for a stiffened element in bending (without the
influenceofholes),fcr,maybedeterminedwithEq.(3.2)wherethebucklingcoefficientk
iscalculatedwithAISIS10007Eq.B2.32(AISIS1002007):
k = 4 + 2(1 + ) + 2(1 + )
3
(3.13)
and istheabsolutevalueoftheratiooftensilestresstocompressivestressappliedto
thestiffenedelement,i.e.:
= f 2 f1 = (h Y ) Y .
(3.14)
When elastic buckling of the stiffened element is governed by the buckling of an

unstiffenedstripadjacenttoahole,thecriticalelasticbucklingstressis:
f crh ,net = min [ f crA , f crB ]
(3.15)
ConsiderationofunstiffenedstripAisrequiredonlyifhA<Y,i.e.,atleastaportionof
theholemustlieinthecompressionregionofthestiffenedelement.Ifthatconditionis
mettheelasticbucklingstressforstripAis:
2
f crA
2E t

= kA
12(1 2 ) h A
(3.16)
TheplatebucklingcoefficientfortheunstiffenedstripAisapproximatedas
kA =
2.70 1.76 A
Y hA
0.578
+
,and A =
2
A + 0.34 0.024 A + 0.035 + (Lhole h A )
Y
(3.17)
Eq. (3.17) is a modification of AISIS10007 Eq. B3.32 (AISIS100 2007)This expression

accounts for the gradient of the compressive stress distribution and the aspect ratio of
50
theunstiffenedstrip(seeAppendixCforderivation).Theequationfor Aisderivedin
Figure3.30.
hA
f1
Neutral Axis
f2
Similar
Triangles
A =
f 2 Y hA
=
f1
Y
Figure3.30DerivationofstressratioforunstiffenedstripA.
Consideration of unstiffened strip B is required only if hA+hhole<Y, i.e., only when
theentireholelieswithinthecompressedregionoftheplate.Forthiscasethebuckling
stressoftheunstiffenedstrip,convertedtoastressatthecompressededge,isfoundas:
f crB
2E t
= kB
12(1 2 ) hB
Y h A hhole
(3.18)
where the final term in Eq. (3.18) converts the buckling stress from the edge of
unstiffenedstripBtotheedgeofunstiffenedstripAasshowninFigure3.31sothat
thetwostresses(fcrAandfcrB)maybecomparedinEq.(3.15)todeterminetheminimum.
TheplatebucklingcoefficientfortheunstiffenedstripBisapproximatedas:
forLhole/hB>2
forLhole/hB2
k B = 0.340 B2 + 0.100 B + 0.573 ,
51
(3.19)
h
0.38 B + 1.6 B + 0.49
Lhole
,
kB =
0.1
h
0.3
0.20 B + B + 0.14
Lhole
1.8
(3.20)
andtheratiooftensiontocompressivestresses(derivedinFigure3.31)is:
B =
h Y
, 0 B 10 .
Y h A h hole
(3.21)
TheplatebucklingcoefficientkBisapplicableoveralargerrangeofBthanAISIS10007
Eq. B3.25 (AISIS100 2007) and accounts for the increase in kB as the unstiffened strip
aspectratiotendstozero(i.e.,awide,shortstripresultingfromasmallhole).Referto
AppendixCforthederivationofkB.
hA
fcrB
Neutral Axis
f1
Similar
Triangles
B =
f2
h Y
=
f1 Y hA hhole
Similar
Triangles
f crB
Y
=
f1 Y hA hhole
Solve for fcrB
f crhB =
Y
f1
Y hA hhole
f2
Figure3.31DerivationBandconversionofthecompressivestressattheedgeofunstiffenedstripBto
thestressfcrBattheedgeoftheplate
ConversiontothegrosssectionforthecomparisonofstressesinEq.(3.12)requires
that:
forhA+hholeY,
f crh = f crh,net (1 + A )
52
hA
,
Y
(3.22)
forhA+hhole<Y,
h
h
f crh = f crh,net 1 hole 2 A hole
Y
Y
(3.23)
Theconversionfromfcrh,netatthenetsectionoftheplatetofcrhonthegrosscrosssectionis
obtainedwithasimilarmethodtothatdescribedinFigure3.14forstiffenedelementsin
uniaxialcompression;thetotalcompressiveforceatthenetandgrosscrosssectionsare
assumed in equilibrium as shown in Figure 3.32 and Figure 3.33. A summary of all
predictionmethodequationsisprovidedinAppendixD.
P
Neutral
Axis
fcrh,net
hA
fcrh
P
fcrh,net
Free Body Diagram
Force Equilibrium
+ A f crh,net
f
P = crh,net
2
Solve for fcrh
f crh = f crh ,net (1 + A )
hA
Y
1
hAt = f crhYt
2
Figure3.32DerivationoffcrhforthecasewhenhA+hholeY(whentheholeislocatedpartiallyinthe
compressedregionandpartiallyinthetensionregionoftheplate)
53
Neutral
Axis
hA
fcrh,net
f3
hhole
fcrh
P
fcrh,net
Free Body Diagram
Force Equilibrium
Define f3 using similar

triangles
Substitute f3 and solve

for fcrh
Simplify
+ f3
f
1
1
hholet = f crhYt
P = Yf cr ,net t A crh ,net
2
2
2
f3
Y hA hhole
h
=
= A hole
f crh,net
Y
Y
f crh = f crh,net
h
hhole
A f crh,net + A hole f crh,net

Y
Y
h
h
f crh = f crh,net 1 hole 2 A hole
Y
Y
Figure3.33DerivationoffcrhforthecasewhenhA+hhole<Y(holeliescompletelyinthecompressedregionof
theplate).
3.4.5 Verification and parameter limits
The elastic buckling prediction method for stiffened elements in bending is now
evaluated with the ABAQUS eigenbuckling results presented in Section 3.4.2 and
Section 3.4.3. The viability of the method is examined for evenly spaced slotted holes
centered transversely or offset in a plate. Parameter limits on the prediction method,
requiredwhenformalizingthemethodforuseindesign,arealsoidentified.
ABAQUS results are compared to predictions in Figure 3.34a and Figure 3.38a.
Figure3.34ademonstratesthatthesimplifiedmethodunderpredictstheelasticbuckling
stressasaspectratiooftheunstiffenedstripAincreases.Adimensionaltoleranceis
imposedtoavoidunconservativepredictions:
Lhole
10 .
hA
54
(3.24)
Eq.(3.24)alsoservesasapracticallimitontheslendernessofanunstiffenedstrip,and
thereforeisalsoisimposedontheunstiffenedstripB:
Lhole
10 .
hB
(3.25)
The prediction method becomes increasingly conservative as hA/Y approaches unity as

showninFigure3.35a.Whenonlyasmallportionoftheholeexistsinthecompressed
regionoftheplate,theobservedbucklingmodeismoreconsistentwithplatebuckling
thanunstiffenedstripbucklingaspredictedbythesimplifiedmethod(SeeFigure3.28,
pictureD).Adimensionallimitissuggestedtopreventexcessiveconservatisminthis
case:
hA
0.6 *
Y
(3.26)
The hole spacing limits defined in Section 3.3.4.3 for stiffened elements in uniaxial
compressionarealsoconsideredhereforastiffenedelementinbending.Theprediction
accuracydegradeswhenholespacingSapproachestheplatewidthhasshowninFigure
3.36a. Predictions can also be unconservative when S is 2 to 3 times the length Lhole as
showninFigure3.37a.WiththelimitsfromEq.(3.8),Eq.(3.9),Eq.(3.24),Eq.(3.25),and
Eq.(3.26)imposed,themethodisobservedtobeviablepredictoroverawiderangeof
hhole/hasshowninFigure3.38b.ThemeanandstandarddeviationoftheABAQUSto
predictedratiowithintheimposeddimensionallimitsare1.22and0.11respectively.
55
3
plate buckling
unstiffened strip "A"
unstiffened strip "B"
1.5
0.5
plate buckling
2.5
fcr,ABAQUS/fcr,predicted
2.5
1.5
0.5
10
15
20
Lhole/hA
25
30
35
40
10
15
20
Lhole/hA
25
30
35
40
Figure3.34InfluenceofLhole/yAontheaccuracyofthepredictionmethodforstiffenedelementsinbending
(a)withoutand(b)withthedimensionallimitsdefinedinEq.(3.9),Eq.(3.24),Eq.(3.25),andEq.(3.26).
3
plate buckling
1.5
0.5
plate buckling
2.5
2.5
1.5
0.5
0.1
0.2
0.3
0.4
0.5
hA /Y
0.6
0.7
0.8
0.9
0.1
0.2
0.3
0.4
0.5
hA /Y
0.6
0.7
0.8
0.9
Figure3.35InfluenceofhA/Yontheaccuracyofthepredictionmethodforstiffenedelementsinbending(a)
withoutand(b)withthedimensionallimitsdefinedinEq.(3.9),Eq.(3.24),Eq.(3.25),andEq.(3.26).
56
3
plate buckling
2.5
2.5
1.5
0.5
plate buckling
1.5
0.5
5
S/h
10
5
S/h
10
Figure3.36InfluenceofS/hontheaccuracyofthepredictionmethodforstiffenedelementsinbending(a)
3
plate buckling
2.5
2.5
1.5
0.5
plate buckling
1.5
0.5
5
S/Lhole
10
5
S/Lhole
10
Figure3.37InfluenceofS/Lholeontheaccuracyofthepredictionmethodforstiffenedelementsinbending(a)
57
3
plate buckling
1.5
0.5
plate buckling
2.5
2.5
1.5
0.5
0.1
0.2
0.3
0.4
0.5
hhole/h
0.6
0.7
0.8
0.9
0.1
0.2
0.3
0.4
0.5
hhole/h
0.6
0.7
0.8
0.9
Figure3.38Influenceofh/hholeontheaccuracyofthepredictionmethodforstiffenedelementsinbending(a)
3.5 Unstiffenedelementinuniaxialcompression
The unstiffened element is modeled with simplysupported boundary conditions

on three sides and unsupported on the fourth side parallel to the application of a
uniformcompressivestressasshowninFigure3.39.
Restrain plate
perimeter in 2 (v=0)
Restrain transverse
midline in 1 (u=0)
Restrain longitudinal
midline in 3 (w=0)
1
Figure3.39ABAQUSboundaryandloadingconditionsforunstiffenedplateloadeduniaxially.
58
3.5.2 Influence of regularly-spaced holes
Eigenbuckling analyses in ABAQUS are performed to evaluate the influence of
evenlyspaced holes on the elastic buckling behavior of an unstiffened element. The

modelloadingandboundaryconditionsaresummarizedinFigure3.39andthematerial
propertiesandmeshingproceduresarethesameasthosedescribedinSection3.2.The
platewidthh,holelengthLhole,andholetype(slotted,circular,rectangular)arevariedin
this study. The hole width remains constant at hhole=1.5 in. The plate and hole
dimensionsaswellasthecriticalelasticbucklingstress,fcrl,forthe91modelsconsidered,
areprovidedinAppendixB.Theparametricrangesconsideredinthisstudyforeach
holetypearesummarizedinTable3.6.
Table3.6Parameterrangeforstudyofregularlyspacedholesonunstiffenedelements.
Hole Type
Slotted
Circular
Square
hhole/h
S/Lhole
S/h
h/t
Min
0.10
1.7
1.0
21
Max
0.70
24.0
42.2
434
Min
0.10
13.3
1.3
62
Max
0.70
13.3
9.3
434
Min
0.10
13.3
1.3
62
Max
0.70
13.3
9.3
434
# of models
77
7
7
A comparison of the ABAQUS results from the 91 models to the theoretical elastic
bucklingstressforalongunstiffenedelement(k=0.425)inFigure3.40demonstratesthat
thecriticalelasticbucklingstressfcrdecreasesasholewidthhholeincreasesrelativetoplate
widthh.Holesalwaysreducethecriticalelasticbucklingstressofunstiffenedelements
in the cases studied. Buckling of the unstiffened strip A between the hole and the
simplysupportededgeisnotobservedinthesimulationsbecauseL/hisalwaysgreater
thanLhole/hA,althoughbucklingoftheunsupportedstripBatthefreeedgeoccursasthe
strip becomes slender (similar to Euler buckling) as shown in Figure 3.41. These
59
important observations are employed in Section 3.5.4 to develop an approximate

predictionmethodforthecriticalelasticbucklingstressofanunstiffenedelementwith
holes.
1.4
1.2
fcr, hole/fcr, no hole
1
0.8
0.6
0.4
0.2
0
0.2
0.4
0.6
0.8
hhole/h
Figure3.40Thepresenceofholescausesadecreaseincriticalelasticbucklingloadforunstiffenedplatesin
uniaxialcompression.
Holes do not influence the buckled shape of unstiffened plates until

the hole width becomes large relative to plate width.
hhole/h=0.10
hhole/h=0.60
hA
Buckling of the strip at the free edge of the plate changes the shape
of the local buckling mode.
Figure3.41Buckledshapesofunstiffenedplateswithholes.
60
3.5.3 Influence of offset slotted holes
ABAQUS eigenbuckling analyses were performed to evaluate the influence of
transverselyoffsetslottedholesontheelasticbucklingofanunstiffenedelement.The
ratiooftransverseoffset, hole,toplatewidthhwasvariedfrom0.375to0.375,wherea
negativeoffsetshiftstheholestowardthesimplysupportededgeandapositiveoffset
shiftstowardsthefreeplateedge(refertoFigure3.2foradefinitionof hole).Themodel
loading and boundary conditions are summarized in Figure 3.39 and the material
propertiesandmeshingproceduresarethesameasthosedescribedinSection3.2.The
plate and hole dimensions as well as the critical elastic buckling stress, fcrl, for the 92
modelsconsidered,aresummarizedinAppendixB.Theparametricrangesconsidered
hereareprovidedinTable3.7.
Table3.7Parameterrangeconsideredforunstiffenedelementstudywithoffsetholes
Slotted
hhole/h
S/Lhole
S/h
h/t
hole/h
Min
0.10
5.00
1.33
62
-0.375
Max
0.70
5.00
9.33
434
0.375
# of models
92
The axial stiffness of an unstiffened element is higher near the simply supported
edge and lower near the free edge. It is hypothesized that holes shifted towards the
simplysupported edge will reduce the critical elastic buckling stress more than hole
materialremovedfromnearthefreeedge.ThishypothesisisconfirmedinFigure3.43
where fcr decreases more when holes are shifted towards the simplysupported edge.
The dimension of the plate strip between the hole and the simplysupported edge, hA
(seeFigure3.3),isidentifiedasausefulparameterwhenpredictingfcr.fcrformsatrend
line when plotted against Lhole relative to yA as demonstrated in Figure 3.43a for offset
61
holes.ThesameplotisproducedusingthedatafromSection3.5.2forcenteredholesin
Figure 3.43b with similar results. This important conclusion, that yA and Lhole are key
parametersinfluencingfcr,isusedinthenextsectiontodevelopanapproximateelastic
bucklingpredictionmethodforunstiffenedelementswithholes.
1.4
1.2
1
0.8
Data groups
correspond to
hhole/h=0.10, 0.20,
0.70
0.6
0.4
0.2
0
-0.4
-0.3
-0.2
-0.1
0.1
0.2
0.3
0.4
hole/h
Figure3.42Thecriticalelasticbucklingstressofastiffenedplatedecreasesasholesareshiftedtowardthe
1.4
1.4
1.2
1.2
1
simplysupportededge(+hole)
0.8
0.6
0.8
0.6
0.4
0.4
0.2
0.2
10
15
20
Lhole/hA
25
30
35
40
10
15
20
Lhole/hA
25
30
35
40
Figure3.43Thecriticalelasticbucklingstressforstiffenedelementswith(a)transverselyoffsetholesand
(b)centeredholes(fromSection3.5.2)decreasesasafunctionofholelengthLholetohA
62
Anapproximateelasticbucklingpredictionmethodforanunstiffenedelementwith
holes is presented here. The method is based on the observations in Section 3.3.2 and
Section3.3.3forlongunstiffenedelementswithevenlyspacedholes.Thewidthofthe
stripbetweentheholeandthesimplysupportededge,hA,andthelengthoftheholeLhole
are utilized as predictors of the critical elastic buckling stress. A summary of the
predictionmethodequationsareprovidedinAppendixD.
3.5.4.1
Derivationofempiricalbucklingcoefficient
An empirical plate buckling coefficient is determined using a linear regression
analysisofthedatainFigure3.43aandFigure3.43bforbothcenteredandoffsetslotted
holes, which was then adjusted to have a slightly conservative bias. The regression
minimizestheerrorbetweentheABAQUSresultsandtheclassicalstabilitysolutionof
an unstiffened element (k=0.425) for the plate models within the following parametric
limits:
Lhole
10
hA
(3.27)
Lhole
10
hB
(3.28)
hhole
0.50
h
(3.29)
Eq. (3.27) is imposed as a practical limit on the slenderness of the strip adjacentto the
hole at the simplysupported plate edge. Eq. (3.28) prevents Euler buckling of
63
unstiffened strip B as shown in Figure 3.41. Eq. (3.29) is imposed because of the
increasedrateofdegradationinfcrobservedinFigure3.40asholesbecomelargerelative
toplatewidth.Theempiricalplatebucklingcoefficientissetas:
L
k = 0.4251 0.062 hole
hA
(3.30)
where the strip of plate between the hole and the simply supported edge, hA, is
calculatedas
hA =
h hhole
hole .
2
(3.31)
Apositivehole(holeoffsetfromthecenterlineoftheplate,SeeFigure3.2)shiftsthehole
towardsthesimplysupportededge.Theempirical bucklingcoefficientinEq.(3.30)is
shown in Figure 3.44a to be a slightly conservative but accurate representation of
ABAQUS predicted buckling coefficients. The mean and standard deviation of the
ABAQUStoempiricalpredictionratioare1.06and0.09respectively.
0.5
1.4
ABAQUS
Eq. (3.30)
0.45
1.2
0.4
fcr, ABAQUS/fcr, predicted
buckling coeff. k
0.35
0.3
0.25
0.2
0.15
0.8
0.6
0.4
0.1
0.2
0.05
0
5
Lhole/y A
10
5
Lhole/y A
10
Figure3.44(a)ComparisonofABAQUSandempiricalplatebucklingcoefficientsforanunstiffenedelement
withholesand(b)ABAQUStopredictedelasticbucklingstressforanunstiffenedelement
64
3.5.4.2
Predictionequations
The elastic buckling stress of an unstiffened element in compression with holes is
thusapproximatedas:
f crl = min[ f cr , f crh ] .
(3.32)
The critical elastic buckling stress prediction for plate buckling of the unstiffened
elementwithoutholes(fcr)iscalculatedwithEq.(3.2),wherek=0.425whenconsidering
long rectangular plates (L/h>4). The minimum critical elastic buckling stress of the
unstiffened element with holes, fcrh, coincides with either buckling of the entire
unstiffenedelementwithholesorbucklingoftheunstiffenedstripAadjacenttothe
holeandthesimplysupportededge,or:
f crh
2E t
h
, f crA 1 hole
= min k
2
h
12 1 h
(3.33)
where k is an empirical plate buckling coefficient derived from finite element

eigenbucklingstudiesinEq.(3.30).fcrAiscalculatedwithEq.(3.4)andmodifiedbythe
factor(1hhole/h)toconvertthestressontheunstiffenedstripAtothestressattheend
oftheplatesothatitcanbecomparedtothebucklingstressoftheunstiffenedelement.
fcrhwillalwaysbepredictedaslessthanorequaltofcrwiththismethod.
65
Chapter 4
Elastic buckling of cold-formed steel
members with holes
The elastic buckling properties of coldformed steel lipped Csection beams and
columns with holes are evaluated in this chapter using thinshell finite element
eigenbucklinganalysesinABAQUS.Theelasticbucklingstudiesareusedtoassessthe
influenceofholesonthelocal,distortional,andglobalcriticalelasticbucklingloadsPcrl,
Pcrd, Pcre. The studies also identify elastic buckling modes unique to coldformed steel
memberswithholes.Elasticbucklingpropertiesofexistingexperimentsoncoldformed
steel columns and beams with holes are summarized and formal buckling modes are
definedinpreparationforthepresentationoftheDirectStrengthMethodforstructural
memberswithholesinChapter8.
66
4.1 Finiteelementmodelingassumptions
Theelasticbucklingbehaviorofthecoldformedsteelstructuralmemberswithholes
areobtainedwitheigenbucklinganalysesinABAQUS(ABAQUS2007a).Allmembers
are modeled with ABAQUS S9R5 reduced integration ninenode thin shell elements.
Thetypicalfiniteelementaspectratiois1:1andthemaximumaspectratioislimitedto
8:1(refertoChapter2foradiscussiononABAQUSthinshellfiniteelementtypesand
finite element aspect ratio limits). Element meshing is performed with a Matlab
(Mathworks2007)programwrittenbytheauthor(refertoAppendixAforadescription
oftheprogram).ColdformedsteelmaterialpropertiesareassumedasE=29500ksiand
=0.3 in the finite element models unless noted otherwise. Py, the squash load of the
column,iscalculatedbymultiplyinganassumedyieldstressof50ksibythegrosscross
sectionalareaofthecolumn.
4.2 Elasticbucklingofcolumnswithholes
4.2.1 Member and hole dimensions
Memberandholedimensionnotationusedthroughoutthischapteris summarized
inFigure4.1.Uppercasedimensions(H,D,B)areouttooutandlowercasedimensions
(b,h)areflatlengthsbetweenpointsofcurvature.
67
B1
D1
bhole
hhole
t
r
D2
b
B2
Figure4.1Csectionandholedimensionnotation
4.2.2 Loading and boundary conditions
Thecoldformedsteelcolumnboundaryconditionsaremodeledaswarpingfreeat
thememberendsandwarpingfixedatthemidlengthofthememberasshowninFigure
4.2,whichmimicsthesemianalyticalfinitestripmethod(Schaferanddny2006).The
columns are loaded at each end with stress distributions applied as consistent nodal
loadsinABAQUS(seeSection3.2fordetailsontheloadingimplementation).
midspan nodes are

supported in 1 (u = 0)
end nodes are
supported in 2 and
3 (v = w = 0)
end nodes are supported

in 2 and 3 (v = w = 0)
Figure4.2Columnsaremodeledwithpinnedwarpingfreeboundaryconditionsandcompressedfrom
bothends
68
4.2.3 Elastic buckling comparison of short C-section

columns versus isolated stiffened elements
This study builds on the results and observations in Chapter 3 for crosssectional
elementswithholesandmarksatransitioninresearchfocusfromelementstofullcold
formed steel members. The influence of one slotted hole on the elastic buckling
behavior of a range of rectangular plates and SSMA coldformed steel structural stud
sectionsiscompared,thegoalbeingtoquantifytherelativeinfluenceofawebholeon
one element in a crosssection (in this case a stiffened element, see Figure 3.1 for
definitionandFigure3.5forABAQUSboundaryconditions)versusafullCsection.The
slottedholehasdimensionsofhhole=1.5in.,Lhole=4in.,andrhole=0.75in.Theplatewidths
are chosen to correspond with the flat web widths of standard SSMA structural studs
(SSMA 2001). Plate aspect ratios are held constant at 4:1. From each plate, a full
structural stud finite element model is developed for comparison. The SSMA member
designations and cross section dimensions considered in this study are listed in Table
4.1.
Table4.1SSMAstructuralstudandplatedimensions
SSMA
Designation
250S162-33
350S162-33
362S162-33
400S162-33
550S162-33
600S162-33
800S162-33
H
in.
2.50
3.50
3.62
4.00
5.50
6.00
8.00
B
in.
1.63
D
r
t
in.
in.
in.
0.50 0.0764 0.0346
h
in.
2.28
3.28
3.40
3.78
5.28
5.78
7.78
b hhole/h
in.
1.40 0.66
0.46
0.44
0.40
0.28
0.26
0.19
L=4h
in.
9.1
13.1
13.6
15.1
21.1
23.1
31.1
Beforeexaminingtheelasticbucklingload,considertheobservedchangesinthefirst
mode shape caused by the addition of a hole as given in Figure 4.3. For the buckled
69
shapeoftheSSMA250S16233inFigure4.3a,thenumberofbuckledhalfwaveschanges
fromfourtothreefortheisolatedplateandfromfivetotwoforthefullmember,when
theholeisadded.Thestripsofplateadjacenttotheholearestiffenedbytheconnected
flangeinthefullmember,causingbuckledhalfwavestoforminthewebawayfromthe
hole.Also,thelengthofthehole,Lhole,isapproximatelyhalfofthelengthofthemember
L in the SSMA 250S16233 member which also prevents local buckling in the web. In
Figure4.3b,theholedecreasesthenumberofbuckledhalfwavesfromfourtothreein
theSSMA440S16233isolatedplatebutdoesnotchangethenumberofhalfwavesinthe
fullmember.Thecrosssectionconnectivityofthefullmemberlimitsdeformationatthe
holeandencouragesbucklinghalfwavestoformalongtheentiremember.Also,there
ismorewebmaterialtoaccommodatelocalbucklingalongthelength(Lhole/L=0.26)when
comparedtotheSSMA250S16233member.
# of Local Buckling
Half-Waves (Typ.)
4
3
2
3
5
(a)
(b)
SSMA 400S162-33
hhole/h=0.40
SSMA 250S162-33
hhole/h=0.66
Figure4.3(a)SSMA250S16233webplateandstructuralstud,and(b)SSMA400S16233webplateand
structuralstud
Figure 4.4 presents the influence of a slotted hole on the critical elastic buckling
stress fcr of the isolated web plates and full members with holes from Table 4.1. These
resultsarecomparedtotheelasticbucklingpredictionforastiffenedelementwithholes
70
developedandpresentedinSection3.3.4.Theinfluenceoftheholeisminimalforsmall
hole width to plate width ratios, but increases to a maximum at hhole/h=0.30 for the
ABAQUS plate results (consistent with the stiffened element prediction). fcr increases
withincreasinghhole/hforfullmembers,demonstratingthatthecrosssectionconnectivity
decreases a members sensitivity to a hole (especially in the range of hhole/h=0.30). The
webisstiffenedthroughbeneficialwebflangeinteractioncreatedbytherelativelystable
edgestiffenedflange.
Asnormalizedholewidthincreases,theelasticbucklingloadexceedsthatofaplate
withoutahole.Thisincreaseinbucklingloadisattributedtotheincreasedaxialstiffness
fromthestripsadjacenttothehole,whichcauseslocalbucklingtooccurawayfromthe
holes (this unstiffened strip effect is discussed in Section 3.3). The length of hole
relativetothelengthofthememberalsocontributestotheincreaseinfcr.Lhole/Lincreases
with increasing hhole/h in this study since L=4h. As demonstrated in Figure 4.3a, the
removal of web material restricts the formation of local buckling in the web of the
member, resulting in shortened halfwaves away from the hole with increased axial
stiffness. The stiffened element prediction in Figure 4.4 is conservative here when
comparedtotheplateresultsbecauseitwasdevelopedforevenlyspacedholesinlong
platesanddoesnotaccountfortheLhole/Lboostinfcr.
71
1.6
ABAQUS, SSMA Stud
ABAQUS, stiffened element
Stiffened element prediction
1.4
1.2
1
0.8
0.6
0.4
0.2
0
0.1
0.2
0.3
0.4
0.5
hhole/h
0.6
0.7
0.8
0.9
Figure4.4.Effectofaslottedholeontheelasticbucklingloadofsimplysupportedplatesandstructural
studs
4.2.4 Influence of slotted hole location on elastic buckling

of an intermediate length structural stud
Thisstudyinvestigatestheelasticbucklingbehaviorofanintermediatelengthcold
formed steel column with one slotted hole. The primary goal here is to identify and
formally define the elastic buckling modes that will be used as predictors of ultimate
strengthwithintheDirectStrengthMethod.Theelasticbucklingbehavioriscompared
as the location of a slotted hole is varied along the length of the column. The typical
compressionmemberinthisstudyhasalength L of48inchesandismodeledwithan
SSMA 36216233 structural channel cross section. A single slotted hole is centered
transverselyintheweb.Theslottedholehasdimensionsofhhole=1.5in.,Lhole=4in.,and
rhole=0.75in.Table4.1summarizesthedimensionsoftheSSMA36216233crosssection.
TheABAQUScolumnboundaryconditionsareconsistentwithFigure4.2.
72
Figure 4.5 compares the local buckling (L) mode shapes of the column with and
without a slotted hole. The lowest buckling mode is local buckling (L) and exists for
boththecolumnwithandwithoutthehole.Thelocationoftheholedoesnotinfluence
PcrforthismodeasobservedinFigure4.8.Platebucklingofthewebawayfromthehole
dominatesforthismode,regardlessofholelocation.
Two unique local buckling modes to the column with a hole, LH and LH2, are
alsoidentifiedintheeigenbucklinganalyses.Thesemodes,showninFigure4.5,exhibit
buckling of the unstiffened strip adjacent to the hole similar to that observed in the
crosssectionalelementstudieswithholesinChapter3.TheLHmodeoccurswhenboth
unstiffenedstripsbuckleinthesamedirectionnormaltothewebplane,whichincreases
the distortional tendencies of the flange in the vicinity of the hole. This localized
distortional buckling is observed in Figure 4.6, which compares the LH modes as the
locationoftheholevariesalongthecolumnlength.TheLHmodeisconsistentwiththe
elastic buckling mode shapes of stiffened elements, where the presence of a hole is
observedtoreducethetransversebendingstiffnesscausinglocalizeddeformationatthe
hole(seeFigure3.9a).
TheLH2modeoccurswhentheunstiffenedstripsbuckleinoppositedirections
relativetothewebplate,resultinginantisymmetricdistortionaldeformationatthehole.
Figure4.8demonstratesthatPcrforthesetwomodesissimilarandthatbothmodesare
minimallyaffectedbythelongitudinallocationoftheholeinthecolumn.
73
LH
LH2
Figure4.5Thepresenceofaholecreatesuniquelocalbucklingmodeswhereunstiffenedstripbuckling
adjacenttotheholeoccurssymmetrically(LH)orasymmetrically(LH2)increasethedistortionaltendency
oftheflanges
C
L hole
Figure4.6SSMAslottedholelocationandlocalbucklingLHmode,L=48in.,x/L=0.06,0.125,0.25,0.375,0.50.
Notethedistortionaltendenciesoftheflangesatthehole.
The pure distortional buckling mode for the column, D, is compared for a column
with and without a hole in Figure 4.7a. Note that distortional halfwavelength is
unchangedwiththepresenceofthehole,althoughlocalbucklingwithhalfwavelengths
shorter than the fundamental L halfwavelength (see Figure 4.5) mix with the D mode
74
forthememberwiththehole.Figure4.8demonstratesthatthepresenceoftheholehas
aminimalinfluenceonPcrdregardlessoflongitudinallocation.
The lowest global buckling mode is flexuraltorsional buckling (GFT) as shown in
Figure4.7b.ThepresenceofaholeresultsinaslightdecreaseinPcreastheholemoves
towards the end of the column as shown in Figure 4.8. As the hole shifts close to the
loadedendofthecolumn(x/L=0.06),localbucklingattheholecombineswiththeGFT
modetoreducePcre.
GFT
Figure4.7Influenceofaslottedholeonthe(a)distortional(D)and(b)globalflexuraltorsional(GFT)modes
ofacoldformedsteelcolumn
75
1
D
GFT
LH2
LH
L
0.9
0.8
D, no hole
0.7
Pcr/Py
0.6
GFT, no hole
0.5
0.4
0.3
0.2
L, no hole
0.1
0
0.1
0.2
0.3
0.4
x/L
0.5
0.6
0.7
0.8
Figure4.8InfluenceofSSMAslottedholelocationonPcrfora362S16233Csection(refertoFigure4.5,
Figure4.6,andFigure4.7forbuckledshapesummaries)
4.2.5 Flange hole study

The research focus up until now has been on the elastic buckling modes of
isolatedwebplatesandcoldformedsteelcompressionmemberswithwebholes.Holes
arealsocommonlypresentintheflangesofCsectioncolumns.Astandardconnection
detail requiring a flange hole is shown in Figure 4.9, where gypsum sheathing is
connected to steel studs with a bolted or screw attachment (Western States Clay
ProductsAssociation2004).
76

Figure4.9Connectiondetailforstructuralstudtoexteriorwallrequiresascreworboltholeplacedinthe
studflange(WesternStatesClayProductsAssociation2004)
Thisstudyevaluatestheinfluenceofcircularflangeholesontheelasticbuckling
behavior of an intermediate length SSMA 362S16233 structural stud. A single hole is
placed at the midlength of both the top and bottom flanges and centered between the
web and lip stiffeners. Five hole diameters consistent with standard bolt holes are
considered:bhole/b=0.178,0.356,0.534,0.713,and0.892(,,,1,1holesina1
flange) where the flat flange width b=1.40 in. The length L is 48 in. for all members,
correspondingtoacommonunbracedlengthofaSSMAstructuralstud.
Figure 4.10 presents the variation in elastic buckling loads for local, distortional,
and global modes as the diameter of the flange holes increase. The local (L) buckling
load, Pcrl, is not influenced by small holes, but decreases as bhole/b exceeds 0.70. Figure
4.11 demonstrates that for large flange holes local buckling is dominated by web and
flangedeformationneartheholes.Thelargeflangeholesadverselyaffectthebeneficial
webflangeinteractioninherentinstructuralstuds(Figure 4.4highlightsthisbeneficial
interaction for Csections with web holes). The interruption of the webflange
interaction by the holes is also reflected in the pure distortional mode (D), as Pcrd
77
decreases slightly as flange hole size increases relative to flange width. The flanges
holes have a minimal effect on the global flexuraltorsional mode (GFT) because their
diameterissmallrelativetothecolumnlength.
1
D
GFT
L
0.9
0.8
D, no hole
0.7
Pcr/Py
0.6
GFT, no hole
0.5
0.4
L, no hole
0.3
0.2
0.1
0
0.1
0.2
0.3
0.4
0.5
bhole/b
0.6
0.7
0.8
0.9
Figure4.10Influenceofflangeholediameteronthelocal(L),distortional(D),andglobal(GFT)elastic
bucklingloadsofanSSMA362S16233structuralstud
Figure4.11Local(L)bucklingisdominatedbyflangeandwebdeformationneartheholesasbhole/bexceeds
0.70
78
4.2.6 Analysis of existing experiments on columns
TheDirectStrengthMethodemploystheelasticbucklingpropertiesofacoldformed
steel member to predict its ultimate strength. To assist in the extension of DSM to
columnswithholes,adatabaseisdevelopedinthissectionwhichsummarizestheelastic
bucklingpropertiesandtestedstrengthsofcoldformedsteelcolumnsexperimentswith
holes from the past 30 years. This database is used Chapter 8 when developing and
verifying DSM for columns with holes. Table 4.2 summarizes the experimental
programscomprisingthedatabase.
Table4.2Summaryofcolumnexperimentaldata
Author
Ortiz-Colberg
Ortiz-Colberg
Miller and Pekz
Sivakumaran
Abdel-Rahman
Pu et al.
Moen and Schafer
Moen and Schafer
4.2.6.1
Publication Date
1981
1981
1994
1987
1997
1999
2008
2008
Types of Specimens
Stub Column
Long Column
Stub Column
Stub Column
Stub Column
Stub Column
Short Column
Intermediate Column
Cross Section
Lipped Cee Channel
End Conditions
Fixed-Fixed
Weak axis pinned
Fixed-Fixed
Fixed-Fixed
Fixed-Fixed
Fixed-Fixed
Fixed-Fixed
Fixed-Fixed
# of Specimens
8
15
12
14
8
9
6
6
Elasticbucklingdatabaseforcolumnexperiments
ABAQUSeigenbucklinganalyseswereconductedforeachspecimeninthedatabase.
Member boundary conditions and loading conditions were modeled to be consistent

with the actual experimental conditions. The ABAQUS implementations of the
boundaryconditionsforeachexperimentalprogramaredescribedinFigure4.12.Local
(L, LH, LH2 see Figure 4.5), distortional (D), and global (G) buckling modes were
manually identified from the buckled modes in ABAQUS using the modal definitions
describedin4.2.4.
79
Nodes bearing on top platen

constrained in 1, 2 and 3
Boundary conditions valid for:
Ortiz-Colberg 1981 (stub columns)

Sivakumaran 1987
Pu et al. 1999
Moen and Schafer 2008
Specimens bear directly on the

platens, ends are not welded
1
2
3
4
6
ABAQUS tied rigid body reference

node constrained in 1, 3 to 6 directions,
ensures that all nodes on contact
surface translate together in 1 direction
and cannot rotate at the platen (crosssection contact edge is welded). The
entire platen can rotate in the 2
direction.

constrained in 1 to 6
(contact edge is welded)
ABAQUS pinned rigid body

reference node constrained in 2 to 6
directions, ensures that all nodes on
loaded surface translate together in 1
direction but can rotate freely at the
platen (no welding)

Miller and Pekoz 1994
Abdel-Rahman 1997
Ortiz-Colberg 1981 (long column)
Specimens are welded to loading

platens, cross-section edge rotation is
restrained at contact location
Specimens are welded to loading

platens, which are attached to pins that
allow weak axis rotation of the platens
ABAQUS tied rigid body reference

node constrained in 2 to 6 directions,
ensures that all nodes on loaded
surface translate together in 1 direction
and cannot rotate at the platen (crosssection contact edge is welded)
ABAQUS tied rigid body reference node

constrained in 3 to 6 directions, ensures that
all nodes on loaded surface translate
together in 1 direction and cannot rotate at
the platen (cross-section contact edge is
welded). The entire platen can also rotate in
the 5 direction.
(a)
(b)
(c)
Figure4.12ExperimentalprogramboundaryconditionsasimplementedinABAQUS
Table 4.3 summarizes the fixedfixed column specimen dimensions and material
properties,includingcrosssectionandholedimensions,testedultimateloadPtest,tested
specimen yield stress Fy, specimen yield force Py,g (calculated with the gross cross
sectional area), and Py,net (calculated with the net crosssectional area). Table 4.4
summarizes the fixedfixed column specimen elastic buckling loads. ABAQUS
eigenbucklingresultsarepresentedfortwodifferenttypesofboundaryconditions,the
experimentboundaryconditionsandCUFSMboundaryconditions(warpingfreeatthe
ends, warpingfixed at the midlength of the column) except for the Moen and Schafer
specimens which were only modeled with experiment boundary conditions. CUFSM
elasticbucklingresultsarealsoprovided,includingthedistortionalhalfwavelengthLcrd.
Thesameexperimentandelasticbucklinginformationissummarizedfortheweakaxis
pinned columns in Table 4.5 and Table 4.6 and together with Table 4.3 and Table 4.4
comprisethedatasetusedtoevaluatetheDSMequationsforcoldformedsteelcolumns
withholesChapter8.
80
Table4.7summarizescrosssectionandmaterialpropertyparameterrangesforthe
fixedfixedspecimensandweakaxispinnedspecimens.Mostoftheweakaxispinned
specimens are long columns, while the majority of the fixedfixed specimens are stub
columns (the exception being the short and intermediate length fixedfixed columns
tested by Moen and Schafer). All column specimens in the database are common
industry shapes and meet the DSM prequalification standards (for members without
holes)summarizedinTable4.8(AISIS1002007).
81
Table4.3Fixedfixedcolumnexperimentdimensionsandmaterialproperties
Member
Study and Specimen Name
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Abdel-Rahman 1997
Abdel-Rahman 1997
Abdel-Rahman 1997
Abdel-Rahman 1997
Abdel-Rahman 1997
Abdel-Rahman 1997
Abdel-Rahman 1997
Abdel-Rahman 1997
Pu et al. 1999
Pu et al. 1999
Pu et al. 1999
Pu et al. 1999
Pu et al. 1999
Pu et al. 1999
Pu et al. 1999
Pu et al. 1999
Pu et al. 1999
Sivakumaran 1987
Sivakumaran 1987
Sivakumaran 1987
Sivakumaran 1987
Sivakumaran 1987
Sivakumaran 1987
Sivakumaran 1987
Sivakumaran 1987
Sivakumaran 1987
Sivakumaran 1987
Sivakumaran 1987
Sivakumaran 1987
Sivakumaran 1987
Sivakumaran 1987
Miller & Pekoz 1994
Miller & Pekoz 1994
Miller & Pekoz 1994
Miller & Pekoz 1994
Miller & Pekoz 1994
Miller & Pekoz 1994
Miller & Pekoz 1994
Miller & Pekoz 1994
Miller & Pekoz 1994
Miller & Pekoz 1994
Miller & Pekoz 1994
Miller & Pekoz 1994
* average of two tests
S4
S7
S6
S8
S5
S3
S14
S15
A-C
A-S
A-O
A-R
B-C
B-S
B-O
B-R
C-2.0-1-30-1
C-2.0-1-30-2
C-2.0-1-30-3
C-1.2-1-30-1
C-1.2-1-30-2
C-1.2-1-30-3
C-0.8-1-30-1
C-0.8-1-30-2
C-0.8-1-30-3
A2
A3
A4
A5
A6
A7
A8
B2
B3
B4
B5
B6
B7
B8
1-12
1-13
1-17
1-19
2-11
2-12
2-14
2-15
2-16
2-24
2-25
2-26
362-1-24-H
362-2-24-H
362-3-24-H
362-1-48-H
362-2-48-H
362-3-48-H
600-1-24-H
600-2-24-H
600-3-24-H
600-1-48-H
600-2-48-H
600-3-48-H
Boundary
Conditions
in.
Fixed-fixed 12.00
12.00
12.00
12.00
12.00
12.00
12.00
12.00
16.73
16.73
18.70
18.70
9.84
9.84
11.81
11.81
14.57
14.57
14.57
14.17
14.17
14.17
14.17
14.17
14.17
7.87
7.87
7.87
7.87
7.87
7.87
8.78
10.43
10.43
10.43
10.43
10.43
10.43
10.43
10.87
10.87
17.95
17.95
10.87
10.87
17.95
17.95
17.95
17.95
17.95
17.95
24.1
24.1
24.1
48.22
48.23
48.2
24.1
24.1
24.1
48.09
48.25
48.06
Material
Hole Dimensions
Cross Section Dimensions
Yield Stress and Force
Test
nu
Hole Type
L hole
h hole
r hole
B1
B2
D1
D2
Fy
P y,g
P y,net
P test
in.
0.0492
0.0493
0.0496
0.0496
0.0498
0.0499
0.0760
0.0760
0.0740
0.0740
0.0740
0.0740
0.0500
0.0500
0.0500
0.0500
0.0787
0.0787
0.0787
0.0472
0.0472
0.0472
0.0315
0.0315
0.0315
0.0630
0.0630
0.0630
0.0630
0.0630
0.0630
0.0630
0.0508
0.0508
0.0508
0.0508
0.0508
0.0508
0.0508
0.0756
0.0752
0.0346
0.0346
0.0752
0.0752
0.0350
0.0346
0.0350
0.0354
0.0354
0.0350
0.0391
0.0383
0.0394
0.0393
0.0391
0.0399
0.0421
0.0412
0.043
0.0428
0.0429
0.0431
ksi
29420
29420
29420
29420
29420
29420
29420
29420
29420
29420
29420
29420
29420
29420
29420
29420
29420
29420
29420
29420
29420
29420
29420
29420
29420
29710
29710
29710
29710
29710
29710
29710
30435
30435
30435
30435
30435
30435
30435
29420
29420
29420
29420
29420
29420
29420
29420
29420
29420
29420
29420
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
Circular
Circular
Circular
Circular
Circular
Circular
Circular
Circular
Circular
Square
Oval
Rectangle
Circular
Square
Oval
Rectangle
Square
Square
Square
Square
Square
Square
Square
Square
Square
Circular
Square
Circular
Square
Circular
Square
Oval
Circular
Square
Circular
Square
Circular
Square
Oval
Rectangular
Rectangular
Rectangular
Rectangular
Rectangular
Rectangular
Rectangular
Rectangular
Rectangular
Rectangular
Rectangular
Rectangular
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
in.
0.75
1.50
1.25
1.75
1.04
0.50
1.04
1.50
2.50
2.50
4.50
4.50
1.50
1.50
4.00
4.00
1.06
1.05
1.05
1.04
1.04
1.04
1.04
1.04
1.04
0.65
0.65
1.30
1.30
1.95
1.95
4.02
1.14
1.14
2.28
2.28
3.43
3.43
4.02
2.76
2.76
2.24
2.24
2.56
2.56
2.24
2.24
2.24
2.24
2.24
2.24
4.00
4.00
4.01
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
in.
0.75
1.50
1.25
1.75
1.04
0.50
1.04
1.50
2.50
2.50
2.50
2.50
1.50
1.50
1.50
1.50
1.06
1.05
1.05
1.04
1.04
1.04
1.04
1.04
1.04
0.65
0.65
1.30
1.30
1.95
1.95
1.50
1.14
1.14
2.28
2.28
3.43
3.43
1.50
1.61
1.61
1.57
1.57
1.50
1.50
1.57
1.57
1.57
1.57
1.57
1.57
1.49
1.50
1.49
1.50
1.50
1.49
1.50
1.49
1.49
1.49
1.50
1.50
in.
0.38
0.75
0.63
0.88
0.52
0.25
0.52
0.75
1.25
--1.25
--0.75
--0.75
--------------------0.32
--0.65
--0.97
--0.75
0.57
--1.14
--1.71
--0.75
------------------------0.75
0.75
0.75
0.75
0.75
0.75
0.75
0.75
0.75
0.75
0.75
0.75
in.
3.50
3.51
3.51
3.51
3.50
3.50
3.52
3.52
7.99
7.99
7.99
7.99
4.00
4.00
4.00
4.00
3.94
3.94
3.94
3.87
3.87
3.87
3.84
3.84
3.84
3.63
3.63
3.63
3.63
3.63
3.63
3.63
6.00
6.00
6.00
6.00
6.00
6.00
6.00
3.62
3.62
5.98
5.98
3.62
3.62
5.98
5.98
5.98
5.98
5.98
5.98
3.58
3.64
3.67
3.62
3.62
3.63
6.04
6.01
6.03
6.01
6.02
6.06
in.
1.62
1.63
1.61
1.62
1.62
1.61
1.67
1.67
1.64
1.64
1.64
1.64
1.64
1.64
1.64
1.64
2.05
2.05
2.05
2.05
2.05
2.05
2.05
2.05
2.05
1.63
1.63
1.63
1.63
1.63
1.63
1.63
1.63
1.63
1.63
1.63
1.63
1.63
1.63
1.46
1.46
1.34
1.34
1.46
1.46
1.38
1.38
1.38
1.38
1.38
1.38
1.65
1.63
1.67
1.60
1.59
1.60
1.59
1.61
1.61
1.60
1.59
1.63
in.
1.49
1.49
1.48
1.48
1.48
1.48
1.49
1.49
1.64
1.64
1.64
1.64
1.64
1.64
1.64
1.64
2.05
2.05
2.05
2.05
2.05
2.05
2.05
2.05
2.05
1.63
1.63
1.63
1.63
1.63
1.63
1.63
1.63
1.63
1.63
1.63
1.63
1.63
1.63
1.46
1.46
1.34
1.34
1.46
1.46
1.38
1.38
1.38
1.38
1.38
1.38
1.60
1.59
1.70
1.60
1.61
1.61
1.61
1.60
1.58
1.63
1.61
1.59
in.
0.49
0.50
0.49
0.49
0.49
0.48
0.51
0.51
0.51
0.51
0.51
0.51
0.51
0.51
0.51
0.51
0.63
0.63
0.63
0.63
0.63
0.63
0.63
0.63
0.63
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.47
0.47
0.31
0.31
0.47
0.47
0.31
0.31
0.31
0.31
0.31
0.31
0.43
0.44
0.42
0.42
0.42
0.40
0.48
0.37
0.36
0.48
0.48
0.37
in.
0.50
0.51
0.51
0.50
0.50
0.50
0.51
0.51
0.51
0.51
0.51
0.51
0.51
0.51
0.51
0.51
0.63
0.63
0.63
0.63
0.63
0.63
0.63
0.63
0.63
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.47
0.47
0.31
0.31
0.47
0.47
0.31
0.31
0.31
0.31
0.31
0.31
0.44
0.39
0.43
0.41
0.40
0.43
0.36
0.50
0.48
0.39
0.36
0.48
in.
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.15
0.15
0.15
0.15
0.10
0.10
0.10
0.10
0.16
0.16
0.16
0.11
0.11
0.11
0.08
0.08
0.08
0.13
0.13
0.13
0.13
0.13
0.13
0.13
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.20
0.20
0.19
0.19
0.19
0.18
0.16
0.15
0.16
0.16
0.16
0.16
ksi
47.1
48.5
51.5
51.5
49.6
49.6
47.4
47.6
55.8
55.8
55.8
55.8
46.2
46.2
46.2
46.2
44.4
33.6
34.4
28.0
28.0
28.0
24.8
24.8
24.8
49.4
49.4
49.4
49.4
49.4
49.4
49.4
38.1
38.1
38.1
38.1
38.1
38.1
38.1
51.9
51.9
44.8
44.8
53.0
53.0
43.8
43.8
43.8
43.8
43.8
43.8
57.9
57.1
56.0
58.6
59.7
58.3
61.9
58.4
60.1
61.4
62.0
61.5
kips
16.7
17.3
18.4
18.4
17.7
17.7
25.8
25.9
48.3
48.3
48.3
48.3
18.2
18.2
18.2
18.2
30.2
22.8
23.4
11.6
11.6
11.6
7.0
7.0
7.0
22.9
22.9
22.9
22.9
22.9
22.9
22.9
19.0
19.0
19.0
19.0
19.0
19.0
19.0
27.3
27.1
13.9
13.9
27.8
27.8
13.9
13.7
13.9
14.0
14.0
13.9
16.4
15.7
16.4
16.6
16.8
16.8
25.0
23.1
24.7
25.2
25.5
25.6
kips
14.9
13.7
15.2
13.9
15.2
16.5
22.0
20.4
37.9
37.9
37.9
37.9
14.8
14.8
14.8
14.8
26.5
20.1
20.6
10.3
10.3
10.3
6.2
6.2
6.2
20.9
20.9
18.8
18.8
16.8
16.8
18.2
16.8
16.8
14.6
14.6
12.4
12.4
16.1
20.9
20.8
11.5
11.5
21.8
21.8
11.5
11.3
11.5
11.6
11.6
11.5
13.0
12.4
13.1
13.1
13.3
13.4
21.1
19.5
20.9
21.3
21.5
21.6
kips
14.2
12.7
13.8
13.6
14.1
14.5
24.6
24.0
26.5*
26.8*
26.6*
25.8*
12.7*
12.7*
12.6*
12.8*
23.6
18.3
18.3
9.4
9.4
9.4
4.6
4.5
4.6
19.3
19.0
18.4
18.3
17.6
17.4
16.3
12.1
12.0
12.0
11.5
10.6
10.6
11.6
25.8
23.6
5.5
5.9
22.2
22.1
6.0
5.8
5.8
6.1
6.1
6.2
10.0
10.4
9.9
9.0
9.2
9.4
12.1
11.6
11.8
11.2
11.7
11.2
82
Table4.4Fixedfixedcolumnexperimentelasticbucklingproperties
ABAQUS elastic buckling with hole,
experiment boundary conditions

Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Abdel-Rahman 1997
Abdel-Rahman 1997
Abdel-Rahman 1997
Abdel-Rahman 1997
Abdel-Rahman 1997
Abdel-Rahman 1997
Abdel-Rahman 1997
Abdel-Rahman 1997
Pu et al. 1999
Pu et al. 1999
Pu et al. 1999
Pu et al. 1999
Pu et al. 1999
Pu et al. 1999
Pu et al. 1999
Pu et al. 1999
Pu et al. 1999
Sivakumaran 1987
Sivakumaran 1987
Sivakumaran 1987
Sivakumaran 1987
Sivakumaran 1987
Sivakumaran 1987
Sivakumaran 1987
Sivakumaran 1987
Sivakumaran 1987
Sivakumaran 1987
Sivakumaran 1987
Sivakumaran 1987
Sivakumaran 1987
Sivakumaran 1987
Miller & Pekoz 1994
Miller & Pekoz 1994
Miller & Pekoz 1994
Miller & Pekoz 1994
Miller & Pekoz 1994
Miller & Pekoz 1994
Miller & Pekoz 1994
Miller & Pekoz 1994
Miller & Pekoz 1994
Miller & Pekoz 1994
Miller & Pekoz 1994
Miller & Pekoz 1994
S4
S7
S6
S8
S5
S3
S14
S15
A-C
A-S
A-O
A-R
B-C
B-S
B-O
B-R
C-2.0-1-30-1
C-2.0-1-30-2
C-2.0-1-30-3
C-1.2-1-30-1
C-1.2-1-30-2
C-1.2-1-30-3
C-0.8-1-30-1
C-0.8-1-30-2
C-0.8-1-30-3
A2
A3
A4
A5
A6
A7
A8
B2
B3
B4
B5
B6
B7
B8
1-12
1-13
1-17
1-19
2-11
2-12
2-14
2-15
2-16
2-24
2-25
2-26
362-1-24-H
362-2-24-H
362-3-24-H
362-1-48-H
362-2-48-H
362-3-48-H
600-1-24-H
600-2-24-H
600-3-24-H
600-1-48-H
600-2-48-H
600-3-48-H
Boundary
Conditions
Fixed-fixed
Pcrl, LH
Pcrl, LH2
Pcrd
Pcre
Pcrl
Pcrd
Pcre
Pcrl
Pcrd
Lcrd
kips
10.7
11.9
11.6
12.5
11.4
11.2
40.5
43.5
16.2
13.7
12.3
12.9
11.2
12.1
11.8
12.3
42.6
42.5
42.6
9.5
9.5
9.5
2.8
2.8
2.8
21.3
21.5
23.5
24.8
30.5
33.4
30.2
6.0
6.1
7.0
7.6
10.3
12.4
6.6
43.2
42.5
1.7
1.7
41.3
41.3
1.8
1.7
1.8
1.9
1.9
1.8
5.9
5.4
5.7
5.3
5.2
5.7
3.3
3.2
3.5
3.4
3.4
3.4
kips
31.5
25.6
25.5
34.6
33.0
--45.5
50.1
12.8
22.4
16.4
15.8
39.5
42.0
23.0
22.7
53.0
52.9
52.9
------------------------30.8
----11.2
--20.2
19.9
11.7
51.1
50.3
2.4
2.4
41.7
41.7
2.5
2.4
2.5
2.6
2.6
2.5
6.4
5.7
6.6
5.7
5.8
6.2
3.1
2.9
3.3
3.2
3.2
3.2
kips
--41.9
33.0
--41.1
--66.9
--21.6
22.5
19.5
19.0
42.4
----29.9
85.5
85.5
85.5
48.5
48.5
48.5
17.4
17.4
17.4
------------32.2
10.4
10.6
16.5
18.6
----16.4
51.6
50.9
2.9
2.9
47.2
47.2
3.0
2.9
3.0
3.1
3.1
3.0
-------------------------
kips
40.0
43.3
41.8
42.8
44.5
41.1
86.5
70.3
24.0
24.7
23.7
24.1
45.3
45.9
30.6
30.5
109.0
109.0
109.0
50.6
50.6
50.6
17.7
17.7
17.7
57.0
58.0
50.7
50.9
81.2
81.2
81.2
12.0
12.3
19.2
19.3
21.1
21.2
18.4
76.4
75.3
3.3
3.3
74.4
74.4
3.4
3.3
3.4
3.5
3.5
3.4
9.2
10.3
9.5
9.1
9.0
9.0
7.0
6.7
7.3
5.1
5.0
5.0
kips
640.0
640.0
640.0
640.0
640.0
640.0
964.0
964.0
1014.2
1014.2
811.8
811.8
1271.6
1271.6
883.3
883.3
1182.8
1182.8
1182.8
777.4
777.4
777.4
528.3
528.3
528.3
2045.6
2045.6
2045.6
2045.6
2045.6
2045.6
1631.1
1742.4
1742.4
1742.4
1742.4
1742.4
1742.4
1742.4
1089.0
1084.1
212.7
212.7
1084.1
1084.1
231.1
231.1
231.1
231.1
231.1
231.1
119.3
112.8
130.6
30.0
29.7
36.2
239.3
238.4
242.6
56.3
53.0
55.8
kips
10.5
11.1
11.0
11.4
10.9
11.0
40.1
40.6
11.4
11.7
11.1
11.5
9.6
9.8
9.9
10.0
41.1
41.1
41.1
9.2
9.2
9.2
2.7
2.7
2.7
20.8
20.8
21.8
22.6
25.2
26.1
25.9
5.8
5.9
6.3
6.7
8.1
9.1
5.9
37.4
36.8
1.7
1.7
36.5
36.5
1.7
1.7
1.7
1.8
1.8
1.7
-------------------------
kips
17.3
17.6
17.7
17.9
17.8
18.0
38.8
40.0
14.0
14.5
13.1
13.1
22.3
23.3
20.8
20.4
52.3
52.3
52.3
19.0
19.0
19.0
11.3
11.3
11.3
35.2
35.4
38.3
39.5
35.1
36.2
33.3
10.5
10.6
12.5
13.4
16.3
15.8
16.5
35.5
35.0
2.3
2.3
35.1
35.1
2.4
2.3
2.4
2.4
2.4
2.4
-------------------------
kips
-------------------------------------------------------------------------------------------------------------------------------
kips
10.8
10.9
11.1
11.1
11.2
11.3
39.8
39.8
11.7
11.7
11.7
11.7
9.6
9.6
9.6
9.6
42.5
42.5
42.5
9.4
9.4
9.4
2.8
2.8
2.8
22.0
22.0
22.0
22.0
22.0
22.0
22.0
5.7
5.7
5.7
5.7
5.7
5.7
5.7
36.0
35.5
1.7
1.7
35.5
35.5
1.7
1.7
1.7
1.8
1.8
1.7
5.7
5.3
5.6
5.2
5.1
5.5
3.4
3.0
3.2
3.3
3.3
3.4
kips
17.7
18.0
18.0
18.0
18.1
18.1
45.5
45.4
13.8
13.8
13.8
13.8
16.9
16.9
16.9
16.9
49.9
49.9
49.9
17.2
17.2
17.2
7.5
7.5
7.5
29.4
29.4
29.4
29.4
29.4
29.4
29.4
9.8
9.8
9.8
9.8
9.8
9.8
9.8
42.1
41.6
2.1
2.1
41.6
41.6
2.1
2.1
2.1
2.1
2.1
2.1
9.5
9.0
9.3
9.0
8.9
8.9
4.9
4.9
5.0
5.1
5.1
5.1
in.
13.8
13.8
13.8
13.8
13.8
13.7
11.3
11.3
15.3
15.3
15.3
15.3
15.9
15.9
15.9
15.9
14.9
14.9
14.9
22.2
22.2
22.2
27.3
27.3
27.3
13.9
13.9
13.9
13.9
13.9
13.9
13.9
16.8
16.8
16.8
16.8
16.8
16.8
16.8
9.7
9.7
8.3
8.3
9.7
9.7
8.3
8.3
8.3
8.3
8.3
8.3
15.7
15.7
15.7
15.7
15.7
15.7
14.8
14.8
14.8
14.8
14.8
14.8
83

CUFSM elastic buckling, no hole
CUFSM boundary conditions
Pcrl
Table4.5Weakaxispinnedcolumnexperimentdimensionsandmaterialproperties
Member
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
L2
L3
L6
L7
L9
L10
L14
L16
L17
L19
L22
L26
L27
L28
L32
Boundary
Conditions
Weak-axis
pinned
Material
Hole Dimensions
nu
in.
63.0
27.0
63.0
63.0
39.0
38.9
39.1
51.0
51.1
27.0
45.0
45.0
27.0
27.0
63.0
in.
0.0490
0.0490
0.0490
0.0490
0.0490
0.0490
0.0490
0.0760
0.0760
0.0760
0.0760
0.0760
0.0760
0.0760
0.0760
ksi
29420
29420
29420
29420
29420
29420
29420
29420
29420
29420
29420
29420
29420
29420
29420
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
Hole Type
Circular
Circular
Circular
Circular
Circular
Circular
Circular
Circular
Circular
Circular
Circular
Circular
Circular
Circular
Circular
Yield Stress and Force
h hole
r hole
B1
B2
D1
D2
Fy
P y,g
P y,net
P test
in.
0.50
1.00
1.00
1.50
1.00
1.50
0.50
1.00
1.50
1.50
1.50
1.00
1.00
1.00
1.00
in.
0.50
1.00
1.00
1.50
1.00
1.50
0.50
1.00
1.50
1.50
1.50
1.00
1.00
1.00
1.00
in.
0.25
0.50
0.50
0.75
0.50
0.75
0.25
0.50
0.75
0.75
0.75
0.50
0.50
0.50
0.50
in.
3.51
3.51
3.51
3.51
3.51
3.51
3.51
3.51
3.51
3.51
3.51
3.51
3.51
3.51
3.51
in.
1.62
1.62
1.62
1.62
1.62
1.62
1.62
1.62
1.62
1.62
1.62
1.62
1.62
1.62
1.62
in.
1.48
1.48
1.48
1.48
1.48
1.48
1.48
1.48
1.48
1.48
1.48
1.48
1.48
1.48
1.48
in.
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
in.
0.51
0.51
0.51
0.51
0.51
0.51
0.51
0.51
0.51
0.51
0.51
0.51
0.51
0.51
0.51
in.
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
ksi
45.7
42.9
46.1
45.5
43.8
42.3
42.9
48.1
48.1
51.5
46.7
45.8
48.3
42.3
47.9
kips
16.2
15.2
16.3
16.1
15.5
15.0
15.2
25.9
25.9
27.7
25.1
24.7
26.0
22.8
25.8
kips
15.0
13.1
14.0
12.7
13.3
11.8
14.1
22.2
20.4
21.9
19.8
21.2
22.3
19.6
22.1
kips
8.5
11.4
8.5
8.5
9.4
10.1
9.6
17.2
15.0
21.2
20.0
19.1
21.9
22.4
13.3
Table4.6Weakaxispinnedcolumnexperimentelasticbucklingproperties
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
Ortiz-Colberg 1981
L2
L3
L6
L7
L9
L10
L14
L16
L17
L19
L22
L26
L27
L28
L32

Boundary
Conditions
Weak-axis
pinned

CUFSM boundary conditions
Pcrl
Pcrl, LH
Pcrl, LH2
Pcrd
Pcre
Pcrl
Pcrd
Pcre
Pcrl
Pcrd
Lcrd
kips
10.5
10.5
10.5
10.6
10.5
10.7
10.5
39.6
40.0
41.0
40.1
39.7
39.9
39.9
39.6
kips
16.2
13.3
12.3
12.5
14.2
12.9
16.5
39.6
41.5
41.4
42.1
39.7
39.9
43.1
39.6
kips
----------16.8
16.5
--43.9
--------45.8
---
kips
18.7
18.7
18.3
19.0
18.6
18.6
18.5
46.0
46.1
49.8
45.0
45.1
49.9
48.8
44.5
kips
8.6
30.0
8.6
8.5
19.9
18.6
19.9
18.5
18.2
50.1
23.1
23.5
52.3
56.5
12.3
kips
10.5
10.5
10.5
10.5
10.5
10.5
10.5
38.9
38.9
39.0
38.9
38.9
39.0
39.0
38.9
kips
17.9
18.9
17.9
17.9
17.7
17.2
17.7
44.6
44.6
46.6
46.1
46.3
47.5
47.5
44.5
kips
6.5
31.6
6.5
6.5
15.5
14.9
15.5
15.4
15.3
45.5
18.9
18.9
45.8
45.8
10.9
kips
10.7
10.7
10.7
10.7
10.7
10.7
10.7
39.6
39.6
39.6
39.6
39.6
39.6
39.6
39.6
kips
17.7
17.7
17.7
17.7
17.7
17.7
17.7
45.5
45.5
45.5
45.5
45.5
45.5
45.5
45.5
in.
13.8
13.8
13.8
13.8
13.8
13.8
13.8
11.3
11.3
11.3
11.3
11.3
11.3
11.3
11.3
84
Test
L hole
Table4.7Parameterrangesforfixedfixedandweakaxispinnedcolumnspecimenswithholes
hhole/h
Lhole/L
F y (ksi)
Specimen type
H/B
D/B
D/t
H/t
B/t
L/H
min
6.3 46.3 19.3
1.9
0.23
1.7
0.16
0.04
24.8
fixed-fixed columns
max 20.0 172.7 65.0
4.9
0.32 13.3
0.60
0.46
62.0
weak-axis pinned
min
6.6 46.2 20.4
2.3
0.33
7.7
0.16
0.01
42.3
columns
max 10.3 71.6 31.7
2.3
0.30 17.9
0.47
0.06
51.5
Table4.8DSMprequalificationlimitsforCsections
Column parameter
4.2.6.2
DSM
prequalification limit
Web slenderness
H/t<472
Flange slenderness
B/t<159
Lip slenderness
4<D/t<33
Web / flange
0.7<H/B<5.0
Lip / flange
Yield stress
0.05<D/B<0.41
Fy<86 ksi.
Boundaryconditioninfluenceonelasticbuckling
TheABAQUSresultsinthecolumnelasticbucklingdatabase,inadditiontoserving
as a resource for extending DSM to columns with holes, can also be used to study the
influenceofcolumnboundaryconditionsonelasticbuckling.Considerthefixedfixed
columns in the database with L/H<4 (most are considered stub columns). Figure 4.13
and Figure 4.14 and compare the influence of the experiment fixedfixed boundary
conditionsforthesecolumnsrelativetowarpingfreeboundaryconditions(i.e.CUFSM
style boundary conditions in Figure 4.2) on Pcrd (distortional buckling) and Pcrl (local
buckling).TheexperimentboundaryconditionsareshowntoincreasePcrdforallofthe
column specimens considered while Pcrl remains relatively unchanged, primarily
becausewarpingdeformationsareintimatelytiedtodistortionalbucklingandnotplate
buckling (Schafer and dny 2006). For stub columns, the length of the fundamental
distortionalhalfwaveisoftenshorterthanthelengthofthecolumn,whichresultsinan
85
increase in Pcrd. The restrained warping at the column ends also contributes to the
shorteningofthehalfwaveandanincreaseinPcrd.ThemagnitudeofthisboostinPcrd
decreasesasL/LcrdincreasesasshowninFigure4.13abecausethewavelengthshortening
required to accommodate distortional buckling in the column can be distributed over
multiplehalfwavesascolumnlengthincreases.Figure4.13bconfirmsthisobservation
bydemonstratingthatPcrdincreaseswithincreasingL/H.Hisinverselyproportionalto
Lcrd for a constant flange width B (i.e., a wider column will have a shorter distortional
halfwavelength) and therefore as L/H increases, the distortional halfwavelength
increasesrelativetothecolumnlengthcausinganincreaseinPcrd.
Pcrlincreasesslightlywithincreasingholesizerelativetocolumnsize(bothforhhole/h
andLhole/L)asshowninFigure4.14duetothefixedfixedboundaryconditions.Forlarge
holes relative to member size the local buckling halfwaves form away from the hole
nearthecolumnends(seeSection3.3).Thesehalfwavelengthsareshortenedrelativeto
their fundamental halfwavelengths by the loaded column edges which are also
restrained from rotating (from welding), resulting in a higher Pcrl when compared to
warpingfreeendconditionswithloadededgesfreetorotate.
86
3
Pcrd, ABAQUS experiment/Pcrd,ABAQUS warping free
2.5
1.5
0.5
0.5
1.5
L/Lcrd
2.5
2.5
1.5
0.5
0.5
1.5
2
L/H
2.5
3.5
Figure4.13InfluenceoffixedfixedboundaryconditionsversuswarpingfreeboundaryconditionsonPcrd
forcolumnexperiments(L/H<4)asafunctionof(a)columnlengthtofundamentaldistortionalhalf
wavelengthcalculatedwithCUFSMand(b)columnlengthtomemberlength.
Pcrl,ABAQUS experiment/Pcrl,ABAQUS warping free
2.5
1.5
0.5
0.1
0.2
0.3
0.4
0.5
hhole/h
0.6
0.7
0.8
0.9
2.5
1.5
0.5
0.05
0.1
0.15
0.2
0.25
Lhole/L
0.3
0.35
0.4
0.45
0.5
Figure4.14InfluenceoffixedfixedboundaryconditionsversuswarpingfreeboundaryconditionsonPcrl
forcolumnexperiments(L/H<4)asafunctionof(a)holewidthrelativetocolumnwidthand(b)holelength
relativetocolumnlength
TheweakaxispinnedboundaryconditionshaveaminimalinfluenceonPcrlandPcrd
inFigure4.15whencomparedtothewarpingfreeboundaryconditions.Thesecolumns
arestillwarpingfixedeventhoughtheyarepinned(aplateisweldedtotheendofthe
member preventing warping deformation), but because the columns are all relatively
long compared to the stub columns and hole size is small relative to column size, the
wavelengthshorteningboostinPcrlandPcrdisnotpronounced.
87
3
2.5
1.5
0.5
0.05
0.1
0.15
0.2
0.25
Lhole/L
0.3
0.35
0.4
0.45
2.5
1.5
0.5
0.5
3
L/Lcrd
Figure4.15Influenceofweakaxispinnedboundaryconditionsversuswarpingfreeboundaryconditionson
(a)Pcrlasafunctionofholelengthtocolumnlengthand(b)Pcrdasafunctionofcolumnlengthtomember
length.
4.2.7 Approximate prediction methods for use in design
The ability to approximate local, distortional, and global critical elastic buckling
loads is central to the extension of the Direct Strength Method (DSM) for coldformed
steel structural members with holes. To facilitate the use of DSM for members with
holes, approximate (and conservative) methods for calculating elastic buckling are
developedherewhichcanbeusedinlieuofafullfiniteelementeigenbucklinganalysis.
Elasticbucklingapproximationsusingthefinitestripmethod(e.g.CUFSM)arederived
for local and distortional buckling, and modifications to the classical column stability
equationsareproposedforglobalbuckling.Thesimplifiedmethodsareintendedtobe
generalenoughtoaccommodatetherangeofholeshapes,sizes,andspacingscommon
inindustry.
4.2.7.1
Localbuckling
An approximate method for predicting the local elastic buckling behavior of cold
formedsteelmemberswithholesispresentedinthissection.Thismethodextendsthe
88
assumptionintheelementbasedmethodsinChapter3thatlocalbucklingoccursas
either plate buckling of the entire crosssection or unstiffened strip buckling at the
locationofthehole.Inthisfinitestripapproximatemethod,localbucklingisassumed
tooccursastheminimumofPcrl occurringfromlocalbucklingonthegrosscrosssection
(ascalculatedintheDirectStrengthMethod)andlocalbucklingoftheunstiffenedstrip
adjacent to the hole. The use of the finite strip method allows for a more realistic
predictionofPcrl forunstiffenedstripbucklingbyincludingtheinteractionofthecross
sectionontheunstiffenedstrip(i.e.,theLHmodefortheCsectioninFigure4.5).The
method is presented through three examples considering industry standard cross
sections with holes which are then verified with ABAQUS thin shell finite element
eigenbucklingresults.Thepredictionmethodisalsovalidatedusingthecolumnelastic
bucklingdatabasedevelopedinSection4.2.6.1.
4.2.7.1.1
Predictionmethod
The local critical elastic buckling load Pcrl is calculated for a coldformed steel
memberwithholesas
Pcrl = min(Pcr , Pcrh ) .
(4.1)
Thecalculationofthelocalcriticalelasticbucklingloadonthegrosscrosssection,Pcr,is
performed using standard procedures defined in Appendix 1 of the AISIS100 (AISI
S1002007).PcrhiscalculatedinCUFSMusingthenetcrosssection,whichisrestrainedto
isolate local buckling from distortional buckling by fixing the column crosssection
cornersasshowninFigure4.16.Itisimportanttoavoidfullyrestrainingacrosssection
89
element (i.e., flange or web), since this prevents Poissontype deformations and
artificially stiffens the crosssection. For example, Figure 4.16a restrains the corners in
the zdirection only to prevent distortional buckling while still accommodating
transverse deformation of the flanges. The only time a corner should be fixed in both
the x and z directions is when two isolated elements intersect (i.e., Csection with a
flange hole, see Figure 4.16a). Finally, when a hole isolates a strip of the net cross
section as shown in Figure 4.16b (e.g., a hat section with flange holes), the isolated
portion of the crosssection should be deleted since it is assumed to no longer
contributestothestiffnessofthecrosssectionoverthelengthofthehole.Iftheisolated
elements are not removed then the critical elastic buckling load calculated in CUFSM
willcorrespondtoEulerbucklingofthisisolatedportionofthecrosssection.
Z
Pinned (typ.)
Roller (typ.)
Web hole
Flange hole
a
b
Web hole
Remove isolated
elements
Flange hole
Figure4.16RulesformodelingacolumnnetcrosssectioninCUFSM
90

Oncethenetcrosssectionisrestrained,aneigenbucklinganalysisisperformed,andan
elasticbucklingcurvesimilartoFigure4.17isgenerated.Lcrhisidentifiedonthecurveas
the halfwavelength corresponding to the minimum buckling load. When Lhole<Lcrh as
shown in Figure 4.17a, Pcrh is equal to the buckling load at the length of the hole. If
LholeLcrh asshowninFigure4.17b,Pcrhistheminimumonthebucklingcurve.Whenno
localminimumexists,thenPcrhisequaltotheelasticbucklingloadcorrespondingtoLhole.
Determining elastic buckling loads at specific halfwavelengths is a new and
fundamentally different use of the finite strip method when compared to its primary
application within DSM, which is calculating the lowest fundamental elastic buckling
modesofcoldformedsteelmembers.
20
20
Lcrh
18
18
16
14
14
Lhole<Lcrh
12
10
10
4
2
Lcrh
12
Pcr
PcrPcr
, kips
Lhole>Lcrh
16
Pcrh
0
0
10
Pcrh
0
1
0
10 10
half-wavelength inches
half-wavelength,
10
half-wavelength inches
half-wavelength,
Figure4.17Localelasticbucklingcurveofnetcrosssectionwhen(a)holelengthislessthanLcrhand(b)
whenholelengthisgreaterthanLcrh
4.2.7.1.2
Methodexamples
Threeexamplesarepresentedherethatapproximatethelocalcriticalelasticbuckling
loadPcrlforcoldformedsteelcolumnswithholesusingCUFSM.Forallexamples,the
91
lengthofthecolumnL=100inchesandfiveslottedholesarespacedatS=20inches.The
typical length of the hole Lhole=4 inches. All ABAQUS eigenbuckling analyses are
modeledwithCUFSMstyleboundaryandloadingconditionsidenticaltothoseshown
inFigure4.2.Themodulusofelasticity,E,isassumedas29500ksiandPoissonsratio,
,as0.3inallfinitestripandfiniteelementmodels.Pcrlisnormalizedwhenplottedby
Py,g, the squash load of the column calculated with the gross crosssectional area and a
yieldstress,Fy,of50ksi.
The first example is an SSMA 362S16233 cross section with a slotted web hole.
Figure4.18comparesthefinitestripandABAQUSmodeshapesforhhole/hC=0.14wherehC
istheCsectionwebdepthmeasuredfromthecenterlineflangetocenterlineflange.The
CUFSMapproximatemethodpredictionsareplottedforarangeofhhole/hCinFigure4.19,
andcomparedwithABAQUSeigenbucklingpredictionstodemonstratetheviabilityof
thepredictionmethod.Forthisexample,smallerholewidthsleadtoreductionsinPcrl
when compared to a member without a hole or members with larger holes. This
counterintuitive result occurs because for small holes the unstiffened strip controls the
local buckling behavior (i.e., the LH mode) and for large holes, local plate buckling
occursbetweentheholes(i.e.,theLmode),whichisconsistentwiththeelasticbuckling
observationsforplates(seeChapter3).(Onemustkeepinmindthatforstrengththenet
section in yielding, as well as the elastic buckling load, ultimately determine the
capacity,notjustPcrl.)
92
Lcrh
hhole
hC
CUFSM Approximation
(SSMA 362S162-33)
ABAQUS
Figure4.18ComparisonofCUFSMandABAQUSpredictionsofunstiffenedstripbuckling.
0.5
ABAQUS
CUFSM Approx. Method
0.45
0.4
0.35
Pcrl /P y,g
0.3
0.25
0.2
0.15
0.1
0.05
0
0.1
0.2
0.3
0.4
0.5
0.6
hhole/hC
0.7
0.8
0.9
Figure4.19ABAQUSresultsverifyCUFSMlocalbucklingpredictionsforanSSMA362S16233column
withevenlyspacedwebholes.
ThenextexampleevaluatestheinfluenceofaslottedflangeholeonPcrlforanSSMA
362S16233 cross section. The unstiffened strip buckled mode shape for this cross
sectionfrombothfinitestripandfiniteelementpredictionsarecomparedinFigure4.20.
It is observed that for both CUFSM and ABAQUS mode shapes, buckling occurs
primarilyinthewebandflangestrip,andthattheflangestriplipportionofthecross
93
section remains stable at Pcrh. The CUFSM prediction method results are plotted for
varying flange hole width bhole relative to centerline flange width bC and compared to
ABAQUSeigenbucklingpredictionsinFigure4.21.Pcrhdecreaseswithincreasingflange
holewidthforbothCUFSMandABAQUSresults.Thedecreasingtrendinthecritical
elastic buckling load demonstrates the importance of the flange in web local buckling
dominatedcrosssections.
Assume Lcrl=Lholein CUFSM

bhole
Lcrl
bC
CUFSM Approximation
(362S162-33)
ABAQUS
Figure4.20CUFSMandABAQUSlocalbucklingmodeshapesareconsistentwhenconsideringaslotted
flangehole.
94
0.5
ABAQUS
0.45
CUFSM Approx. Method, Lcrl=Lhole
0.4
0.35
Pcrl /Py,g
0.3
0.25
0.2
0.15
0.1
0.05
0
0.1
0.2
0.3
0.4
0.5
0.6
bhole/bC
0.7
0.8
0.9
Figure4.21ABAQUSresultsverifyCUFSMpredictionsforanSSMA362S16233crosssectionwithevenly
spacedflangeholes.
ThethirdexampleisanSSMA1200S16268crosssectionwithaslottedholecentered
in the web. Figure 4.22 provides the CUFSM and ABAQUS buckled shapes when
hhole/hC=0.16.TheassumptionintheCUFSMpredictionmethodthatLcrlisequaltoLhole=4
in.producesaPcrhhigherthanPcrwithoutthehole(becauseLcrlisshorterthanthelocal
buckling halfwavelength of the column) and therefore Pcr controls in the prediction
method as shown in Figure 4.23. The approximate method correctly predicts that
unstiffened strip buckling does not occur as observed in the ABAQUS buckled shape,
and that the actual local buckling halfwavelength Lcrl is similar to that of a column
without holes. The prediction for Pcrl is unconservative here though (ABAQUS results
are10%lowerthanPcr),becausetheholecausesamixedlocaldistortionalmodethatis
not captured by the CUFSM netsection model (with pinned corners) or the CUFSM
grosscrosssectionmodel(withouttheinfluenceofthewebhole). Forsectionssuchas
95
this where local and distortional buckling have similar halfwavelengths and critical
elasticbucklingloads,afullfiniteelementeigenbucklinganalysismaybewarrantedto
evaluatethepresenceofholes.
Distortional buckling mixes

with local buckling
hhole
ABAQUS
CUFSM Approximation
(SSMA 1200S162-68)
Figure4.22ABAQUSpredictslocalplatebucklingwithdistortionalbucklinginteractionwhichisnot
detectedinCUFSM.
0.5
ABAQUS
0.45
CUFSM Approx. Method Lcrl=Lhole
0.4
Pcrl /Py,g
0.35
0.3
Pcrh from CUFSM is always

higher than Pcr for this
cross-section
0.25
0.2
0.15
0.1
0.05
0
Pcr
0.1
0.2
0.3
0.4
0.5
hhole/hC
Figure4.23ABAQUSresultsareslightlylowerthanCUFSMpredictions,CUFSMpredictscorrectlythat
platelocalbucklingcontrolsoverunstiffenedstripbuckling.
4.2.7.1.3
Methodvalidationusingelasticbucklingdatabase
96
TheelasticbucklingdatabasedevelopedinSection4.2.6.1isutilizedheretoevaluate
theCUFSMapproximatemethodforpredictingPcrl.CustomMatlabcodewaswrittento
calculatePcrhforall78specimensinthedatabase(Mathworks2007).Thecodeperformed
aCUFSManalysisofthenetcrosssection(crosssectioncontainingahole)withpinned
corners (x and zdirections). The predicted Pcrl of each column specimen is the
minimum of Pcrh (unstiffened strip buckling at the net section) and Pcr (Table 4.4 and
Table4.6,CUFSMelasticbucklingresults,nohole).
Figure 4.24 compares Pcrl reflecting the experimental boundary conditions in
ABAQUS(fromTable4.4andTable4.6)relativetoPcrandPcrh.Forallspecimens,Pcr(no
hole,grosscrosssection)islowerthanPcrh(hole,netcrosssection)becausethestripsof
webadjacenttotheholearestifferthanthecrosssectionawayfromtheholes(similarto
the SSMA362S16233 crosssection with hhole/h>0.20, see Figure 4.19). Even for those
column specimens with small holes relative web width, the holes are often circular or
squareandthereforePcrhispredictedhigherthanPcrsincethebucklinghalfwavelength
of the unstiffened strip is assumed equal to the diameter of the hole. This prediction
resultisconsistentwiththeactualbuckledbehaviorofstiffenedelementswithcircular
andsquarecircularholesshowninFigure3.19.
Figure 4.25 compares the ABAQUS experiment Pcrl to the predicted Pcrl and
demonstrates the approximate method is accurate for smaller holes relative to column
sizeandbecomesincreasingconservativeasholesizeincreasesrelativetocolumnsize
(hhole/h and Lhole/L). The prediction becomes conservative because it does not take into
accountthewavelengthstiffeningeffects(discussedinSection3.3.2)whichboostPcrlas
97
theholebecomeslargerelativetothecolumn.Themeanandstandarddeviationofthe
ABAQUStopredictedratioare1.11and0.18respectively,demonstratingtheviabilityof
themethodforthespecimensconsidered.
3
3
Pcrh (CUFSM net section)
1.5
0.5
Pcr (CUFSM gross section)
2.5
Pcrl,ABAQUS/Pcrl,predicted
Pcrh (CUFSM net section)
Pcr (CUFSM gross section)
2.5
1.5
0.5
0.1
0.2
0.3
0.4
0.5
hhole/h
0.6
0.7
0.8
0.9
0.1
0.2
0.3
0.4
0.5
Lhole/L
0.6
0.7
0.8
0.9
2.5
2.5
Figure4.24PredictedPcrh(CUFSM,bucklingofthenetcrosssection)andPcr(CUFSM,bucklingofthegross
crosssection,nohole)arecomparedrelativetotheABAQUSPcrlwithexperimentboundaryconditionsasa
functionof(a)holewidthtoflatwebwidthand(b)holelengthtocolumnlength
1.5
0.5
1.5
0.5
0.1
0.2
0.3
0.4
0.5
hhole/h
0.6
0.7
0.8
0.9
0.1
0.2
0.3
0.4
0.5
Lhole/L
0.6
0.7
0.8
0.9
Figure4.25PredictedPcrl(CUFSMapproximatemethod)iscomparedrelativetotheABAQUSPcrlwith
experimentboundaryconditionsasafunctionof(a)holewidthtoflatwebwidthand(b)holelengthto
columnlength
4.2.7.2
Distortionalbuckling
An approximate method utilizing the finite strip method is developed here for
predictingthedistortionalcriticalelasticbucklingloadPcrdofcoldformedsteelcolumns
withholes.Themethodsimulatesthelossinbendingstiffnessofacrosssectionwithin
98
a distortional buckling halfwave by modifying the crosssection thickness in CUFSM.

Two different approaches to simulating this loss in stiffness are evaluated. The first
approach reduces the member thickness in the regions of the crosssection with holes
basedontheratioofholelengthtodistortionalhalfwavelength.Thesecondapproach
isdevelopedforCsectionswithwebholesandismechanicsbased,wherethethickness
of the entire web is reduced based on an assumed relationship between web bending
stiffness (derived with observations from ABAQUS thin shell elastic FE analyses) and
thebendingstiffnessmatrixtermsofafinitestripelement.Thestepsforimplementing
these methods in CUFSM are described, and an example is provided where the
predictionmethodforPcrdiscomparedtoABAQUSeigenbucklingresultsforanindustry
standardSSMA250S16268crosssectionwithevenlyspacedwebholesalongthelength
of a column. An empirical equation is derived to account for the increase in Pcrd from
warping fixed end conditions and then the viability of the approximate method is
evaluatedagainstPcrdfromthecolumnexperimentdatabaseinSection4.2.6.1.
4.2.7.2.1
Predictionmethod
The prediction method presented here for Pcrd assumes that the change in cross
sectional stiffness within a distortional halfwave caused by the presence of a hole (or
holes) can be simulated by assuming a reduced thickness of the crosssection. The
distortional halfwavelength of the crosssection, Lcrd, without holes is determined first.
TheelasticbucklingcurveiscalculatedusingthegrosssectionofthecolumninCUFSM
and Lcrd is read off of the curve at the location of the distortional local minimum as
99
shown in Figure 4.26 (this elastic buckling curve corresponds to an SSMA 250S16268
crosssection,whereLcrd=12in.).ThepredictionmethodassumesthatLcrddoesnotchange
with the presence of holes. The crosssection is then modified to approximate the
presence of holes within a distortional halfwavelength. Two approaches for this
modificationsteparepresentednextinSection4.2.7.2.1.1andSection4.2.7.2.1.2. Once
the cross section is modified to account for the presence of a hole in CUFSM, another
elastic buckling curve is generated and Pcrd (including the presence of the hole) is
determinedastheelasticbucklingloadoccurringatLcrdasshowninFigure4.26.
100
without hole
with hole
90
80
70
Lcrd (determined at
local minimum of no
hole curve)
P cr , kips
60
50
40
30
20
10
0
0
10
Pcrd (includes
influence of hole)
10
half-wavelength, in.
10
Figure4.26CUFSMapproximatemethodforcalculatingPcrdforacolumnwithholes.
.7.2.1.1 Weighted average approach for predicting hole influence
100
Theholeinfluenceondistortionalbucklingofanopenthinwalledcrosssectioncan
beapproximatedbymodifyingthecrosssectionthicknessinCUFSMatthelocationofa
holewiththefollowingequation:
L
t hole = 1 hole t .
Lcrd
(4.2)
The implementation of the reduced thickness in a Csection with a single web hole is
provided in Figure 4.27. This approach is an intuitive first cut at approximating the
reduction in bending stiffness of the cross section. A more refined mechanicsbased
approachispresentednext.
thole
hhole
Figure4.27ModifiedcrosssectiontobeusedinCUFSMtopredictPcrdforacolumnwithholes.
.7.2.1.2 Mechanics-based approach for predicting hole influence
A plate model is developed in ABAQUS to study the influence of a hole on the
bendingstiffnessofaSSMA250S16268columnexperiencingdistortionalbuckling.The
stiffness reduction observed in ABAQUS is quantified and then equated to finite strip
bending stiffness matrix terms to derive a reduced web thickness expression based on
finite strip mechanics. The plate dimensions in ABAQUS are chosen to correspond to
thewebofthe250S16268sectionoveronedistortionalhalfwave.Theplatewidthhis
101
2.4 in., the plate length L=12 inches (consistent with Lcrd=12 in.), and t=0.0713 in. One
slotted hole with Lhole=4 in. is centered in the plate. The width of the hole is varied,
hhole=0.5 in., 0.96 in., 1.20 in., 1.5 in., and 1.75 in. (and subsequently rhole varies). The
modulusofelasticity,E,isassumedas29500ksiandPoissonsratio,,as0.3forallfinite
element models considered here. The ABAQUS boundary conditions and applied
loading are described in Figure 4.28. The plate is simplysupported and loaded with
imposedrotationsatthelongedgesoftheplatewithmagnitudesvaryingasahalfsine
wavetosimulatedistortionaldeformationoveronehalfwavelength.
Restrain midline node in
1 and 3
1 (x)
Restrain midline node in 3

Apply imposed rotation at
plate edges in the shape of
0.001*sin(x/L) radians
Restrain perimeter in 2
Figure4.28ABAQUSboundaryconditionsandimposedrotationsforwebplate
Thedeformedshapeoftheplatewhenhhole/h=0.50isprovidedinFigure4.29.Ateach
node where an imposed rotation is applied, the associated moment is obtained from
ABAQUS and plotted in Figure 4.30 as a transverse bending stiffness per unit length.
(Notethatnearx=0in.andx=12in.,thedeformedshapeinABAQUSresultsinasmall
negative bending stiffness which is not plotted in Figure 4.30 and does not affect the
overallresultshere.Thenegativestiffnessisnotpredictedinthefinitestripformulation
102
because the longitudinal shape function is enforced as a halfsine wave). The hole
causes a sharp reduction in bending stiffness at the location of the hole, but has a
minimal influence on bending stiffness away from the hole. The stiffness reduction is
shown to be relatively insensitive to the ratio of hole width to plate width except for
peaksinstiffnessthatincreasewithhhole/hattheroundededgesoftheslottedhole.The
results in Figure 4.30 confirm the intuitive assumption employed to develop Eq. (4.2);
the ratio of the length of the hole to the length of the distortional halfwave is an
importantparameterwhenpredictingthelossinbendingstiffness.
Figure4.29Platedeformationfromimposededgerotations,hhole/h=0.50
103
0.07
hhole/h=0.63
rotational stiffness, k*in/rad
0.06
0.05
no hole
0.04
0.03
0.02
Increasing
hhole/h
0.01
4
6
8
distance along plate, in.
10
12
Figure4.30Transverserotationalstiffnessoftheplateissignificantlyreducedinthevicinityoftheslotted
hole
If K represents the cumulative transverse bending stiffness for the plate without a
hole(areaunderthecurveinFigure4.30),thenthereducedKincludingthepresenceof
theholecanbeapproximatedas:
L
K hole = 1 hole K .
Lcrd
(4.3)
TheglobalbendingstiffnessKforasimplysupportedfinitestripelementisderivedby
applyingaunitrotationatthestripedges:
keb [d w ]
k11
k
= 12
k31
k 41
k12
k 22
k13
k 23
k32
k 42
k33
k 43
k14 0 V
k 24 1 K
= ,
k34 0 V
k 44 1 K
(4.4)
where the keb is the bending stiffness matrix and dw=[w1 1 w2 2] (Schafer and dny
2006).SolvingEq.(4.4)forK:
104
K = k 22 k 24 .
(4.5)
Sincek22andk24arebothfunctionsofthewebthickness(tweb)3,KandKholecanbeequated
directlyas:
K hole t web,hole
=
.
3
K
t web
( )
(4.6)
Substituting Eq. (4.6) and rearranging in terms of tweb,hole, the reduced web thickness
correspondingtothereducedtransversebendingstiffnessfromtheholeis:
t web,hole
L
= 1 hole
Lcrd
1/ 3
t web .
(4.7)
Eq. (4.7) is an improvement over Eq. (4.2) because it reflects the underlying plate
bending mechanics involved in distortional buckling and is actually simpler to
implement in CUFSM since the entire web thickness of a Csection is reduced to tweb,hole
insteadofchangingthesheetthicknessjustatthelocationoftheholeasshowninFigure
4.27.Asimilarmodificationtothasbeenproposedforwebslottedthermalstructural
studs(Kesti2000).
4.2.7.2.2
Methodexample
ThedistortionalcriticalelasticbucklingloadPcrdiscalculatedherewiththeCUFSM
predictionmethodforalongcolumn(L=100in.)withanSSMA250S16268crosssection
andfiveevenlyspacedslottedwebholes(S=20in.,Lhole=4in.).Thewidthoftheholeis
varied relative to the web width, and ABAQUS eigenbuckling results are used to
evaluatetheviabilityofthemethod.AllABAQUSfiniteelementmodelshaveCUFSM
style boundary and loading conditions as shown in Figure 4.2. The modulus of
105
elasticity,E,isassumedas29500ksiandPoissonsratio,,as0.3inallfinitestripand
finiteelementmodels.PcrdisnormalizedbyPy,gwhenplotted.Py,gisthesquashloadof
thecolumncalculatedwiththegrosscrosssectionalareaandassumingFy=50ksi.
AcomparisonoftheCUFSMpredictionmethod(employingtheweightedaverage
thicknessapproximation)andABAQUSdistortionalbucklingmodeshapesareprovided
inFigure4.31whenhhole/h=0.63.Ninedistortionalhalfwavesformalongthememberin
ABAQUS, with every other halfwave containing one slotted hole. The CUFSM
prediction method employing both the weighted average and mechanicsbased
thickness modifications to the crosssection are compared over a range of hhole/h to
ABAQUS eigenbuckling results in Figure 4.32. Pcrd for the pure ABAQUS distortional
(D) buckling mode is plotted to demonstrate that prediction method is viable for this
crosssection and hole spacing. The CUFSM prediction for Pcrd with the weighted
averagethicknessreductionattheholedecreaseswithincreasingholewidthsincethe
webprovideslessbendingstiffnesstotheflangesasmoreholematerialisremoved.The
CUFSMpredictionemployingthemechanicsbasedreductioninwebthicknessisnot
afunctionofhhole/handisshowntobeamorerealisticpredictorofPcrdthantheweighted
average approach. These approximate methods are evaluated against ABAQUS
distortionalbucklingpredictionsfromthecolumndatabaseinSection4.2.7.2.4.
106
Lcrd
hhole
CUFSM Approximation
(250S162-68)
ABAQUS
Figure4.31ComparisonofCUFSMandABAQUSdistortionalbucklingmodeshapes.
3
CUFSM Approx. Method with "weighted average" t hole
CUFSM Approx. Method with "mechanics-based" t w eb,hole
2.5
ABAQUS D mode
Pcrd/Py,g
1.5
0.5
0.1
0.2
0.3
0.4
0.5
hhole/h
0.6
0.7
0.8
0.9
Figure4.32CUFSMdistortionalbucklingpredictionmethodisconservativewhenconsideringanSSMA
262S16268columnwithuniformlyspacedholes.
4.2.7.2.3
Warpingfixeddistortionalamplificationfactor
Longitudinalwarpingdeformationsoccurasaresultofdistortionalbucklingincold
formed steel columns. When this warping deformation is restrained, the distortional
107
buckling halfwavelength is shortened relative to the fundamental distortional half

wavelengthofthecolumncrosssection,Lcrd.Thischangeinhalfwavelengthresultsin
anamplificationofthedistortionalcriticalelasticbuckling loadofthecolumn,Pcrd(see
Figure4.13aforboostinPcrdforstubcolumns).Theelasticbucklingdatabasedeveloped
in Section 4.2.6.1 (Table 4.4 and Table 4.6) provides an opportunity to derive an
empiricalamplificationfactorsinceforallcolumnsinthedatabase,Pcrdforbothwarping
fixed (ABAQUS) and warping free (CUFSM) boundary conditions are known and the
fundamentaldistortionalhalfwavelength,Lcrd,hasbeencalculatedinCUFSM.
ThewarpingfixedboundaryconditioneffectonPcrdisplottedforthe78specimensin
thecolumndatabaseinFigure4.33.TheboostinPcrdishighestwhenthecolumnisshort
relativetoLcrdbecausethewavelengthshorteningmustbeaccommodatedoverjustone
halfwave.Anempiricalequation(alsoplottedinFigure4.33)isfittothedatatrend:
2
D boost
1L
= 1 + crd
2 L
(4.8)
ThisamplificationfactorcanbeusedwiththeCUFSMpredictionmethoddevelopedin
Section 4.2.7.2.1 when the column being considered has warpingfixed boundary
conditions.Eq.(4.8)isconsistentwiththedistortionalbucklingboostfactorprovidedin
theDSMDesignGuide(AISI2006)asshowninFigure4.33.
108
Pcrd,fixed-fixed /Pcrd,CUFSM warping free
3
Specimen data
Eq.(4.8)
DSM Design Guide
2.5
1.5
0.5
0.5
1.5
2.5
L/Lcrd
3.5
4.5
Figure4.33WarpingfixedboundaryconditionamplificationofPcrd
4.2.7.2.4
VerificationofCUFSMapproximatemethodwithcolumndatabase
The CUFSM approximate method for distortional buckling is now evaluated using
theelasticbucklingpropertiesofthe78columnspecimensfromtheexperimentdatabase
developed in 4.2.6.1. The ABAQUS distortional critical elastic buckling load Pcrd,
determined with the experiment boundary conditions, is plotted against the
approximate method predictions in Figure 4.34. The predictions including the
distortionalamplificationfactorfromEq.(4.8).Theapproximatemethodemployingthe
weighted average reduced web thickness at the hole from Eq. (4.2) and the
mechanicsbasedreducedwebthicknessapproachfromEq.(4.7)arebothpresented.
Theaccuracyofthepredictionmethodimprovesasthecolumnlengthincreasesrelative
tothefundamentaldistortionalhalfwavelengthLcrd.Thepredictionaccuracyishighly
109
variable when L/Lcrd<1, primarily because of the variation in the boundary condition
influence described in Section 4.2.7.2.3 for stocky columns. As expected, the
mechanicsbased thickness approach (with ABAQUS to predicted ratio mean and
standard deviation of 1.19 and 0.29) is more accurate over the 78 columns than the
weightedaverageapproach(meanof1.24andstandarddeviationof0.29).
3
weighted average t hole
mechanics-based t w eb,hole
Pcrd,ABAQUS Exp. BC/Pcrd,predicted
2.5
1.5
0.5
3
L/Lcrd
Figure4.34AccuracyoftheCUFSMapproximatemethodforpredictingPcrdimprovesascolumnlength
increasesrelativetothefundamentaldistortionalhalfwavelengthforwarpingfixedcolumns
4.2.7.3
Globalbuckling
Twodifferentapproximatemethodsforcalculatingthecriticalelasticbucklingload
ofacolumn,Pcre,areevaluatedinthissection.Bothmethodsemployweightedaverages
ofthemembersectionpropertiesintheclassicalcolumnstabilityequationtoaccountfor
theinfluenceofholes,oneusingaweightedcrosssectionalthicknessatthelocationsof
theholeandtheotherusingtheweightedaverageofthegrossandnetcrosssections.The
110
approximatemethodsarecomparedtoABAQUSeigenbucklingresultsforalongcold
formed steel column (SSMA 1200S16268 cross section) with uniformly spaced circular
holes.Theaveragetorsionalconstants,JandCw,arecalculateddirectlyusingABAQUS
for this column and then compared to their associated weighted average estimates.
Basedonthesestudies,recommendationsaremaderegardingtheapproximatemethod
mostsuitableforpredictingPcreforcolumnswithholes.
4.2.7.3.1
Descriptionofmethods
.7.3.1.1 Weighted Properties Method
Theequationforpredictingtheglobal(flexuralonly)criticalelasticbucklingloadPcre
of a column with holes along its length can be solved using energy methods, and is
derived for a column with two holes located symmetrically about the longitudinal
midlineofthecolumninAppendixE.TheequationthatevolvesfromtheRaleighRitz
derivationforthiscaseis:
Pcre =
2 E I g L NH + I net LH
L2
(4.9)
whereIgisthemomentofinertiaofthegrosscrosssection,Inetisthemomentofinertiaof
thenetcrosssection,LNHisthelengthofcolumnwithoutholesandLHisthelengthof
columnwithholes(notethatLNH+LH=L).Theaveragemomentofinertiaofthecolumn
with holes is shown in Eq. (4.9) to be equivalent to the weighted average of the gross
andnetcrosssectionsalongthecolumnlength.Anapproximatemethodforcalculating
PcreisproposedherewhichextendsthisweightedpropertiesmethodologyinEq.(4.9)
111
toallofthecrosssectionpropertiesofthecolumnrequiredtosolvetheclassical cubic
bucklingequationforcolumns(Chajes1974):
(P
cre , y
P )(Pcre, x P )(Pcre, P ) (Pcre, y
P 2 xo2
P 2 y o2
P ) 2 (Pcre, x P ) 2 = 0
ro
ro
(4.10)
including the crosssectional area A, moment of inertia Ix and Iy, St. Venant torsional
constant J, and shear center location. The computer program CUTWP solves Eq. (4.10)
for any general crosssection and is freely available (Sarawit 2006). The net section
properties can be calculated in CUFSM (or CUTWP) by reducing the sheet strip
thickness to zero at the location of the hole. This approximation is referred to as the
weightedpropertiesmethod.AgeneralformofEq.(4.9)isalsoderivedinAppendix
E which can be used with the weighted properties method for the case of a column
withasingleholeormultiplearbitrarilyspacedholes.
.7.3.1.2 Weighted Thickness Method
Thisapproximatemethodapproachesthecalculationoftheaveragecolumnsection
propertiesinadifferentway,byusingaweightedmemberthicknessatthelocationof
theholesinthecrosssectiontocalculatetheaveragecrosssectionalproperties:
t hole =
L LH
t .
L
(4.11)
An example of a Csection with a reduced thickness at the location of a web hole is

provided in Figure 4.35. This weighted thickness method is more convenient to
implement than the weighted properties method presented in the previous section
112
becausethemodifiedcrosssection(withreducedthickness)canbeinputdirectlyintoa
computerprogramsuchasCUTWP.
hhole
thole
Figure4.35Aweightedthicknesscrosssectioncanbeinputdirectlyintoaprogramthatsolvesthe
classicalcubicstabilityequationforcolumns(e.g.CUTWP).
4.2.7.3.2
Exampleandverification
ABAQUSglobaleigenbucklingresultsarecomparedinthissectiontotheweighted
properties and weighted thickness prediction methods for an industry standard

SSMA 1200S16268 long column (SSMA 2001) with evenly spaced circular holes. The
lengthofthecolumnL=100in.,theholespacingS=20in.,andthediameterofthecircular
holeisvariedfromhhole/H=0.10tohhole/H=0.90whereHistheouttooutdepthofthecross
section (see Figure 4.1 for crosssection dimension notation). All ABAQUS finite
elementmodelsareloadedincompressionatthememberendsandhavewarpingfree
boundaryconditionsasshowninFigure4.2.Themodulusofelasticity,E,isassumedas
29500ksiandPoissonsratio,,as0.3inallCUTWPandfiniteelementmodels.
The three global buckling modes of this SSMA 1200S16268 long column without
holesarecalculatedinCUTWP:(1)weakaxisflexuralbucklingoccursatPcr=7.91kips,
(2) flexuraltorsional buckling occurs at Pcr=13.39 kips, and (3) strongaxis flexural
113
buckling occurs at Pcr=604.17 kips. The first two buckling modes are the focus of this
study since the strongaxis buckling mode is much higher than the squash load of the
column (Py,g= 56.30 kips assuming Fy=50 ksi) and will not influence the design of the
column. Figure 4.36 provides an example of the weakaxis flexural and flexural
torsional buckling modes when hhole/H=0.50. Note that shell FE predicts local buckling
mixingwiththeweakaxisflexuralmodewhenhhole/H>0.50becausePcreisreducedbythe
presence of holes to a magnitude similar to the local critical elastic buckling load
(Pcrl=6.69kips). Local buckling is not observed to mix with global buckling in the
flexuraltorsional(column)orlateraltorsional(beam)bucklinginthisstudy.
Web local buckling
mixes with global mode
Weak Axis Flexural

Pcre=6.96 kips
Flexural-Torsional
Pcre=10.64 kips
Global column buckling modes

(SSMA 1200S162-68, hhole/h=0.50)
Figure4.36WeakaxisflexuralandflexuraltorsionalglobalbucklingmodesforanSSMA1200S16268
columnwithevenlyspacedcircularholes.
.7.3.2.1 Section property calculations at the net section
To draw meaningful conclusions regarding the weighted properties and
weighted thickness prediction methods it is first helpful to understand how circular

holediameterinfluencesthecolumnssectionproperties.Figure4.37comparesthenet
sectiontothegrosscrosssectionalareaA,thestrongandweakaxismomentofinertiaIx
andIy,theSt.VenanttorsionalconstantJ,andthewarpingtorsionalconstantCwofthe
114
columnastheyvarywithhhole/H.Allnetsectionpropertiesinthisfigurearedetermined
withtheCUFSMsectionpropertycalculatorbyreducingthesheetthicknesstozeroat
the location of the hole. A and J decrease linearly with hole diameter while Ix and Iy
decreasenonlinearly.Iyismostsensitivetothepresenceoftheholebecausetheholeis
locatedinthechannelwebforthiscase,whichprovidesmuchofthecontributiontothe
weakaxismomentofinertia.Cw,calculatedhereassumingzerothicknessbutcontinuity
through the hole, varies nonlinearly with hhole/H. It is unclear if the net section Cw
calculatedinthiswayproducesthebestapproximationofthecolumnsactualphysical
behavior in torsional buckling. The magnitude of Cw is influenced heavily by cross
sectioncontinuitysince alineintegralaroundthecrosssectionisusedtosolveforthe
warpingfunction.FurtherinvestigationofJandCwforcolumnswithholesispresented
inSection4.2.7.3.2.3.
1.5
Cw
Ix (strong)
net section/gross section
Iy (weak)
A
J
0.5
0.1
0.2
0.3
0.4
0.5
hhole/H
0.6
0.7
0.8
0.9
Figure4.37Variationinnetsectionpropertiesascircularholediameterincreases.
115
.7.3.2.2 Average section property calculations for the column - A, Ix, and Iy
The average section properties of the 1200S16268 column with circular holes
calculated using the weighted thickness and weighted properties methods are
compared in Figure 4.38 through Figure 4.40. For this example problem there are
minimal differences between the methods for A and Ix, although Iy calculated with the
weightedpropertiesmethoddecreasesinstiffnessrelativetotheweightedthickness
methodasholediameterincreasesrelativetocolumnwidth.
1.5
weighted thickness
weighted properties
Aavg/Ag
0.5
0.1
0.2
0.3
0.4
0.5
hhole/H
0.6
0.7
0.8
0.9
Figure4.38Comparisonofweightedthicknessandweightedpropertiescrosssectionalarea.
116
1.5
Ix,avg/Ix,g (strong axis)
weighted thickness
weighted properties
0.5
0.1
0.2
0.3
0.4
0.5
hhole/H
0.6
0.7
0.8
0.9
Figure4.39Comparisonofweightedthicknessandweightedpropertiesstrongaxismomentofinertia.
1.5
Iy,avg/Iy,g (weak axis)
weighted thickness
weighted properties
0.5
0.1
0.2
0.3
0.4
0.5
hhole/H
0.6
0.7
0.8
0.9
Figure4.40Comparisonofweightedthicknessandweightedpropertiesweakaxismomentofinertia.
.7.3.2.3 Average section property calculations for the column - J and Cw
The average J and Cw of the 1200S16268 column with circular holes is determined
directly using ABAQUS and then compared to the weighted properties and
117
weightedthicknesspredictionsinthissection.Thedifferentialequationfortorsionis
definedas(Timoshenko1961):
d
d 3
T = GJ
ECw 3 ,
dx
dx
(4.12)
where is the angle of twist of the cross section and G is the shear modulus of steel
(G=11346ksiinthiscase).Eq.(4.12)isusedinconjunctionwithstaticABAQUSanalyses
(noteigenbuckling!)tosolveforJavgandCw,avgofthecolumnashhole/Hvariesfrom0.10to
0.90.Javg iscalculatedbyapplyingaunittwistattheendofthecolumnaboutthegross
crosssectionshearcenterwhilekeepingtheoppositecolumnendfixedagainsttwistas
showninFigure4.41.Ifbothendsofthecolumnarefreetowarp,thevariationintwist
alongthecolumnis constantasshowninFigure4.42andwarpingresistancedoesnot
contributetotheresultingtorsion(d3/dx3=0).Thevariationintwistwasnotsensitiveto
increasingholediameterinthiscase,andthereforethelineshowninFigure4.42isthe
same regardless of hole diameter. The twist along the column is measured in
ABAQUSastherelativerotationoftheflangewebcorners.Thetwistmagnitudealong
the column length remains unchanged with increasing hhole/H. Javg for the warping free
columniscalculatedbyrearrangingEq.(4.12):
J avg =
To L
G o
(4.13)
o,G,andLareknownandToisthetorqueresultingfromtheunitrotationo,whichis
readdirectlyfromABAQUS.
118
shear center reference

node (determined with
gross cross section)
Cross-section kinematically restrained to shear

center reference node in 2, 3 (warping free)
SECTION A-A
1 (x)
Cross-section dof
fixed in 2 and 3
(warping free)
Node at centerline
of web fixed in 1 to
prevent rigid body
motion
Figure4.41ABAQUSboundaryconditionsforwarpingfreeandappliedunittwistatx=0in.andwarping
freebutrotationrestrainedatx=100in.
1.4
/ 0, normalized angle of twist
1.2
0.8
0.6
0.4
0.2
10
20
30
40
50
60
70
distance along column, in.
80
90
100
Figure4.42AngleoftwistdecreaseslinearlyintheSSMA1200S16268columnwithwarpingfreeend
conditions.
The resulting Javg from ABAQUS is compared against the weighted properties and
weighted thickness calculations of Javg. (Note that the weighted properties Javg is
calculated with Jnet from Figure 4.37 using the CUFSM section property calculator and
119
assuming the thickness is zero at the hole). It is clear from Figure 4.43 that the
weighted properties calculation of Javg is most consistent with Javg derived from
ABAQUS.
1.5
weighted thickness
weighted properties
ABAQUS
Javg/Jg
0.5
0.1
0.2
0.3
0.4
0.5
hhole/H
0.6
0.7
0.8
0.9
Figure4.43TheweightedpropertiesapproximationforJavgmatchescloselywiththeABAQUSprediction
fortheSSMA12S0016268columnwithholes
The ABAQUS boundary conditions are now modified such that warping is
restrained at the fixed column end as shown in Figure 4.44. A unit twist, o, is again
appliedatthefreeend,andtheresultingangleoftwistalongthelengthofthecolumn
is measured in ABAQUS. Because of the warpingfixed end condition, is nonlinear
along the length of the column and warping resistance contributes to the torsional
stiffnessofthecolumn.Sincethedistributionofalongthecolumnisnotinfluencedby
hhole/H as observed in ABAQUS, an indirect solution of Cw,avg as a function of Cw,g can be
derived:
120
C w,avg
C w, g
d
(x = 0)
dx
,
d
(x = 0)
GJ g
dx
To GJ avg
To , g
(4.14)
where for each ABAQUS model (hhole/H=0.10 to 0.90), the torque To associated with the
unittwistoisreaddirectlyfromABAQUSandd/dx(x=0)iscalculatedfromFigure4.45.
AswasthecaseforthewarpingfreecaseinFigure4.42,thevariationintwistwasnot
sensitivetoincreasingholediameterandthereforethelineshowninFigure4.45isthe
same regardless of hole diameter. The influence of holes on the variation in twist is
expected to be more pronounced as column length decreases relative to hole length.
Futureresearchisplannedtoevaluatetheinfluenceofmemberlengthonthetorsional
propertiesofcolumnswithholes.
Cross-section kinematically restrained to shear

center reference node in 2, 3 (warping free)
shear center reference
node (determined with
gross cross section)
SECTION A-A
1 (x)
Cross-section dof
fixed in 1,2, and 3
(warping fixed)
Figure4.44ABAQUSboundaryconditionsforwarpingfreeandappliedunittwistatx=0in.andwarping
fixedandrotationrestrainedatx=100in.
121
1.4
/0, normalized angle of twist
1.2
0.8
0.6
0.4
0.2
10
20
30
40
50
60
70
distance along column, in.
80
90
100
Figure4.45AngleoftwistisnonlinearalongtheSSMA1200S16268columnwithwarpingfixedend
conditionsatx=100in.
Figure 4.46 demonstrates that the weighted properties and weighted thickness
approximations both overestimate Cw,avg when compared to the ABAQUS derived Cw,avg
demonstrating that neither is an accurate predictor of Cw,avg. A modified version of the
weightedpropertiesapproximationisalsoplotted,whereCw,netistakenequaltozero
instead of Cw,net taken from the results in Figure 4.37. This assumption for Cw,net is
motivated by the idea that the hole separates the cross section into two pieces, where
each piece on its own contributes minimally to warping resistance. This modified
weighted properties approximation results in a conservative lower bound on Cw,avg
which is useful from a design perspective until more accurate approximations are
developed.
122
1.5
weighted thickness
weighted properties
derived from ABAQUS
weighted properties Cw ,net=0
Cw,avg /Cw,g
0.5
0.1
0.2
0.3
0.4
0.5
hhole/H
0.6
0.7
0.8
0.9
Figure4.46Comparisonofweightedthicknessandweightedpropertiesapproximationstothe
ABAQUSderivedwarpingtorsionconstantCw,avg.
4.2.7.3.2.4
Comparison of prediction accuracy between methods
Figure 4.47 compares the weakaxis flexural critical elastic buckling load of the
1200S16268 column calculated with the weighted thickness and weighted

properties prediction methods to ABAQUS eigenbuckling results. The ABAQUS
calculation of Pcre is systematically 10% lower than the prediction method (even for a
columnwithoutholes),whichresultsfromtheassumptionofarigidcrosssectioninthe
classical stability equations. The column crosssection as modeled in ABAQUS is
allowed to change shape along the length, resulting in a lower axial stiffness of the
column. (The reduction in Pcre was also confirmed in CUFSM, which like ABAQUS,
accounts for platetype deformations in elastic buckling calculations.) Beyond this
systematic difference, both approximate methods are accurate predictors of Pcre for
123
hhole/h0.60andtheweightedpropertiesmethodremainsaccurateforevenlargerholes.
ThepredictionofweakaxisflexurePcreusingthenetsectionpropertiesfromFigure4.37
isalsoplottedinFigure4.47,demonstratingaconservativealternativetotheweighted
propertiesandweightedthicknessmethods.
1.5
Pcre,ABAQUS/Pcre,prediction
weighted thickness
weighted properties
net section
1
Difference is caused by classical rigid

cross-section assumption which
increases axial column stiffness
when compared to ABAQUS
0.5
0.1
0.2
0.3
0.4
0.5
hhole/H
0.6
0.7
0.8
0.9
Figure4.47Comparisonofweightedthicknessandweightedpropertiespredictionmethodsforthe
SSMA1200S16268weakaxisflexuralbucklingmode.Predictionsusingnetsectionpropertiesarealso
plottedasaconservativebenchmark.
Figure4.48comparestheweightedthicknessandweightedpropertiesmethods
toABAQUSresultsforthesecondglobalmode,flexuraltorsionalcolumnbuckling.The
accuracy of the prediction methods decrease with increase hhole/H for both methods,
confirmingwhatwasobservedinFigure4.46,thattheweightedapproximationsforCw
are not accurate representations of the average warping torsion stiffness, especially as
hhole/Hbecomeslarge.WarpingtorsiondominatesoverSt.Venanttorsioninthismode
since both weighted average methods predict similar variations in Pcre, even though J
variesbetweenthetwomethods(seeFigure4.43).Theweightedpropertiesmethod
124
withCw,avgreplacedwithCw,avgpredictedinABAQUS(seeFigure4.46)accuratelypredicts
Pcre until hhole/H exceeds 0.80, although this method may not be practical from a design
perspective since it requires thin shell FE analysis. The modified weight properties
approach,calculatedassumingCw,net=0,isshowntobemoreaccuratethanusingjustthe
net section properties and is a conservative method for predicting Pcre of flexural
torsional buckling modes. Future research is planned to develop a mechanicsbased
approximationfortheaverageCwofacolumnincludingtheinfluenceofholes.
Pcre,ABAQUS/Pcre,prediction
1.5
weighted thickness
weighted properties
net section
weighted properties, ABAQUS Cw ,avg
weighted properties, Cw ,net=0
0.5
0.1
0.2
0.3
0.4
0.5
hhole/H
0.6
0.7
0.8
0.9
SSMA1200S16268flexuraltorsionalcolumnbucklingmode.Predictionsusingnetsectionpropertiesare
alsoplottedasaconservativebenchmark.
4.3 Elasticbucklingofbeamswithholes
4.3.1 Analysis of existing tests on beams
A column experiments database was assembled in Section 4.2.6.1 to serve as a
resourceinthedevelopmentandvalidationoftheDirectStrengthMethodforcolumns
125
withholes.Inthissection,asimilardatabaseisdevelopedthatsummarizestheelastic
buckling properties and tested strengths of coldformed steel beam experiments with
holes. This database is used in Chapter 8 when developing and verifying DSM for
beamswithholes.
The beam experiments considered in this study were conducted by Shan,
LaBoube, Schuster, and Batson in the early 1990s and consist of three separate test
sequences(Batson1992;Schuster1992;ShanandLaBoube1994).TestSequences1and2
were performed at the University of MissouriRolla (UMR) and Test Sequence 3 was
executed at the University of Waterloo. Each specimen is made up of two Csections
orientedtoetotoeasdepictedinFigure4.49.3/4x3/4x1/8aluminumanglesconnect
thetopandbottomflangesofthetwochannelswithoneselfdrillingscrewperflange.
Theanglesprovideaclosedbeamsectionthatpreventslateraltorsionalbucklingofthe
individualchannels.
4.3.1.1
Memberandholedimensionnotation
BeamcrosssectionandholedimensionnotationispresentedinFigure4.50.The
Csectioninsidecornerradiiareassumedtoequaltwicethemeasuredthicknessofthe
specimen.Twoholeshapeswereconsideredintheexperiments,anindustrystandard
slottedholeandatrislottedholewiththecurvedholeendsreplacedbytriangulartips.
Theholesarecenteredinthewebandaremechanicallypunchedat24inchesoncenter
longitudinallywithaholeatthecenterofthespan.
126
3/4x3/4x1/8 angle(top and bottom)
Cee channel (typ.)
self-drilling screw (typ.)
Figure4.49CrosssectionofbeamspecimenshowingaluminumstrapanglesconnectedtoCflanges
6 in.
B11
B21
D1
Channel 1
H1
Lhole
D21
rhole
Channel 2
hhole
H2
Slotted hole
hhole/2
t
D2
B12
D22
Tri-slotted hole
B22
Figure4.50Csectionandholedimensionnotation
4.3.1.2
Testedboundaryconditionsandloading
All beam specimens are tested as simply supported in fourpoint bending to

create a region of constant momentbetween the load points at the center of the beam.
Thepointloadsareappliedthroughstubchannelsattachedtothebeamwebswithself
drilling screws. The stub channels prevent web crippling by distributing the
concentratedloadandbyrestrainingtheweb.Lateralbracingisprovidedinthevicinity
oftheconstantmomentregionbystrutsconnectingthetopflangealuminumanglestoa
127
reactionframe.Theendsofthebeamspecimensarelaterallybracedbyverticalrollers.
AsummaryofthebeamtestsetupisprovidedinFigure4.51.
128

L
6-0
X
P
*
lateral brace point
at angle (typ.)
X
P
a
24 hole spacing (typ.)
12 spacing (typ.) Test Sequence 1, 3
6 spacing (typ.) Test Sequence 2
3/4x3/4x1/8 angle connected with one

screw to each flange (top and bottom)
3/4x3/4x1/8 angle
Channel 2
Channel 1
Section a-a
Figure4.51Experimenttestsetupwithholespacing,locationoflateralbracing,spacingofaluminumanglestraps,andloadpoints
129
4.3.1.3
FiniteElementModeling
The elastic buckling properties of the 72 beam specimens are obtained with
eigenbucklinganalysesinABAQUS(ABAQUS2004).AllbeamsaremodeledwithS9R5
reduced integration ninenode thin shell elements. Refer to Section 2.1 for a detailed
discussion of the S9R5 element. Coldformed steel material properties are assumed as
E=29500ksiand=0.3.
Special care is taken to simulate the experimental boundary conditions when
modeling in ABAQUS. The simple supports with vertical roller restraints, the
aluminum angle straps connecting the top and bottom channel flanges, the lateral
bracing of the top flange in the constant moment region, and the application of load
through the webs and are all considered. Figure 4.52 summarizes the ABAQUS
boundaryconditionassumptions.
Beam end restrained in 2
and 3 (v, w=0)
*
Bottom flange restrained in
1 at support (u=0)
Restrain node at midline of

top flange in 3 (w=0) (Typ.)
Rigid body connection

between top (and bottom)
flange midline nodes (Typ.)
Beam end restrained in 2

and 3 (v, w=0)
Figure4.52Finiteelementmodelboundaryconditionsforbeameigenbucklinganalyses
Tosimulatethesimplysupportedconditionswithverticalrollers,theendsofthe
beams are modeled as warping free except for the bottom flange at one end which is
130
restrainedtopreventlongitudinalrigidbodymovement.Thechannelcrosssectionsare
restrainedfromverticalandlateraltranslationatbothbeamends.
A rigid body restraint is used to model the connectivity between the top and
bottom Csection flanges provided by the aluminum angle straps connected with self
drilling screws. Figure 4.53 demonstrates how each angle is modeled as a rigid body
made up of one midline flange node from each channel. The rigid body definition
requiresthatthemotion(bothtranslationalandrotational)ofthetwonodesisgoverned
by a single reference node, in this case the midline flange node of Channel 1. The
formulation allows for rigid body motion but requires that the relative position of the
twonodesremainsconstant.Adisadvantageofthisrigidbodyrestraintisthatflange
movementsareonlyrestrainedatthemidlinenodeanddonotsimulatecontactbetween
the channel flange and aluminum angle, which sometimes results in distortional
bucklingmodesthatwouldnotbephysicallypossible.
Rigid body connection
between top (and bottom)
flange midline nodes
Rigid body reference node

1 inch mesh spacing (typ.)
Channel 1
Channel 2
Figure4.53Channelandholemeshingdetailsandmodelingofaluminumanglestraps
ElementmeshingisperformedwithacustombuiltMatlabprogramwrittenbythe
author(SeeAppendixA).Figure4.53providesanexampleofatypicalFEmesh,where
131
theholesaredefinedwithaseriesofelementlinesradiatingfromtheopening.Figure
4.54 provides a closeup view of the rounded corner meshing of the channels. Two
elementsmodeltheroundedcornersherebecauseS9R5elementshavequadraticshape
functions which allow the initial curved geometry. Refer to Section 2.2 for more
informationonmodelingroundedcornerswithS9R5elements.
2-ABAQUS S9R5 finite elements used for rounded corners
of channels (max element aspect ratio of 16 to 1)
Figure4.54ABAQUSmeshingdetailsforCsectionroundedcorners
Concentrated loads are applied to the beam specimens through vertical stub
channels connected to the beam webs with selfdrilling screws. To simulate the
distributionoftheloadintoachannelweb,theconcentratedloadisappliedasagroup
ofwebpointloadsinABAQUS.Figure4.55demonstrateshowtheconcentratedloads
are applied to the beam webs in the finite element models. The web local buckling
restraint(essentiallydoublingupofthewebattheloadingpoint)providedbythestub
channels is not modeled in ABAQUS because it was observed to have a negligible
influenceontheelasticbucklingbehaviorintherelativelylongconstantmomentregions
ofthebeams.
132
Applied load
Stub channel
The transfer of load from the stub channels through the

self-drilling screws is simulated as a series of web point
loads in ABAQUS
Figure4.55ModelingofthebeamconcentratedloadsinABAQUS
4.3.1.4
Elasticbucklingresultsandmodedefinitions
The beam specimen elastic buckling modes were reviewed in ABAQUS by the
authortoidentifythepurelocal(L)anddistortional(D)bucklingmodesaswellasany
mixedelasticmodescreatedbytheadditionofwebholes.Lateraltorsionalbucklingis
restrainedbythetopflangelateralbracingandaluminumanglestraps(seeFigure4.51),
althoughotherpossibleglobal(G)bucklingmodesarepossibleasdiscussedinSection
4.3.1.4.3.
The mode identification process for beams with holes is guided by the
experiences obtained in Section 4.2.4 for coldformed steel compression members with
holes. Csection columns with web holes exhibited unique mixed buckling modes
wheredistortionoftheflangesneartheholemixeswithlocalbuckling(LHmode).In
133
thisbeamstudymixedlocaldistortionalmodesareagainobserved,aswellaslocalweb
holemodesinitiatedbythecompressioncomponentofthestressgradientfrombending.
4.3.1.4.1
LocalBuckling
Slotted holes in the beam specimen webs initiate unique local buckling modes
and reduce the critical elastic local buckling moment Mcrl in most cases. The shallow
beamspecimenwithoutholes(nominaldepthof2.5inches)inFigure4.56exhibitslocal
buckling in both the top flange and web. The addition of slotted web holes creates a
newlocalbucklingmode,theLH2mode.TheLH2modeoccurswhenthestripofweb
abovetheholebucklesintwohalfwaves.Thismodeoccursbecausethefundamental
local buckling halfwavelength of the crosssection, Lcrl, is less than the length of the
hole.ThecriticalelasticbucklingmomentfortheLH2modeis8percentlessthanthatof
the pure L mode, suggesting that this local hole mode may influence the load
deformationresponseofthebeam.
Figure 4.57 compares the elastic buckling behavior of a slightly deeper beam
(nominal depth of 3.625 inches) with and without holes. The addition of slotted web
holes again creates the LH2 mode with a critical elastic buckling moment that is 17
percentlessthanthepureLmode.Figure4.58,Figure4.59,andFigure4.60summarize
the influence of slotted holes on the local buckling behavior of deeper beams with
nominal heights of 6 inches, 8 inches, and 12 inches respectively. The LH mode is
identifiedinthesedeeperbeamdepthsasthebucklingofthestripofwebabovethehole
intoasinglehalfwave.TheLH2modeisobservedinthe6in.and8in.deepbeamsbut
134
withahighercriticalelasticbucklingmomentthattheLHmode.TheLH2modeisnot
observedforthe12in.deepspecimensinceLcrlforthisspecimenexceedsthelengthof
thehole.
LH2
Mcrl/Myg=0.77
L
Mcrl/Myg=0.82
L
Mcrl/Myg=0.83
Figure4.56Localbucklingmodesforspecimen2B,20,1&2(H)withandwithoutholes
LH2
Mcrl/Myg=2.49
L
Mcrl/Myg=3.00
L
Mcrl/Myg=3.02
135
LH
Mcrl/Myg=0.75
L
Mcrl/Myg=1.05
LH2
Mcrl/Myg=0.87
L
Mcrl/Myg=1.07
LH
Mcrl/Myg=0.63
L
Mcrl/Myg=0.78
LH2
Mcrl/Myg=0.72
L
Mcrl/Myg=0.79
Figure4.59LocalbucklingmodesforspecimenBP40(H)withandwithoutholes
136
LH
Mcrl/Myg=0.85
L
Mcrl/Myg=0.96
L
Mcrl/Myg=0.89
4.3.1.4.2
DistortionalBuckling
Figure 4.61 compares the influence of slotted web holes on the distortional
bucklingofashallowbeamspecimen(nominalheightof2.5inches).Forthespecimen
withholes,auniqueDH+Lmodeisobservedwithacriticalelasticbucklingmoment20
percentlessthatthepureDmode.ThismodehassimilarcharacteristicstotheLHmode
in beams (see Section 4.3.1.4.1), especially the buckling of the strip above the hole into
one halfwave. The DH+L mode is expressed more as a distortional mode though
because the compression flange is wide relative to the unstiffened strip. The D mode
without holes becomes a mixed distortionallocal mode (D+L) when holes are added,
although the critical elastic buckling moment is not significantly affected in this case.
This specimen is sensitive to mixing of local and distortional modes because of the
137
relativelythinsteelsheetthicknesstof0.0346inches.Itisalsonotedthattheholehas
onlyasmallinfluenceonthepureDmodehalfwavelengthforthisspecimen.
The DH distortional buckling mode at the hole is also observed for a slightly
deeper beam specimen (nominal height of 3.625 inches) in Figure 4.62. The sheet
thicknessforthesechannelsisroughlydoublethatofthepreviouslydiscussedspecimen
(t=0.71inches)andtheholedepthisunchanged.McrlishigherthanMcrdbecauseofthe
increased thickness, resulting in DH and D modes without local buckling interaction
whentheslottedholesarepresent.ThecriticalelasticbucklingmomentoftheDHmode
is13percentlessthanthatofthepureDmode.
DH+L
Mcrd/Myg=0.83
D
Mcrd/Myg=1.08
Half wavelength=13 inches
D+L
Mcrd/Myg=1.03
Figure4.61Distortionalbucklingmodesforspecimen2B,20,1&2(H)withandwithoutholes
138
D
Mcrd/Myg=2.31
DH
Mcrd/Myg=2.00
D
Mcrd/Myg=2.31
Figure4.63andFigure4.64comparetheinfluenceofslottedwebholesonbeams
with nominal heights of six inches and eight inches respectively, both having a steel
sheet thickness of t=0.047 inches. For these specimens the unstiffened strip buckling
modeabovetheholeisidentifiedasLHbuckling(seeFigure4.58,Figure4.59)insteadof
DHbucklingbecausethemajorityofthebucklingdeformationoccursintheweb.The
similaritiesbetweentheLHandDHmodescanmakethemdifficulttoclassifyinsome
cases.Researchonamechanicsbasedmodalidentificationmethodisunderway.
D
Mcrd/Myg=1.56
D
Mcrd/Myg=1.56
139
D
Mcrd/Myg=1.02
D
Mcrd/Myg=1.00
Figure4.64DistortionalbucklingmodesforspecimenBP540(H)withandwithoutholes
D+L
Mcrd/Myg=1.03
D
Mcrd/Myg=1.02
Thedistortionalbucklingmodesofthedeepestbeamspecimenconsideredinthis
study (nominal depth of 12 inches) are provided in Figure 4.65. Identifying the
distortional buckling modes for the channels making up this beam are inherently
challenging because even for a member without holes, there is not a clear distinction
betweentheLandDmodes.ThecriticalelasticbucklingmomentsforasingleCsection
fromthebeamcrosssectionareprovidedatvarioushalfwavelengthsfromafinitestrip
analysis (CUFSM) in Figure 4.66 (including the modal participation factors calculated
withtheconstrainedfinitestripmethod).Onlyoneminimumexists,suggestingthatthe
modesatorneartheminimumbucklingloadareamixtureofLandDmodes.Themost
suitablemodeidentifiedbytheauthor(forthespecimenwithoutahole)inFigure4.65
alternatesbetweenlargerdistortionalhalfwavesandshorterlocalbucklinghalfwaves
intheconstantmomentregionofthechannels.Forthespecimenwiththewebholesin
140
Figure 4.65, the local halfwaves are not present and the mode resembles more of a
pure D mode. The DH mode is not observed for this specimen, which is consistent
with the buckling behavior of stiffened elements in bending (see Figure 3.25).
Unstiffenedstripbuckling(theplatemodethatishypothesizedtoinitiatetheDHmode
inbeams)doesnotoccurwhenhhole/hissmall.
D
This plot summarizes the
modal participation (L, D,
G, O) as a function of
half-wavelength
Figure4.66Elasticbucklingcurvefor12deepspecimenwithmodalparticipationsummarized,notethat
selectedLandDaremixedlocaldistortionalmodes
4.3.1.4.3
Globalbuckling
Lateraltorsional buckling is a common global (G) elastic buckling mode in

beams, although this mode is eliminated for the specimens considered here by
connectingthetwoCsectionstoetotoewithaluminumanglesandbyprovidinglateral
bracing at the compression flange in the constant moment region of the beams (see
Figure 4.51). Twisting of an individual channel about its longitudinal axis is still
possible though, even with the top flange restrained. Figure 4.67 depicts the potential
twistingmode.CUFSMisusedtoconservativelyquantifytheelasticbucklingmoment
forthismode,anditisdeterminedthatMcreismorethantentimestheyieldmomentMy
141
forthespecimensinthisstudy.SincetheMcredoesnotinfluencetheDSMpredictionas
longasMcre>2.78My,theglobaltwistingmodeisnotsummarizedinthedatabase.
Lateral bracing of top flange
Figure4.67Possibleglobalbucklingmodeoccursaboutthecompressionflangelateralbracepoint
4.3.1.5
Elasticbucklingdatabaseforbeamswithholes
Table 4.11 summarizes the dimensions and material properties of each channel
making up the beam (Channel 1 and Channel 2), including cross section and hole
dimensions,testedultimatepointloadPtest(foreachchannel)andultimatemomentMtest
(for each channel), tested specimen yield stress Fy, specimen yield moment My,g
(calculated with the gross crosssection), and My,net (calculated with the net cross
section). Fy varies from 22.0 ksi to 93.3 ksi with a mean of 48.6 ksi and standard
deviationof14ksi.Thislargevariationinyieldstresswassomewhatunexpected.
ABAQUS eigenbuckling results are summarized in Table 4.12 for each channel
considering the experiment boundary conditions both with and without holes. These
resultsareusedinSection4.3.1.6toevaluatetheinfluenceofholesonMcrlandMcrd.The
CUFSM elastic buckling results are also provided, including the fundamental
distortional halfwavelength Lcrd, which are used in Section 4.3.1.7 to evaluate the
influenceofexperimentboundaryconditionsonMcrlandMcrdandthedistortionalhalf
wavelength.
142
Table 4.10 presents the crosssection parameter ranges of the beam Csections
contained in the experiment database. All of the beam specimens have crosssection
dimensions that meet the DSM prequalification standards for ultimate strength
prediction summarized in Table 4.9 (AISIS100 2007). Four of the beam specimens
exceedtheyieldstressprequalificationlimitofFy<70ksi.
Table4.9DSMprequalificationlimitsforbeamCsections
DSM
prequalification limit
Beam parameter
Web slenderness
H/t<321
Flange slenderness B/t<75

Lip slenderness
0<D/t<34
Web / flange
1.5<H/B<17
Lip / flange
Yield stress
0<D/B<0.70
Fy<70 ksi
Table4.10Parameterrangesforbeamspecimenswithholes
min
max
D/t
5.5
22.1
H/t
40.5
257.1
B/t
16.3
58.3
H/B
1.5
7.7
143
D/B
0.18
0.42
hhole/h
0.13
0.67
F y (ksi)
22.0
93.3
Table4.11Beamexperimentcrosssectiondimensions,materialproperties,andtestedstrengths
Test
Sequence
Shan and LaBoube 1994
Schuster 1992
1
1
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
3
3
Member Length, Load

Location and Thickness
L
2,16,1&2(H)
2,20,1&2(H)
2,20,3,4(H)
3,14,1&2(H)
3,14,3&4(H)
3,18,1&2(H)
3,18,3&4(H)
3,20,1&2(H)
3,20,3&4(H)
12,14,1&2(H)
12,14,3&4(H)
12,16,1&2(H)
12,16,3&4(H)
2B,16,1&2(H)
2B,16,3&4(H)
2B,20,1&2(H)
2B,20,3&4(H)
3B,14,1&2(H)
3B,14,3&4(H)
3B,18,1&2(H)
3B,18,3&4(H)
3B,20,1&2(H)
3B,20,3&4(H)
3B,20,5&6(H)
3B,20,1&2(T)
3B,20,3&4(T)
6B,18,1&2(H)
6B,18,3&4(H)
6C,18,1&2(H)
6C,18,3&4(H)
6D,18,1&2(H)
6D,18,3&4(H)
6B,20,1&2(H)
8A,14,1&2(H)
8A,14,3&4(H)
8A,14,5&6(H)
8A,14,7&8(H)
8A,14,9&10(H)
8B,14,1&2(T)
8B,14,3&4(T)
8B,14,5&6(T)
8B,14,7&8(T)
8D,14,1&2(T)
8D,14,3&4(T)
8B,18,1&2(H)
8D,18,1&2(H)
8D,18,3&4(H)
8A,20,1&2(H)
8A,20,3&4(H)
8B,20,1&2(T)
8B,20,3&4(T)
8B,20,5&6(T)
8B,20,7&8(T)
8D,20,1&2(T)
8D,20,3&4(T)
8D,20,5&6(T)
12B,16,1&2(H)
12B,16,3&4(H)
12B,16,5&6(H)
12B,16,7&8(H)
BP440(H)
BP540(H)
BP640(H)
BP765(H)
BP865(H)
BP965(H)
CP440(T)
CP540(T)
CP640(T)
CP765(T)
CP865(T)
CP965(T)
in.
150.0
150.0
150.0
150.0
150.0
150.0
150.0
150.0
150.0
192.0
192.0
192.0
192.0
150.0
150.0
150.0
150.0
150.0
150.0
150.0
150.0
150.0
150.0
150.0
150.0
150.0
192.0
192.0
192.0
192.0
192.0
192.0
192.0
192.0
192.0
192.0
192.0
192.0
192.0
192.0
192.0
192.0
192.0
192.0
192.0
192.0
192.0
192.0
192.0
192.0
192.0
192.0
192.0
192.0
192.0
192.0
192.0
192.0
192.0
192.0
168.0
168.0
168.0
168.0
168.0
168.0
168.0
168.0
168.0
168.0
168.0
168.0
shear
span
in.
39.0
39.0
39.0
39.0
39.0
39.0
39.0
39.0
39.0
60.0
60.0
60.0
60.0
39.0
39.0
39.0
39.0
39.0
39.0
39.0
39.0
39.0
39.0
39.0
39.0
39.0
60.0
60.0
60.0
60.0
60.0
60.0
60.0
60.0
60.0
60.0
60.0
60.0
60.0
60.0
60.0
60.0
60.0
60.0
60.0
60.0
60.0
60.0
60.0
60.0
60.0
60.0
60.0
60.0
60.0
60.0
60.0
60.0
60.0
60.0
48.0
48.0
48.0
48.0
48.0
48.0
48.0
48.0
48.0
48.0
48.0
48.0
t
in.
0.062
0.039
0.039
0.077
0.077
0.044
0.044
0.044
0.044
0.098
0.098
0.055
0.055
0.059
0.059
0.033
0.033
0.071
0.071
0.044
0.044
0.036
0.036
0.036
0.029
0.029
0.046
0.046
0.048
0.048
0.046
0.046
0.033
0.074
0.074
0.074
0.065
0.065
0.067
0.067
0.065
0.065
0.065
0.065
0.045
0.046
0.046
0.031
0.031
0.031
0.031
0.031
0.031
0.043
0.043
0.043
0.060
0.060
0.060
0.060
0.047
0.047
0.047
0.047
0.047
0.047
0.048
0.048
0.048
0.048
0.048
0.048
Yield Stress and Moment of Inertia
Material
Properties
E
ksi
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
29500
nu
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
Hole Dimensions
Hole Type
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Tri-slotted
Tri-slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Tri-slotted
Tri-slotted
Tri-slotted
Tri-slotted
Tri-slotted
Tri-slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Tri-slotted
Tri-slotted
Tri-slotted
Tri-slotted
Tri-slotted
Tri-slotted
Tri-slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Tri-slotted
Tri-slotted
Tri-slotted
Tri-slotted
Tri-slotted
Tri-slotted
L hole
in.
2.000
2.000
2.000
4.000
4.000
4.000
4.000
4.000
4.000
4.000
4.000
4.000
4.000
4.000
4.000
4.000
4.000
4.000
4.000
4.000
4.000
4.000
4.000
4.000
4.500
4.500
4.000
4.000
4.000
4.000
4.000
4.000
4.000
4.000
4.000
4.000
4.000
4.000
4.500
4.500
4.500
4.500
4.500
4.500
4.000
4.000
4.000
4.000
4.000
4.500
4.500
4.500
4.500
4.500
4.500
4.500
4.000
4.000
4.000
4.000
4.02
4.02
4.02
4.53
4.53
4.53
4.65
4.65
4.65
4.61
4.61
4.61
h hole
in.
0.750
0.750
0.750
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.500
1.5
1.5
1.5
2.48
2.48
2.48
1.69
1.69
1.69
2.52
2.52
2.52
r hole
in.
0.375
0.375
0.375
0.750
0.750
0.750
0.750
0.750
0.750
0.750
0.750
0.750
0.750
0.750
0.750
0.750
0.750
0.750
0.750
0.750
0.750
0.750
0.750
0.750
----0.750
0.750
0.750
0.750
0.750
0.750
0.750
0.750
0.750
0.750
0.750
0.750
------------0.750
0.750
0.750
0.750
0.750
--------------0.750
0.750
0.750
0.750
0.750
0.750
0.750
0.750
0.750
0.750
-------------
H1
in.
2.510
2.500
2.510
3.680
3.690
3.750
3.650
3.650
3.670
12.080
12.050
11.960
12.070
2.460
2.470
2.420
2.420
3.650
3.640
3.610
3.620
3.610
3.610
3.600
3.560
3.560
6.060
6.050
5.960
5.950
6.020
6.020
5.920
8.060
8.070
8.070
8.030
8.040
8.050
8.050
8.020
8.030
7.950
7.950
7.950
8.000
8.000
7.930
7.930
7.970
7.960
7.950
7.950
7.940
7.940
7.950
11.950
11.980
11.960
11.970
7.99
7.99
7.99
7.99
7.99
7.99
7.99
7.99
8.03
7.99
8.03
7.99
144
H2
in.
2.510
2.480
2.520
3.680
3.690
3.650
3.640
3.710
3.690
12.070
12.000
11.970
11.960
2.460
2.460
2.420
2.430
3.620
3.630
3.630
3.630
3.600
3.610
3.600
3.570
3.560
6.050
6.020
5.960
5.980
6.020
6.020
5.920
8.060
8.070
8.070
8.030
8.040
8.050
8.040
8.020
8.030
7.960
7.950
7.940
8.000
8.000
7.930
7.920
7.970
7.960
7.950
7.950
7.940
7.940
7.950
11.950
12.020
11.970
11.960
7.99
7.99
7.99
7.99
7.99
7.99
7.99
7.99
8.03
7.99
7.99
7.99
Channel
1
2
B 11
B 21
B 12
B 22
D 11
D 21
D 12
D 22
in.
1.610
1.600
1.590
1.650
1.630
1.560
1.560
1.560
1.560
1.640
1.640
1.570
1.560
1.620
1.630
1.630
1.630
1.620
1.630
1.610
1.620
1.630
1.640
1.630
1.620
1.620
1.620
1.620
1.980
1.970
2.420
2.430
1.630
1.380
1.380
1.370
1.390
1.390
1.640
1.640
1.630
1.630
2.480
2.470
1.590
2.420
2.420
1.380
1.370
1.630
1.630
1.630
1.630
2.490
2.460
2.490
1.630
1.630
1.630
1.630
1.61
1.61
1.61
1.58
1.61
1.61
1.58
1.58
1.61
1.61
1.58
1.61
in.
1.610
1.600
1.620
1.640
1.620
1.560
1.580
1.640
1.590
1.630
1.600
1.570
1.570
1.630
1.620
1.640
1.640
1.660
1.620
1.650
1.660
1.620
1.630
1.630
1.650
1.680
1.620
1.620
1.990
1.980
2.430
2.430
1.620
1.380
1.380
1.380
1.390
1.380
1.630
1.640
1.640
1.630
2.500
2.490
1.580
2.450
2.450
1.390
1.380
1.640
1.630
1.630
1.630
2.450
2.460
2.460
1.630
1.630
1.630
1.630
1.61
1.61
1.61
1.58
1.58
1.58
1.58
1.61
1.61
1.61
1.61
1.61
in.
1.630
1.600
1.580
1.630
1.640
1.570
1.560
1.550
1.550
1.690
1.670
1.570
1.570
1.620
1.620
1.630
1.630
1.630
1.620
1.650
1.650
1.630
1.640
1.620
1.680
1.690
1.550
1.550
1.980
1.990
2.430
2.430
1.520
1.380
1.380
1.380
1.390
1.380
1.640
1.640
1.640
1.630
2.470
2.470
1.580
2.440
2.450
1.380
1.390
1.630
1.620
1.630
1.640
2.450
2.440
2.450
1.630
1.620
1.630
1.620
1.61
1.61
1.61
1.58
1.61
1.58
1.58
1.58
1.58
1.61
1.58
1.63
in.
1.610
1.600
1.600
1.630
1.630
1.580
1.570
1.590
1.610
1.630
1.710
1.560
1.580
1.610
1.630
1.620
1.620
1.630
1.630
1.620
1.640
1.620
1.630
1.630
1.600
1.610
1.550
1.550
1.990
1.980
2.430
2.430
1.530
1.380
1.380
1.370
1.400
1.380
1.640
1.640
1.630
1.630
2.490
2.480
1.580
2.430
2.430
1.380
1.370
1.620
1.630
1.630
1.630
2.490
2.480
2.480
1.630
1.630
1.630
1.630
1.61
1.61
1.61
1.58
1.58
1.58
1.58
1.58
1.58
1.61
1.61
1.61
in.
0.400
0.420
0.360
0.570
0.530
0.580
0.560
0.520
0.600
0.690
0.650
0.500
0.420
0.470
0.470
0.420
0.420
0.540
0.540
0.510
0.500
0.460
0.460
0.460
0.590
0.580
0.470
0.470
0.640
0.600
0.700
0.700
0.440
0.490
0.500
0.410
0.430
0.460
0.630
0.640
0.640
0.660
0.640
0.660
0.470
0.610
0.600
0.410
0.450
0.610
0.620
0.610
0.610
0.640
0.640
0.620
0.530
0.470
0.510
0.480
0.47
0.47
0.47
0.47
0.47
0.47
0.51
0.51
0.53
0.51
0.51
0.53
in.
0.450
0.410
0.420
0.550
0.530
0.560
0.570
0.560
0.560
0.600
0.640
0.610
0.530
0.460
0.520
0.420
0.410
0.550
0.470
0.520
0.500
0.470
0.470
0.460
0.640
0.630
0.470
0.480
0.590
0.650
0.620
0.700
0.470
0.480
0.410
0.500
0.480
0.440
0.640
0.640
0.630
0.610
0.480
0.480
0.470
0.690
0.700
0.440
0.430
0.610
0.580
0.600
0.620
0.590
0.590
0.620
0.540
0.500
0.500
0.550
0.47
0.47
0.47
0.47
0.47
0.47
0.51
0.51
0.51
0.51
0.51
0.51
in.
0.420
0.420
0.470
0.560
0.620
0.580
0.540
0.550
0.520
0.600
0.650
0.520
0.580
0.510
0.520
0.500
0.500
0.490
0.490
0.500
0.520
0.460
0.470
0.460
0.620
0.620
0.500
0.500
0.590
0.640
0.620
0.610
0.440
0.410
0.410
0.490
0.480
0.450
0.670
0.660
0.640
0.610
0.470
0.450
0.480
0.690
0.700
0.450
0.430
0.600
0.580
0.600
0.610
0.590
0.590
0.620
0.520
0.550
0.510
0.560
0.47
0.47
0.47
0.47
0.47
0.47
0.51
0.51
0.51
0.51
0.51
0.51
in.
0.430
0.410
0.410
0.520
0.550
0.540
0.540
0.560
0.590
0.620
0.640
0.430
0.530
0.510
0.460
0.500
0.500
0.500
0.540
0.500
0.520
0.470
0.470
0.470
0.610
0.570
0.500
0.510
0.640
0.630
0.700
0.620
0.420
0.430
0.500
0.410
0.450
0.480
0.660
0.650
0.630
0.660
0.620
0.610
0.470
0.620
0.600
0.430
0.440
0.620
0.620
0.610
0.620
0.640
0.650
0.630
0.530
0.530
0.520
0.490
0.47
0.47
0.47
0.47
0.47
0.47
0.51
0.51
0.51
0.51
0.53
0.51
in.
0.124
0.078
0.078
0.154
0.154
0.088
0.088
0.088
0.088
0.196
0.196
0.110
0.110
0.118
0.118
0.066
0.066
0.142
0.142
0.088
0.088
0.072
0.072
0.072
0.058
0.058
0.092
0.092
0.096
0.096
0.092
0.092
0.066
0.148
0.148
0.148
0.130
0.130
0.134
0.134
0.130
0.130
0.130
0.130
0.090
0.092
0.092
0.062
0.062
0.062
0.062
0.062
0.062
0.086
0.086
0.086
0.120
0.120
0.120
0.120
0.094
0.094
0.094
0.094
0.094
0.094
0.096
0.096
0.096
0.096
0.096
0.096
Fy
ksi
37.23
33.70
33.70
63.72
63.72
46.92
46.92
46.82
46.82
35.93
35.93
49.11
49.11
53.59
53.59
67.15
67.15
81.36
81.36
53.13
53.13
63.71
63.71
63.71
25.51
25.51
47.17
47.17
75.08
75.08
30.77
30.77
93.26
31.04
31.04
31.04
56.29
56.29
32.58
32.58
53.14
53.14
54.71
54.71
72.32
22.00
22.00
37.96
37.96
44.89
44.89
44.89
44.89
38.59
38.59
38.59
60.64
60.64
60.64
60.64
38.87
38.87
38.87
38.87
38.87
49.02
49.02
49.02
49.02
49.02
49.02
49.02
Channel
1
2
Experimental
Results
M y,g
M y,g
M y,net
M y,net
P test
M test
k*in
12.25
7.13
7.17
43.11
43.83
19.05
18.18
18.11
18.14
167.26
167.02
124.61
126.99
16.68
16.80
11.99
11.99
49.80
49.49
20.81
20.98
20.36
20.48
20.22
6.82
6.84
37.34
37.26
70.77
71.28
32.19
32.16
50.95
53.04
53.16
53.77
86.17
85.70
58.37
58.29
91.54
91.07
113.41
113.04
82.33
33.87
33.97
27.95
27.94
37.40
37.13
37.27
37.42
54.64
54.46
54.95
169.78
170.43
169.55
170.58
46.76
46.76
46.76
46.33
46.76
58.55
60.19
60.19
60.81
60.75
60.62
61.09
k*in
12.18
7.04
7.19
42.72
43.07
18.30
18.20
18.90
18.93
164.89
166.45
123.30
125.25
16.62
16.57
11.95
12.01
49.51
49.91
20.81
21.05
20.25
20.40
20.33
6.68
6.64
37.26
37.10
71.51
71.59
32.55
32.20
50.90
53.24
53.93
53.02
86.08
86.09
58.25
58.09
91.20
91.71
116.36
115.71
81.96
33.58
33.48
27.90
27.80
37.47
37.42
37.32
37.41
55.30
55.30
55.27
170.15
171.33
169.96
169.32
46.76
46.76
46.76
46.33
46.33
58.43
60.19
60.32
60.75
60.75
60.98
60.75
k*in
12.18
7.09
7.14
42.35
43.05
18.74
17.86
17.79
17.83
167.09
166.85
124.48
126.86
15.93
16.06
11.46
11.46
48.89
48.58
20.44
20.61
20.00
20.12
19.86
6.70
6.73
37.14
37.06
70.43
70.94
32.06
32.02
50.66
52.88
53.00
53.61
85.91
85.45
58.22
58.14
91.30
90.83
113.16
112.79
82.10
33.80
33.90
27.86
27.85
37.30
37.03
37.17
37.32
54.52
54.34
54.83
169.61
170.26
169.38
170.41
46.64
46.64
46.64
45.74
46.18
57.82
59.95
59.95
60.57
59.96
59.84
60.30
k*in
12.11
7.00
7.16
41.96
42.30
17.98
17.88
18.59
18.61
164.73
166.28
123.17
125.12
15.88
15.84
11.43
11.49
48.61
48.99
20.45
20.68
19.89
20.04
19.97
6.56
6.52
37.06
36.89
71.17
71.25
32.41
32.07
50.61
53.08
53.77
52.86
85.83
85.83
58.10
57.93
90.95
91.47
116.10
115.45
81.73
33.51
33.41
27.81
27.71
37.37
37.32
37.22
37.31
55.18
55.18
55.15
169.98
171.16
169.79
169.15
46.64
46.64
46.64
45.74
45.74
57.69
59.95
60.08
60.52
59.96
60.19
59.96
kips
1.04
0.46
0.46
3.7
3.54
1.35
1.37
1.35
1.43
7.16
7.50
4.38
4.79
1.345
1.36
0.6
0.635
4.31
4.255
1.6
1.51
1.2
1.1
1.335
0.425
0.455
1.64
1.7
3.425
3.445
1.67
1.7
1.15
3.675
3.7
3.64
4.37
4.31
3.225
3.89
3.735
5.375
5.895
5.925
2.76
2.1
1.84
1.005
0.985
1.37
1.4
1.4
1.4
2.5
2.7
2.6
6.5
6.4
6.4
6.7
3.2
3.1
3.2
3.1
3.2
3.2
3.5
3.3
3.5
3.4
3.4
3.4
k*in
10.1
4.5
4.5
36.1
34.5
13.2
13.4
13.2
13.9
107.4
112.5
65.7
71.9
13.1
13.3
5.9
6.2
42.0
41.5
15.6
14.7
11.7
10.7
13.0
4.1
4.4
24.6
25.5
51.4
51.7
25.1
25.5
17.3
55.1
55.5
54.6
65.6
64.7
48.4
58.4
56.0
80.6
88.4
88.9
41.4
31.5
27.6
15.1
14.8
20.6
20.6
21.4
20.7
38.1
39.8
39.0
97.3
96.6
95.8
100.0
37.9
36.8
38.2
37.7
38.2
38.2
41.4
39.4
41.6
41.3
40.9
40.8
Table4.12Beamexperimentelasticbucklingresults
Channel 1
M cr l ,LH M cr l ,LH2 M cr l ,L
Test Sequence
Shan and LaBoube 1994
Schuster 1992
1
1
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
3
3
ABAQUS elastic buckling WITHOUT holes,

ABAQUS elastic buckling with holes, experiment boundary conditions
2,16,1&2(H)
2,20,1&2(H)
2,20,3,4(H)
3,14,1&2(H)
3,14,3&4(H)
3,18,1&2(H)
3,18,3&4(H)
3,20,1&2(H)
3,20,3&4(H)
12,14,1&2(H)
12,14,3&4(H)
12,16,1&2(H)
12,16,3&4(H)
2B,16,1&2(H)
2B,16,3&4(H)
2B,20,1&2(H)
2B,20,3&4(H)
3B,14,1&2(H)
3B,14,3&4(H)
3B,18,1&2(H)
3B,18,3&4(H)
3B,20,1&2(H)
3B,20,3&4(H)
3B,20,5&6(H)
3B,20,1&2(T)
3B,20,3&4(T)
6B,18,1&2(H)
6B,18,3&4(H)
6C,18,1&2(H)
6C,18,3&4(H)
6D,18,1&2(H)
6D,18,3&4(H)
6B,20,1&2(H)
8A,14,1&2(H)
8A,14,3&4(H)
8A,14,5&6(H)
8A,14,7&8(H)
8A,14,9&10(H)
8B,14,1&2(T)
8B,14,3&4(T)
8B,14,5&6(T)
8B,14,7&8(T)
8D,14,1&2(T)
8D,14,3&4(T)
8B,18,1&2(H)
8D,18,1&2(H)
8D,18,3&4(H)
8A,20,1&2(H)
8A,20,3&4(H)
8B,20,1&2(T)
8B,20,3&4(T)
8B,20,5&6(T)
8B,20,7&8(T)
8D,20,1&2(T)
8D,20,3&4(T)
8D,20,5&6(T)
12B,16,1&2(H)
12B,16,3&4(H)
12B,16,5&6(H)
12B,16,7&8(H)
BP440(H)
BP540(H)
BP640(H)
BP765(H)
BP865(H)
BP965(H)
CP440(T)
CP540(T)
CP640(T)
CP765(T)
CP865(T)
CP965(T)
k*in
65.6
17.0
16.9
------------------------9.3
9.3
----28.3
28.1
15.6
15.6
15.6
8.6
8.6
28.0
28.1
36.4
35.5
35.0
34.9
10.7
103.0
104.6
98.0
69.2
70.8
86.1
86.5
79.2
80.1
96.1
96.7
25.9
34.0
33.9
8.0
8.1
8.9
9.0
8.9
8.9
28.3
28.1
28.1
59.2
57.6
59.0
57.9
29.3
29.3
29.3
26.9
27.4
27.4
30.2
30.3
31.0
30.7
30.4
31.1
k*in
------153.7
149.2
29.6
29.7
29.4
29.8
--------53.8
53.4
9.3
9.3
124.1
--29.5
29.3
16.2
16.1
16.1
8.7
8.7
32.6
32.6
41.9
41.1
39.7
39.6
12.3
117.5
117.9
109.1
81.0
80.2
98.7
99.2
90.7
91.6
109.5
110.0
29.4
38.3
38.2
9.1
9.2
10.1
10.2
10.1
10.1
32.1
31.9
31.9
--------33.5
33.5
33.5
30.8
30.8
31.2
34.9
35.0
35.6
35.2
34.8
35.4
k*in
67.0
17.0
17.1
193.1
194.3
38.6
38.2
37.7
38.3
248.8
241.7
44.6
41.9
57.3
58.5
9.9
9.7
150.2
151.0
36.7
36.4
20.2
20.1
20.1
10.4
10.5
39.9
39.9
50.6
49.8
47.3
45.3
15.1
131.3
134.2
128.6
88.5
90.3
110.6
111.1
101.7
102.6
121.2
121.9
32.8
42.3
42.2
10.1
10.3
11.4
11.4
11.4
11.3
35.6
35.4
35.3
62.0
60.4
61.6
60.7
37.3
37.3
37.3
37.1
37.4
37.4
39.9
39.9
40.5
40.3
39.8
40.5
M crd,DH
k*in
50.3
----115.7
109.9
28.3
28.5
28.0
29.2
--------43.5
43.0
10.0
10.2
99.5
102.0
35.7
35.3
19.7
19.6
19.6
--------59.8
57.3
47.3
47.4
--133.0
133.0
120.3
90.9
90.6
------------------------------53.6
56.5
56.6
---------------------------------
Channel 1
Channel 2
M crd,D
k*in
38.5
15.1
13.0
106.5
101.8
39.3
36.6
33.8
40.3
250.9
243.5
51.0
44.0
39.9
41.6
12.4
12.3
114.9
114.3
38.9
40.5
24.4
23.0
23.0
15.9
16.1
58.3
58.6
73.2
74.9
59.2
66.7
28.1
129.6
130.2
119.9
86.8
89.0
152.3
153.3
137.1
140.3
137.4
138.3
46.7
76.7
75.9
16.0
16.7
30.4
28.6
27.9
28.3
62.6
62.6
63.3
71.6
69.0
75.6
69.6
47.2
47.2
47.2
47.0
47.6
47.7
51.7
52.6
54.8
53.5
52.8
53.6
L crd *
in.
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
14.0
14.0
14.0
14.0
12.0
12.0
12.0
14.0
14.0
16.0
16.0
25.0
25.0
12.0
12.0
20.0
23.0
24.0
27.0
12.0
6.0
6.0
5.0
5.0
6.0
12.0
12.0
12.0
14.0
24.0
24.0
10.0
27.0
24.0
12.0
10.0
17.0
14.0
14.0
13.0
24.0
24.0
24.0
11.0
11.0
11.0
11.0
12.0
12.0
12.0
12.0
12.0
12.0
11.0
12.0
12.0
12.0
12.0
12.0
M cr l ,LH M cr l ,LH2 M cr l ,L
k*in
66.3
16.8
17.0
------------------------9.3
9.3
----28.4
28.1
15.6
15.6
15.6
8.7
--28.0
28.1
35.7
36.3
34.0
34.9
10.8
103.0
98.3
104.7
71.5
69.5
86.1
86.5
79.2
78.3
90.2
90.2
25.8
35.2
35.3
8.1
8.1
8.9
8.8
8.9
8.9
27.5
27.6
27.8
59.2
58.5
58.4
59.9
29.3
29.3
29.3
26.9
27.0
27.0
30.2
30.7
30.7
30.7
30.8
30.7
k*in
------153.7
149.2
29.6
29.8
30.2
29.8
--------53.2
54.6
9.3
9.3
124.1
--29.5
29.3
16.2
16.1
16.1
8.7
8.7
32.6
32.6
41.2
41.9
38.8
39.6
12.4
116.6
109.6
118.0
78.5
78.7
98.7
99.2
90.7
89.7
103.9
103.9
29.3
39.5
39.6
9.1
9.1
10.2
10.0
10.1
10.1
31.3
31.4
31.6
--------33.5
33.5
33.5
30.8
30.8
30.8
34.9
35.3
35.3
35.2
35.2
35.2
k*in
67.0
17.0
17.2
193.1
194.2
38.6
38.2
37.6
38.4
234.0
236.6
46.8
45.0
57.3
58.5
9.9
9.7
149.3
150.1
36.7
36.4
20.2
20.1
20.1
10.4
10.2
39.9
40.1
49.9
50.6
44.5
45.3
15.3
132.6
129.0
133.1
91.1
88.7
110.8
111.3
101.7
100.6
115.2
115.2
32.7
43.5
43.7
10.3
10.2
11.4
11.3
11.3
11.4
34.8
34.9
35.1
62.2
61.2
61.3
62.5
37.3
37.3
37.3
37.1
37.0
37.1
39.9
40.3
40.2
40.3
40.3
40.2
L crd * approximated in ABAQUS
145
M crd,DH
k*in
50.3
17.5
--115.7
110.0
28.3
28.5
28.1
28.5
--------43.0
47.0
10.0
10.0
99.5
91.7
35.7
35.3
19.7
19.6
19.6
--------58.8
59.7
46.6
47.4
--130.8
120.7
133.2
91.4
88.6
------------------------------53.6
56.5
56.6
---------------------------------
M crd,D
k*in
41.7
14.6
15.0
103.9
101.8
36.9
37.1
36.4
36.9
234.6
238.1
56.4
46.1
39.9
41.6
12.4
11.9
113.5
100.5
39.1
40.4
24.6
23.3
23.0
15.9
15.8
58.3
58.6
73.2
77.5
59.2
66.7
28.1
128.3
120.3
130.2
90.2
87.2
152.3
153.3
137.1
136.8
125.6
126.2
46.7
76.7
75.9
16.8
16.7
30.4
27.0
28.2
28.3
62.6
62.6
63.3
71.6
70.6
70.3
73.6
47.2
47.2
47.2
47.0
45.9
47.1
51.7
53.3
53.2
53.5
53.6
53.6
L crd *
in.
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
14.0
14.0
14.0
14.0
12.0
12.0
12.0
14.0
14.0
16.0
16.0
25.0
25.0
12.0
12.0
20.0
23.0
24.0
27.0
12.0
6.0
6.0
5.0
5.0
6.0
12.0
12.0
12.0
10.0
13.0
12.0
10.0
27.0
24.0
10.0
10.0
17.0
12.0
14.0
13.0
24.0
24.0
24.0
11.0
11.0
11.0
11.0
12.0
12.0
12.0
12.0
12.0
12.0
11.0
12.0
12.0
12.0
12.0
12.0
M cr l
M crd
k*in
67.56
17.41
17.29
192.81
192.98
38.82
38.07
37.81
38.28
279.02
268.64
47.71
45.90
57.15
57.44
9.95
9.96
149.82
150.47
36.63
36.36
20.17
20.08
20.09
10.41
10.24
39.38
39.39
49.97
49.16
44.76
44.85
14.93
129.97
130.42
123.80
86.93
87.44
107.87
108.35
99.22
100.03
118.25
118.89
31.96
41.15
41.02
9.86
9.99
11.06
11.11
11.07
11.05
33.85
34.45
34.43
63.23
61.49
62.74
61.83
36.39
36.39
36.39
36.12
36.44
36.49
38.88
38.90
39.48
39.17
38.72
39.43
k*in
38.67
15.11
13.02
108.11
103.74
39.60
37.12
33.96
40.56
259.43
250.06
51.22
45.66
47.86
48.17
12.96
13.09
114.50
115.07
41.75
41.23
24.92
24.86
24.91
20.76
20.28
58.49
58.87
65.68
78.83
64.75
68.02
28.30
128.02
128.44
116.30
85.59
95.68
151.94
152.82
140.34
141.76
137.30
143.78
47.69
78.20
77.69
17.06
17.28
34.95
33.01
32.28
33.66
64.10
64.07
65.10
72.18
71.40
68.34
73.80
47.48
47.48
47.48
47.00
47.68
47.73
52.75
52.84
55.34
53.34
52.57
55.28
Channel 2
L crd *
in.
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
15.0
15.0
15.0
15.0
20.0
20.0
12.0
12.0
19.0
19.0
24.0
24.0
12.0
6.0
6.0
6.0
5.0
7.0
11.0
11.0
11.0
12.0
24.0
24.0
11.0
24.0
24.0
10.0
11.0
18.0
18.0
16.0
16.0
26.0
26.0
24.0
11.0
11.0
11.0
12.0
11.0
11.0
11.0
12.0
12.0
12.0
11.0
11.0
11.0
12.0
12.0
12.0
M cr l
M crd
k*in
67.97
17.26
17.27
192.81
193.09
38.46
38.07
37.48
38.22
271.46
260.27
49.93
48.11
57.01
58.26
9.90
9.92
149.82
150.13
36.63
36.36
20.17
20.08
20.09
10.41
10.34
39.38
39.64
49.25
49.91
44.05
44.85
15.07
127.47
124.20
128.76
88.37
85.87
107.98
108.50
99.22
98.12
112.42
112.38
31.88
42.37
42.50
10.02
9.95
11.08
10.96
11.03
11.08
34.64
33.93
34.16
63.49
62.15
62.41
63.79
36.39
36.39
36.39
36.12
36.09
36.13
38.88
39.28
39.19
39.17
39.18
39.14
k*in
41.93
14.63
15.02
105.69
103.74
37.25
37.83
37.25
37.44
239.70
243.99
57.30
50.85
46.82
53.24
12.96
12.75
114.50
101.52
41.75
40.72
25.40
25.31
24.91
20.76
20.48
58.49
58.87
77.15
85.21
63.81
68.02
28.30
126.88
116.81
128.47
96.81
91.40
151.94
152.82
140.34
141.76
126.65
127.17
47.69
78.20
77.69
17.06
17.28
34.95
31.93
32.71
33.66
64.10
64.07
65.10
72.30
70.50
68.34
73.80
47.48
47.48
47.48
47.00
47.03
47.09
52.75
53.55
53.43
53.34
53.44
53.37
Channel 1
L crd *
in.
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
12.0
15.0
15.0
15.0
15.0
20.0
20.0
12.0
12.0
19.0
19.0
24.0
24.0
12.0
6.0
6.0
6.0
5.0
7.0
11.0
11.0
11.0
12.0
14.0
14.0
11.0
24.0
24.0
10.0
11.0
18.0
18.0
16.0
16.0
26.0
26.0
24.0
11.0
11.0
11.0
12.0
11.0
11.0
11.0
12.0
12.0
12.0
11.0
11.0
11.0
12.0
12.0
12.0
Channel 2
M cr l
M crd
L crd
M cr l
M crd
L crd
k*in
67.34
17.23
17.13
192.63
192.17
38.47
38.02
37.81
38.26
280.76
276.49
48.73
47.84
57.50
57.27
9.85
9.85
152.67
151.76
36.82
36.62
19.95
19.84
19.92
10.43
10.43
39.13
39.16
49.74
48.80
44.52
44.52
14.82
127.62
128.58
126.98
85.18
87.17
107.41
107.91
98.79
99.81
118.12
118.74
31.78
40.75
40.58
9.80
9.94
11.00
11.07
11.01
10.99
34.47
34.29
34.19
62.52
60.45
61.98
60.88
36.22
36.22
36.22
35.95
35.95
35.95
38.71
39.10
39.03
38.99
38.92
38.99
k*in
36.38
14.46
12.57
105.73
100.37
34.97
33.32
31.56
35.31
236.79
247.64
50.74
46.59
37.15
37.08
9.82
9.82
85.95
85.53
30.06
29.50
17.75
17.67
17.71
14.39
14.18
42.30
42.28
54.18
52.00
47.53
47.37
19.61
122.79
124.46
108.98
83.61
87.94
123.59
124.67
116.37
118.65
108.07
111.45
41.89
50.55
49.76
15.47
16.75
23.66
23.97
23.66
23.64
45.07
45.48
44.07
71.58
67.40
70.20
68.02
46.03
46.03
46.03
45.72
45.72
45.72
51.31
51.56
51.52
51.50
51.46
51.50
in.
8.90
12.90
10.80
10.80
10.90
16.10
15.70
15.70
15.80
11.50
11.50
11.40
13.90
10.60
10.60
15.10
15.10
10.80
10.70
15.50
15.50
15.50
15.50
15.50
22.30
18.40
14.80
14.80
21.20
21.20
25.80
25.80
17.40
9.30
9.30
9.30
9.20
9.30
16.30
16.30
16.20
16.20
19.40
23.40
13.30
23.60
23.60
13.30
13.30
23.50
23.50
23.40
23.40
28.20
28.20
28.30
16.60
16.60
16.60
16.60
13.40
13.40
13.40
13.40
13.40
13.40
13.40
16.20
16.20
16.20
16.20
16.20
k*in
68.45
17.07
17.13
193.11
193.72
37.99
37.70
37.13
37.67
271.13
272.01
49.73
47.84
56.85
58.09
9.77
9.79
149.19
148.57
36.32
35.99
20.05
19.99
19.91
10.26
10.08
39.16
39.37
48.85
49.63
43.55
44.49
14.96
127.00
127.39
128.71
88.03
85.54
107.55
108.08
98.95
97.56
111.09
111.08
31.73
42.16
42.33
9.98
9.91
11.02
10.88
10.96
11.02
33.49
33.56
33.87
62.75
61.49
61.63
63.16
36.22
36.22
36.22
35.95
35.95
35.95
38.71
39.10
39.03
38.99
38.92
38.99
k*in
40.36
14.02
14.38
102.82
100.85
33.35
33.41
32.52
33.19
236.79
240.79
60.03
52.46
36.11
40.76
9.75
9.58
85.39
75.98
29.93
28.88
18.16
18.11
17.72
15.02
14.61
42.28
42.90
50.96
54.97
42.89
47.37
20.78
120.90
109.29
124.56
90.23
84.53
124.50
124.73
115.58
112.93
83.24
83.37
41.80
56.35
56.95
16.63
16.19
23.69
22.79
23.38
23.93
42.24
42.15
44.53
72.24
69.30
69.43
73.02
46.03
46.03
46.03
45.72
45.72
45.72
51.31
51.56
51.52
51.50
51.46
51.50
in.
10.80
12.80
13.10
10.80
10.90
15.70
15.60
15.90
15.80
11.50
11.50
13.80
11.40
10.60
12.70
15.10
15.20
12.90
10.70
15.60
15.60
15.50
15.50
15.50
22.30
22.30
14.80
14.70
17.60
21.30
25.80
25.80
17.40
9.30
9.30
9.30
11.20
9.30
16.30
16.30
16.20
13.50
16.10
16.10
13.30
28.50
28.50
13.30
13.30
23.50
19.40
23.40
23.40
23.40
23.40
23.40
16.60
16.60
16.60
16.60
13.40
13.40
13.40
13.40
13.40
13.40
13.40
16.20
16.20
16.20
16.20
16.20
4.3.1.6
Influence of holes on beam local and distortional
criticalelasticbucklingloads
4.3.1.6.1
Localbuckling
TheABAQUSlocalbucklingeigenbucklingresultsforeachbeamspecimenCsection
(Channel 1 and Channel 2) with holes is compared to the same beam specimen but
without holes in Figure 4.68. The variation in Mcr for the LH, LH2, and L modes (see
Section4.3.1.4.1fordefinition)withholesizetoflatwebdepthishighlightedinFigure
4.68a. The LH mode (buckling of the compressed unstiffened strip above a hole) is
observed only when 0.20<hhole/h<0.40, and is always the lowest buckling mode when it
exists. As hhole/h exceeds 0.40 the lowest mode switches to the LH2 mode. This trend
occurs because as h decreases, the local buckling halfwavelength decreases causing
multiple halfwaves to form in the unstiffened strip at the hole. When hhole/h<0.20 the
unstiffenedstripabovetheholeisrelativelystiff(i.e.,deeprelativetoholelength)and
platebucklingcontrolsasthelowestlocalbucklingmode.TheminimumMcrfortheLH,
LH2, and L is plotted in Figure 4.68b exhibits a similar trend to that observed for
stiffened elements in bending (see Figure 3.26a), where the maximum hole influence
occurswhenhhole/hisbetween0.30and0.40.Unstiffenedstripbuckling(LHandLH2)of
fullmemberscontrolsforhhole/hexceeding0.50whichisalsoconsistentwiththebehavior
ofastiffenedelementinbending(anddifferentfromacolumnwithholes,whereweb
localbucklingoccursawayfromtheholeforlargehhole/h).ThepresenceoftheCsection
flangesreducesthemagnitudeoftheholeinfluenceinafullmemberwhencomparedto
146
a stiffened element, which is consistent with similar observations for compression

members(seeFigure4.4).
1.5
1.5
LH controls
LH2 controls
L controls
Mcrl,ABAQUS hole/Mcrl,ABAQUS no hole
Mcrl,ABAQUS hole/Mcrl,ABAQUS no hole
LH
LH2
L
0.5
0.1
0.2
0.3
0.4
0.5
hhole/h
0.6
0.7
0.8
0.9
0.5
0.1
0.2
0.3
0.4
0.5
hhole/h
0.6
0.7
0.8
0.9
Figure4.68InfluenceofholesonbeamspecimenMcrl (Channel1andChannel2plotted)considering(a)all
localbucklingmodesand(b)thelowestlocalbucklingmode
4.3.1.6.2
TheABAQUSdistortionalbucklingeigenbucklingresultsforeachbeamspecimenC
section(Channel1andChannel2)withholesiscomparedtothesamebeamspecimen
but without holes in Figure 4.69. The variation in Mcr for the DH and D modes (see
Section4.3.1.4.2fordefinition)withholesizetoflatwebdepthishighlightedinFigure
4.69a. The DH mode is often the lowest distortional mode in Figure 4.69b, especially
whenhhole/hisbetween0.20and0.40.Thismodeisinitiatedbyunstiffenedstripbuckling
and is related to the LH mode, and therefore its maximum influence in this region is
expected.
Theratioofwebdepthtoflangewidthisanimportantparametertoconsiderwhen
differentiatingbetweentheLHandDHmodesforbeamswithholes.TheDHmodeis
most prevalent in the range 2<H/B<6 as shown in Figure 4.70. As the beam depth
increases relative to flange width (H/B>6) the distortional tendency associated with
147
unstiffened strip buckling decreases and the DH mode transitions to the LH (or LH2)
mode.
1.5
1.5
DH controls
D controls
Mcrd,ABAQUS hole/Mcrd,ABAQUS no hole
DH
D
0.5
0.1
0.2
0.3
0.4
0.5
hhole/h
0.6
0.7
0.8
0.9
0.5
0.1
0.2
0.3
0.4
0.5
hhole/h
0.6
0.7
0.8
0.9
Figure4.69InfluenceofholesonbeamspecimenMcrd(Channel1andChannel2plotted)asafunctionof
holedepthtoflatwebdepthconsidering(a)alldistortionalbucklingmodesand(b)thelowestdistortional
bucklingmode
1.5
1.5
DH controls
D controls
DH
D
0.5
4
H/B
0.5
4
H/B
Figure4.70InfluenceofholesonbeamspecimenMcrd(Channel1andChannel2plotted)asafunctionof
webdepthtoflangewidthconsidering(a)alldistortionalbucklingmodesand(b)thelowestdistortional
bucklingmode
4.3.1.7
Influence of experiment boundary conditions on beam
localanddistortionalcriticalelasticbucklingloads
4.3.1.7.1
Localbuckling
The influence of experiment boundary conditions on the elastic buckling

behavior is evaluated by comparing the ABAQUS critical elastic buckling moment
148
Mcrl (without holes) of each Csection making up the beam specimens to the local
buckling moment determined with the finite strip software CUFSM. Since the finite
strip method considers elastic buckling of each channel individually under a constant
moment, the comparison of ABAQUS and CUFSM results isolate the influence of the
aluminumanglestrapsatthetopandbottomflanges,aswellasthelateralbracingand
theapplicationoftheconstantmomentasaseriesofpointloadsintheexperiments.The
experiment loading and boundary conditions have a minimal influence on Mcrl for the
specimens considered in this study as shown in Figure 4.71. This result is consistent
withthelocalbucklingmodeshapesinSection4.3.1.4.1,whereitwasobservedthatthe
formationoflocalbucklinghalfwavesintheconstantmomentregionwereunimpeded
bytheanglestraps.
Mcrl,no hole, ABAQUS/Mcrl,CUFSM
1.5
0.5
50
100
150
H/t
200
250
Figure4.71InfluenceoftestboundaryconditionsonMcrl
149
300
4.3.1.7.2
The influence of the experiment boundary conditions on the distortional

buckling behavior is evaluated by comparing the critical elastic buckling moment Mcrd
(without holes) from the ABAQUS eigenbuckling analyses to the buckling moment of
each channel individually determined with the finite strip software CUFSM. The
comparison of ABAQUS and CUFSM results isolates the influence of the aluminum
angle straps, lateral bracing and the load application method on the critical elastic
moment results. The experiment test conditions provide a significant boost to Mcrd as
showninFigure4.72b,whichishypothesizedtoberelatedtotherestraineddistortional
bucklingcausedbythealuminumanglestraps.Thishypothesisissupportedbyexisting
research on unrestrained elastic distortional beam buckling (no compression flange
connections), which observed similar CUFSM and ABAQUS eigenbuckling results (Yu
and Schafer 2006). The pure D distortional halfwavelengths approximated from
ABAQUS(forspecimenswithoutholes)inFigure4.72bareoftenshorterrelativetothe
predicted halfwavelengths from a finite strip analysis. This trend is a direct result of
the angle spacing (12 on center for Test Sequences 1 and 3, 6 for Test Sequence 2),
whichislessthanthefundamentalLcrdformanyoftheCsections.Thechangeinhalf
wavelength away from the natural halfwavelength of the distortional mode increases
thecriticalelasticbucklingmoment.ThisboostinMcrddecreaseswithincreasingH/Bas
shown in Figure 4.72a, since the fundamental Lcrd also decreases as beam web depth
increasesrelativetoflangewidth.
150
1.6
1.5
1.2
Lcrd,ABAQUS no hole/Lcrd,CUFSM
Mcrd,no hole, ABAQUS/Mcrd,CUFSM
1.4
1
0.8
0.6
0.4
0.5
0.2
0
4
H/B
4
H/B
Figure4.72Influenceoftestboundaryconditionson(a)Mcrdand(b)onthedistortionalhalfwavelength
The boost in Mcrd from the restraint of the beam compression flanges exhibits a
linear trend when plotted against the ratio of Lcrd (from CUFSM) versus the restraint
spacing Sbrace in Figure 4.73. A linear equation is fit to this trend, resulting in a useful
approximationoftherestraintboost:
L
Dboost = 0.15 crd
S brace
L
+ 0.85 , crd 1
S brace
(4.15)
2
Beam database
Eq. (4.15)
DSM Design Guide
1.8
Mcrd,ABAQUS no hole/Mcrd,CUFSM
1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
0
0.5
1.5
2
2.5
3
Lcrd,CUFSM/Sbrace
3.5
4.5
Figure4.73BoostinMcrdfromtheanglerestraintsincreasesasthefundamentaldistortionalhalfwavelength
increasesrelativetotherestraintspacingSbrace
151
TheDSMDesignGuidessuggestedmodificationtoMcrdwhenL<=Lcrdisalsoplottedin
Figure4.73(AISI2006,Section4.2):
*
M crd ( L < Lcrd ,CUFSM ) = M crd
( L Lcrd ,CUFSM )
ln (L Lcrd ,CUFSM
(4.16)
where M*crd is the minimum distortional critical elastic buckling moment read from
CUFSM.LisassumedequaltoSbracewhenplottedinFigure4.73,i.e.thedistortionalhalf
waveisassumedtoformbetweentheflangebraces.TheDSMDesignGuideprediction
forMcrdishigherthanthatproposedbyEq.(4.16)becauseformanyofthebeamsinthe
ABAQUSgenerated elastic buckling database, Lcrd was shortened but not completely
restrainedbetweenbraces.Ontheotherhand,theABAQUSeigenbucklinganalysesdid
not simulate contact between the angles and the flanges (only the bending and shear
stiffness of the angles), and therefore the actual Mcrd most likely lies between the two
predictions.
4.3.2 Approximate prediction methods for use in design

4.3.2.1
4.3.2.1.1
Localbuckling
Predictionmethod
The approximate method for predicting the local elastic buckling behavior of cold
formed steel beams is similar to the method for columns presented in Section 4.2.7.1.1
LocalbucklingisassumedtooccurastheminimumofMcr ofthegrosscrosssection(as
calculated in the Direct Strength Method) and local buckling of the compressed
unstiffenedstripadjacenttothehole,Mcrh.Themethodcapturesthelowestunstiffened
stripbucklingmode,eithertheLHorLH2mode,withtheproceduredescribedinFigure
152
4.17. When the hole length is longer than the fundamental halfwavelength of the net
crosssectionLcrh,thentheLH2modegoverns.WhentheholelengthislessthanLcrh,the
LHmodegoverns.
TopredictMcrhfromthenetcrosssectioninCUFSM,thecrosssectionisrestrainedto
isolatelocalbucklingfromdistortionalbucklingasshowninFigure4.74.Compressed
corners should be restrain in the direction normal to the neutral axis about which
bendingoccurs(cornersexperiencingtensionneednotberestrained).Itisimportantto
avoid fully restraining a crosssection element, since this prevents Poissontype
deformations and artificially stiffens the crosssection. The only time both the x and z
directions of a corner should be restrained is if a hole isolates two compressed
intersecting elements (as in the case of a flange hole in a Csection, see Figure 4.74a).
Finally, when holes isolate two compressed elements of a crosssection (similar to the
flangeholeinthecolumnhatsection,seeFigure4.16b),theisolatedelementshouldbe
removedfromthecrosssection.Thispredictionmethodisvalidatedinthenextsection
usingthebeamelasticbucklingdatabasedevelopedinSection4.3.1.5.
153
Restrain compressed corners

in direction normal to neutral
axis (typ.)
Compression
z
Compression
Restrain isolated
compressed corners in
x and z
Neutral axis
Tension
Tension
Web hole
Flange hole
a
b
Compression
Compression
Neutral axis
Tension
Tension
Figure4.74GuidelinesforrestrainingbeamnetcrosssectionsintheCUFSMlocalbucklingapproximate
method
4.3.2.1.2
Methodverificationusingelasticbucklingdatabase
The finite strip prediction method is used to predict Mcrl for the 144 Csections
describedinTable4.11.ThesepredictionsarecomparedtotheABAQUSeigenbuckling
resultsfromTable4.12(theminimumofL,LH,andLH2modes),anddemonstratesthat
thefinitestripapproximatemethodisviableandconservativeoverawiderangeofhole
widthsandbeam depths.AcleartransitionfromLandLH2bucklingtoLHbuckling
occursastheCsectionsincreasesindepthasshowninFigure4.75a.Thisobservationis
consistent with finite element eigenbuckling observations (see Figure 4.56 to Figure
4.60), where as beam depth increases the halfwavelength of the netsection increases
beyond the length of the hole, resulting in a switch from unstiffened strip buckling in
twohalfwaves(LH2)toonehalfwave(LH).Themeanandstandarddeviationofthe
ABAQUStopredictedratioforMcrlare1.14and0.16respectively.
154
2
L predicted
LH2 predicted
LH predicted
1.8
1.6
1.4
Mcrl,ABAQUS/Mcrl,predicted
Mcrl,ABAQUS/Mcrl,predicted
1.6
1.2
1
0.8
0.6
1.4
1.2
1
0.8
0.6
0.4
0.4
0.2
0.2
50
100
150
H/t
200
250
L predicted
LH2 predicted
LH predicted
1.8
300
0.1
0.2
0.3
0.4
0.5
hhole/h
0.6
0.7
0.8
0.9
Figure4.75ComparisonofABAQUStopredictedMcrlforCsectionswithholesinthebeamdatabaseasa
functionof(a)webdepthand(b)holewidthrelativetoflatwebdepth
4.3.2.2
4.3.2.2.1
Predictionmethod
Theweightedaverageandmechanicsbasedfinitestripmethodsforpredicting
McrdofcolumnswithholesintroducedinSection4.2.7.2.1areemployedheretopredict
the distortional critical elastic buckling load of coldformed steel beams with holes.
These approximate methods are evaluated against the ABAQUS Mcrd (the minimum of
theDHandDmodes)fromthebeamexperimentdatabaseinTable4.12.
4.3.2.2.2
Methodverificationusingelasticbucklingdatabase
Figure 4.76 plots Mcrd determined with ABAQUS versus the predictions using the
weightedaverage and mechanicsbased approximate methods. The ABAQUS Mcrd

(withholes)ismultipliedbytheratioofMcrdfromCUFSMtoMcrdfromABAQUSwithout
holestoeliminatetheinfluenceoftheboundaryconditionsandtoallowforaconsistent
comparison between the ABAQUS results (with only the hole influence) and the
prediction method. The ABAQUS to mechanicsbased prediction ratio is more
155
accurate (ABAQUS to predicted mean of 1.04and standard deviation of 0.02) than the
weightedaverage prediction (ABAQUS to predicted mean of 1.10 and standard
deviationof0.06),whichisconsistentwiththeverificationstudyforcolumnswithholes
Mcrd,ABAQUS/Mcrd,predicted*Mcrd, CUFSM/Mcrd,ABAQUS no hole
inSection4.2.7.2.1.
2
weighted average
mechanics-based
1.8
1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
0
0.1
0.2
0.3
0.4
0.5
hhole/h
0.6
0.7
0.8
0.9
Figure4.76ComparisonofmechanicsbasedandweightedaveragepredictionmethodstoABAQUS
resultsforthedistortionalbucklingloadMcrdofCsectionswithholesintheelasticbucklingdatabase
4.3.2.3
Globalbuckling
The weighted thickness and weighted properties approximate methods
presented in Section 4.2.7.3.1 are now implemented to predict Mcre for a beam with
uniformlyspacedholesloadedwithaconstantmoment.
156
4.3.2.3.1
DescriptionofPredictionMethod
TheweightedthicknessandweightedpropertiesapproximatesforIy,J,andCw
are employed with the classical lateraltorsional stability equation to predict Mcre of a
beamwithholes(Chajes1974):
2
M cre = EI y GJ + ECw 2 .
L
(4.17)
4.3.2.3.2
ExampleandVerification
The long SSMA 1200S16268 member evaluated in Section 4.2.7.3.2 as a column is
now evaluated as a beam with a uniform moment along the member to compare
prediction methods to ABAQUS results. The ABAQUS boundary conditions and
appliedloadingaredescribedinFigure4.77. Thebeamendsaremodeledaswarping
freeandthecrosssectionatthelongitudinalmidlineiswarpingfixed.Warpingatthe
memberendsisvisibleinFigure4.77forthislateraltorsionalbucklingmode.
Cross-section dof fixed in 2

and 3 (warping free)

and 1 at longitudinal
midline (warping fixed)
2
Warping free end detail
1 (x)

and 3 (warping free)
Moment applied as
consistent nodal loads on
cross section (typ.)
Figure4.77ABAQUSboundaryconditionsandappliedloadingforan
SSMA1200S16268beamwithholes(hhole/h=0.50shown)
157
Figure 4.78 demonstrates that both the weighted stiffness and weighted
properties models are accurate predictors of Mcre for hhole/h0.50 in this particular case.
Forhhole/h>0.50,thereductioninpredictionaccuracyoccursbecausetheweightedaverage
approximationsforCwarenotconsistentwiththeactualphysicalbehavior(Jwasshown
tobeconsistentwiththeweightpropertiesmethodforcalculatingsectionproperties
in Section ). If a designer does not know Cw,avg, then using the net section properties
(calculatedwithCUFSM,seeFigure4.37)ortheweightedpropertiespredictionwith
Cw,net=0arebothviableoptionsforconservativelypredictingMcre.
Mcre,ABAQUS/Mcre,prediction
1.5
weighted thickness
weighted properties
net section
weighted properties, ABAQUS Cw ,avg
weighted properties, Cw ,net=0
0.5
0.1
0.2
0.3
0.4
0.5
hhole/h
0.6
0.7
0.8
0.9
SSMA1200S16268lateraltorsionalbeambucklingmode.Predictionsusingnetsectionpropertiesarealso
plottedasaconservativebenchmark.
158
Chapter 5
Experiments on cold-formed steel
columns with holes
TheelasticbucklingmodesdiscussedinChapter4andtheirinfluenceontheload
deformation response of coldformed steel columns can be readily observed and
quantified with experiments. In this study, 24 coldformed steel lipped Csection
columnswithandwithoutslottedwebholesaretestedtofailure.Thecolumnlengths
andcrosssectiondimensionsarespecificallychosentoexploretheconnectionbetween
local, distortional, and global elastic buckling modes, ultimate strength, and the
resulting failure mechanisms. The elastic buckling behavior is evaluated for each
specimenwithafiniteelementeigenbucklinganalysis,takingcaretoaccuratelysimulate
the tested boundary conditions and measured specimen dimensions. These elastic
bucklingresultsareusedtoprovideameansofunderstandingthevarieddeformation
responseunderload.Thecolumnsaretestedwithfrictionbearingboundaryconditions
wheretheendsofeachspecimenaremilledflatandparallel,andbeardirectlyagainst
159
steelplatens.Recommendationsaremadetoadviseotherresearchersontheviabilityof
the frictionbearing boundary conditions when testing short and intermediate length
columns.
5.1 Acknowledgements
Thecoldformedsteelcolumntestsdescribedinthischapterwerecompletedwitha
teameffortfromtheindividualsbelow:
EricHarden
LatrobeHallMachineShop
WalterKrug
MarylandHallMachineShop
MichaelFranckowiak
MarylandHallMachineShop
Dr.RachelSangree
JohnsHopkinsPostdoctoralResearcher
JackSpangler
SeniorMechanicalEngineerStructuresLab
NickolayLogvinosky
StructuresLabTechnician
MarioFasano
JohnsHopkinsSenior
RebeccaPierce
JohnsHopkinsFreshman
DawneshiaSanders
BaltimorePolytechnicInstituteSenior
AlexanderPei
HighSchoolIntern
Clark Western Building Systems in Dundalk, MD graciously donated the structural

studs.
5.2 TestingProgram
Twentyfour coldformed steel lipped Csection columns with and without pre
punchedslottedwebholesweretestedtofailure.Theprimaryexperimentalparameters
are column crosssection, column length, and the presence or absence of slotted web
holes. The specimen naming convention, as it relates to the testing parameters, is
definedinFigure5.1.
160
No Holes
362-1-24-NH
362-2-24-NH
362-3-24-NH
SSMA 362S162-33
362-1-48-NH
362-2-48-NH
362-3-48-NH
600-1-24-NH
600-2-24-NH
600-3-24-NH
SSMA 600S162-33
600-1-48-NH
600-2-48-NH
600-3-48-NH
Holes
362-1-24-H
362-2-24-H Short Column
362-3-24-H
362-1-48-H
Intermediate
362-2-48-H
Column
362-3-48-H
600-1-24-H
600-2-24-H Short Column
600-3-24-H
600-1-48-H
Intermediate
600-2-48-H
Column
600-3-48-H
362-1-24-NH
Specimen with holes
(H) or without holes
(NH)
Cross- section type
Specimen
number within
common group
(1,2,3)
Nominal specimen
length, 24 in. or 48 in.
Figure5.1Columntestingparametersandnamingconvention
5.2.1 Rationale for selecting specimen dimensions

5.2.1.1
Crosssectiontypes
TwoindustrystandardcrosssectionsfromtheSteelStudManufacturersAssociation
(SSMA2001),362S16233and600S16233,wereevaluatedinthisstudy.The362S16233
crosssectionhasanominalwebwidthof3.62in.,whilethe600S16233webiswiderat
6.00 in. Both sections have a 1.62 in. flange and nominal sheet thickness of 0.0346 in.
SpecificmeasureddimensionsareprovidedinSection5.2.4.
Thebucklinghalfwavelengthsthatformalongthelengthofthespecimensarecross
section dependent, and can be calculated with the semianalytical finite strip method
(FSM)(Schaferanddny2006).FSMassumessimplysupportedboundaryconditions,
and therefore the local and distortional halfwavelengths for the crosssections studied
here,asprovidedinTable5.1,areonlyaguideastotheexpectedhalfwavelengthinthe
fixedfixedtests.TheFSMhalfwavelengthsarestillausefulreferencewhendecidingon
specimenlengths(seeSection5.2.1.2)andidentifyingbucklingmodes(seeSection5.3.2),
especially as specimen length increases and local and distortional buckling half
161
wavelengths converge to the fundamental (simply supported) halfwavelengths

reportedinTable5.1.
Table5.1FSMlocalanddistortionalbucklinghalfwavelengthsfornominal362S16233and600S16233
crosssections
Cross-section
362
600
5.2.1.2
Elastic buckling half-wavelength

Local (L)
Distortional (D)
in.
in.
2.8
15.4
4.7
12.2
Columnlengths
Morethan80%ofthetestedspecimenswithholesavailableintheliteraturearestub
columns,asdepictedinthespecimenlengthhistogramoftestedspecimensprovidedin
Figure5.2.(Thehistogramisconstructedwiththespecimensfromtheelasticbuckling
database inSection 4.2.6.) Stub columns accommodate local buckling halfwaves, but
due to their short length, distortional buckling is typically restrained from forming at
relevantstresslevels.Thespecimenlengthsselectedinthisstudy,a24in.shortcolumn
anda48in.intermediatelengthcolumn,ensurethatatleastonedistortionalhalfwave
and multiple local halfwaves can form along the length of the column (see Table5.1).
Further,atleastforNorthAmericanpractice,theselectedlengthsaremoretypicalofthe
unbraced length of actual coldformed steel columns in an allsteel design with
bridginginplacetobracethestuds.
162
20
20
Stub column tests
18
16
16
14
14
number of specimens
number of specimens
18
12
Specimens with
holes in this study
10
8
6
12
10
8
6
10
20
30
40
50
column specimen length (in.)
60
70
10
L/H
12
14
16
18
20
Figure5.2Testedlengthsofcoldformedsteelcolumnswithholesasafunctionof(a)columnlengthLand
and(b)LversusouttooutcolumnwidthH
5.2.1.3
Holetypeandlocation
Oneslottedwebholeislocatedatthemidheightoftheshortcolumntoevaluateits
influence at the midlength of one distortional buckling halfwave. Two slotted web
holesareorientedintheintermediatelengthcolumnswithanindustrystandardspacing
of 24 in. (SSMA 2001). The holes also coincide with the locations where distortional
bucklinghalfwavesareexpectedtohavetheirmaximumdisplacementunderload.A
typical short column and intermediate length column specimen with slotted holes is
providedinFigure5.3.
163
24 in.
Short column
Intermediate length column
Figure5.3Typicalcolumnspecimenswithslottedholes
5.2.2 Column test setup

The column tests were performed with the 100 kip capacity twopost MTS
machineshowninFigure5.4.Theuppercrossheadandloweractuatorarefittedwith12
in. x 12 in. x 1 in. thick chromemoly 4140 steel platens ground flat and parallel. The
column specimens bear directly on the steel platens as they are compressed. Friction
betweenthecolumnendsandthesteelplatensaretheonlylateralforcesthatrestrainthe
column crosssection under load. An MTS load cell (model number 66123A02)
measuredtheappliedcompressiveforceoneachspecimen,andaninternalMTSlength
voltagedisplacementtransducer(LVDT)reportedactuatordisplacement.
164
All column specimens were loaded in displacement control at a constant rate of

0.004inchesperminute.Thisratewasselectedtoensurethatthe3ksiaxialstressper
minuteupperlimitintheSpecificationforstubcolumntestingwouldnotbeexceeded
(AISITS2022001).AnMTS407controllerwasusedtooperatethehydraulicactuator
duringthecompressiontests.
Fixed
Crosshead
Load Cell
Friction-bearing
boundary conditions
(specimen bears
directly on steel
platen)
Position
transducers
(with
magnet tips)
Figure5.4Columntestsetupandinstrumentation
Two Novotechnik T Series position transducers fitted with balljointed magnet

tips measured the eastwest displacements of the specimen flangelip intersections at
columnmidheight.Eachtransducerhasastrokeofsixinchesandispoweredbyone9
volt battery. The battery strengths were checked periodically to ensure that a drop in
batterychargedidnotinfluencethetransducerreadings.Theloadcellandtransducer
readings are transmitted as voltage to a PC fitted with a National Instruments data
acquisitioncard.Thevoltagesarethenconvertedtoforcesanddisplacementswiththe
conversionfactorssummarizedinTable5.2.Alldisplacementconversionfactorswere
165
determinedbytheauthorwithavoltmeteranddigitalcalipers.The dataisplottedto
the PC screen and recorded in a text file with a custom LabVIEW program (Labview
2005).
Magnetic tip
Figure5.5Novotechnikpositiontransducerwithballjointedmagnetictip
Table5.2Voltageconversionfactorsforcolumntestinstrumentation
Measurement
Tensile Force
Actuator Displacement
West Flange Displacement
East Flange Displacement
Source
MTS Load Cell
MTS Internal LVDT
Novotechnik Position Transducer
Novotechnik Position Transducer
Conversion
1 Volt = 1000 lbf
1 Volt = 0.300 in.
1 Volt = 0.678 in.
1 Volt = 0.678 in.
5.2.3 Column specimen preparation

All column specimens were cut from 8 ft. structural studs using the Central
Machinery 4 inch metal cutting band saw shown in Figure 5.15. For short columns
without holes, the whole series of specimens (for example 362124NH, 362224NH,
and 362224NH) was cut from a single 8 ft. structural stud. For all other specimen
types, each specimen was cut from its own individual stud. Tensile coupons for
materialstestingwereobtainedfromtheleftoverstudlengths.
166

Figure5.6CentralMachinerymetalbandsawusedtoroughcutcolumnspecimens
The specimen ends were milled to ensure flat and parallel bearing surfaces for
testing.Theflatnesstoleranceacrossthespecimenendisrecommendedas0.001inches
for stub columns and was adopted as the goal for this study (Galambos 1998a). The
short columns were sidemilled with a Fadel computer numericallycontrolled (CNC)
verticalmillingmachine.TheintermediatelengthcolumnsweretoolongfortheCNC
machine, and were instead sidemilled with a Bridgeport manual milling machine.
Duringinitialtrialsthemillingprocesscausedtroublesomevibrationsofthespecimen.
Thelargeclampingforcesrequired todampenthevibrationalsotendedtomodifythe
shapeoftheCsectionduringthemillingprocess.Unsatisfactoryflatnessresultswere
obtainedinthesetrials,withflatnessvariationsofupto0.010inches.
Themillingprocedurewasimprovedbyencasingthespecimenendsinbismuth
diaphragms before milling as demonstrated in Figure 5.7. The diaphragms preserved
the undeformed shape of the specimens, dampened vibration during the milling
167
process, and reduced the clamping force required to hold the specimens in place.
Bismuth is a chemical element that is relatively soft compared to steel at room
temperatureandmeltsat158degreesFahrenheit.
Figure5.7362S16233shortcolumnspecimenwithbismuthenddiaphragms
Liquidbismuthwaspouredintocustomwoodformsatthespecimenends.Once
thebismuthwasset,thespecimen(withbismuthenddiaphragms)waspositionedinthe
milling machine (Figure 5.8 through Figure 5.11). Several passes were made until the
steelcrosssectionandbismuthdiaphragmwereflush.Bothcolumnendsweremilled
without removing the specimen from the milling table to reduce the chances of
unparallel bearing ends. The bismuth diaphragms were removed from the specimen
with a few taps of a wooden mallet and then melted down for use with the next
specimen. The flatness tolerance of 0.001 inches was achieved for all but four
specimens(seeSection5.2.4.4,themaximumoutofflatnesswas+0.003in.).
168
Figure5.8600S16233shortcolumnspecimenorientedinCNCmachine
Figure5.9Anendmillisusedtopreparethecolumnspecimens
169
Figure5.10Theintermediatelengthspecimenswereendmilledinamanualmillingmachine
Figure5.11Thespecimensareclampedatthewebsonlytoavoiddistortionofthecrosssection
170
5.2.4 Column specimen measurements and dimensions

5.2.4.1
Specimenreferencesystemanddimensionnotation
All column dimensions are measured with reference to the orientation of the
specimen in the testing machine. The assumed reference system and specimen
dimensionnotationareprovidedinFigure5.15.
5.2.4.2
Crosssectionmeasurements
Theouttooutdimensionsoftheweb,flanges,andlipstiffenersweremeasured
withdigitalcalipersandaluminumreferenceplatesatthemidlengthofthespecimens.
The measurement procedure for a typical crosssection is summarized in Figure 5.12
(specimen setup) and Figure 5.13 (cross section dimensions). The outside corner radii
weremeasuredusingasetofradiusgaugeswith1/32in.increments.Thecrosssection
dimensions,basedontheaverageofthreeindependentmeasurements,areprovidedfor
eachspecimeninTable5.3.
171
Check levelness of measuring platform with

the angle indicator. The slope
perpendicular to the length of the specimen
should be as close to zero as possible.
Find and mark the longitudinal midline of

the specimen.
Clamp the specimen to the measuring

platform.
Figure5.12Setupprocedureformeasuringspecimencrosssectiondimensions
172
Clamp a beveled aluminum plate to the

flange. Use the veneer caliper to measure
the distance between the edge of the lip and
the outside face of the beveled plate. The
true dimension (D1 or D2) is then found by
subtracting the thickness of the beveled
plate from the veneer caliper reading.
Clamp beveled alumninum plates to the lip

and web, ofsetting them longituinally by
about 1/2 inch. Make sure that the beveled
faces are oriented so that they are touching
the channel.
Use the extension on the veneer caliper to

measure the distance between the outside face
of the lip plate and the inside face of the web
plate. Make sure that the extension is flush with
the flange surface. The true dimension (B1 or
B2) is found by subtracting the the thickness of
the beveled plate from the veneer caliper
reading.
Clamp beveled alumninum plates to each flange, ofsetting them

longituinally by about 1/2 inch. Make sure that the beveled faces are
oriented so that they are touching the channel.
Use the extension on the veneer caliper to measure the distance
between the outside face of one flange plate and the inside face of the
other flange plate. Make sure that the extension is flush with the web
surface. The true dimension H is found by subtracting the the
thickness of the beveled plate from the veneer caliper reading.
Figure5.13Procedureformeasuringspecimencrosssectiondimensions
173
Clamp beveled aluminum plate to flange.
Measure the flange angle with the angle

indicator (F1 and F2).
Clamp the beveled aluminum plate to the

stiffener lip. Measure the flange angle
using the angle indicator (S1 and S2).
Figure5.14Procedureformeasuringflangelipandflangewebangles
The four corner angles of each Csection are measured with a digital angle
indicator as demonstrated in Figure 5.14. The angle indicator has a precision of 0.1
degrees.TheflangelipanglesS1andS2aremeasuredatthemidlengthofthespecimens;
thewebflangeanglesF1andF2aremeasuredatmultiplepointsalongthespecimenas
denotedinTable5.4.TheCsectioncorneranglemagnitudes,basedontheaverageof
twoindependentmeasurements,areprovidedforeachspecimeninTable5.4.
174
Table5.3Summaryofmeasuredcrosssectiondimensions
H
in.
3.654
3.712
3.623
3.583
3.645
3.672
3.624
3.624
3.614
3.622
3.623
3.633
6.037
6.070
6.030
6.040
6.011
6.032
6.018
6.017
6.026
6.010
6.017
6.062
Specimen
362-1-24-NH
362-2-24-NH
362-3-24-NH
362-1-24-H
362-2-24-H
362-3-24-H
362-1-48-NH
362-2-48-NH
362-3-48-NH
362-1-48-H
362-2-48-H
362-3-48-H
600-1-24-NH
600-2-24-NH
600-3-24-NH
600-1-24-H
600-2-24-H
600-3-24-H
600-1-48-NH
600-2-48-NH
600-3-48-NH
600-1-48-H
600-2-48-H
600-3-48-H
North
West
B2
D1
D2
RT1
RT2
RB1
RB2
in.
1.550
1.586
1.677
1.650
1.627
1.674
1.611
1.609
1.604
1.602
1.594
1.604
1.599
1.582
1.601
1.594
1.608
1.606
1.621
1.596
1.585
1.598
1.589
1.632
in.
1.621
1.585
1.679
1.595
1.593
1.698
1.605
1.585
1.599
1.595
1.610
1.610
1.631
1.614
1.591
1.606
1.602
1.577
1.609
1.601
1.627
1.625
1.607
1.588
in.
0.411
0.416
0.425
0.430
0.440
0.418
0.413
0.407
0.425
0.420
0.425
0.395
0.488
0.472
0.369
0.484
0.369
0.360
0.486
0.482
0.489
0.480
0.476
0.366
in.
0.431
0.422
0.399
0.437
0.391
0.426
0.426
0.421
0.401
0.412
0.403
0.432
0.365
0.380
0.483
0.359
0.500
0.478
0.374
0.357
0.338
0.388
0.356
0.480
in.
0.188
0.172
0.188
0.188
0.188
0.188
0.172
0.188
0.188
0.172
0.172
0.172
0.172
0.203
0.156
0.172
0.172
0.172
0.172
0.172
0.172
0.188
0.172
0.172
in.
0.188
0.203
0.172
0.203
0.188
0.188
0.172
0.172
0.188
0.172
0.172
0.172
0.156
0.203
0.172
0.172
0.172
0.172
0.172
0.172
0.172
0.156
0.172
0.172
in.
0.172
0.266
0.281
0.281
0.281
0.266
0.281
0.297
0.266
0.281
0.281
0.281
0.250
0.266
0.266
0.250
0.203
0.250
0.234
0.234
0.266
0.250
0.234
0.219
in.
0.188
0.281
0.281
0.281
0.281
0.266
0.281
0.281
0.266
0.281
0.281
0.250
0.203
0.266
0.219
0.219
0.234
0.203
0.219
0.234
0.219
0.219
0.234
0.250
S1
RT1
B1
S2
D1
RT2
D2
B1
East
B2
F1
F2
hhole
rhole=hhole/2
L
RB1
RB2
Section a-a
Lhole
Hole detail
South
Front View
W1
Section b-b
W2
Figure5.15Specimenmeasurementnomenclature
175
Table5.4Summaryofmeasuredlipflangeandflangewebcrosssectionangles
X
X
S1
S2
F1
F2
X
F1
F2
X
F1
F2
X
F1
F2
X
F1
F2
Specimen
in. degrees degrees in. degrees degrees in. degrees degrees in. degrees degrees in. degrees degrees in. degrees degrees
362-1-24-NH 12 12.767
8.367
6 82.600 84.500 12 86.033 86.833 18 84.533 87.000
362-2-24-NH 12 11.367 11.567 6 86.800 84.800 12 87.600 85.467 18 86.400 83.700
362-3-24-NH 12 9.567
9.433
6 85.700 85.000 12 86.300 85.400 18 85.600 83.000
362-1-24-H 12 11.130 10.930 6 83.200 83.970 12 87.600 85.600 18 84.330 86.430
362-2-24-H 12 4.367
10.267 6 86.000 85.133 12 86.333 85.167 18 84.400 84.500
362-3-24-H 12 10.533 10.833 6 85.200 86.333 12 87.700 86.133 18 87.667 89.033
362-1-48-NH 12 7.800
10.100 12 85.100 85.600 18 84.300 85.000 24 85.000 85.600 30 84.000 85.200 36 85.300 85.700
362-2-48-NH 12 8.000
10.800 12 85.500 84.900 18 84.800 85.100 24 84.200 84.600 30 84.800 85.300 36 85.200 84.900
362-3-48-NH 12 9.100
12.200 12 86.900 84.000 18 85.800 83.900 24 85.300 84.100 30 86.400 83.400 36 86.100 83.700
362-1-48-H 12 8.500
9.800 12 86.500 84.800 18 86.600 85.000 24 85.600 84.200 30 85.500 85.100 36 86.400 84.400
362-2-48-H 12 8.300
11.200 12 86.800 84.800 18 86.500 84.200 24 85.600 83.800 30 85.500 84.100 36 86.700 83.800
362-3-48-H 12 9.700
7.300 12 85.300 85.200 18 84.700 86.100 24 84.100 85.300 30 84.400 84.700 36 85.200 85.000
600-1-24-NH 24 1.567
2.133
6 90.567 92.033 12 92.467 93.733 18 91.433 93.767
600-2-24-NH 24 1.733
2.333
6 91.000 92.033 12 91.167 94.067 18 91.467 93.333
600-3-24-NH 24 -2.167
3.500
6 93.700 89.767 12 94.067 91.033 18 92.733 89.667
600-1-24-H 24 0.967
2.033
6 89.000 91.000 12 90.400 92.267 18 91.200 92.600
600-2-24-H 24 1.800
1.100
6 94.433 90.900 12 93.233 88.733 18 91.967 89.000
600-3-24-H 24 0.100
4.100
6 93.500 90.000 12 93.300 89.300 18 90.100 86.300
600-1-48-NH 24 0.167
1.400 12 91.033 92.933 18 90.833 92.700 24 90.600 92.800 30 91.333 92.900 36 91.667 93.200
600-2-48-NH 24 2.000
2.367 12 90.767 91.900 18 90.233 92.300 24 89.900 91.867 30 90.967 92.000 36 91.467 92.767
600-3-48-NH 24 2.600
2.300 12 90.000 92.100 18 89.200 91.900 24 90.000 92.100 30 90.700 92.600 36 90.900 92.500
600-1-48-H 24 2.533
2.100 12 90.933 92.167 18 91.000 92.767 24 90.000 92.633 30 91.000 92.000 36 91.100 92.967
600-2-48-H 24 2.400
1.000 12 89.000 90.700 18 89.200 91.000 24 88.900 91.200 30 89.600 91.600 36 90.200 92.200
600-3-48-H 24 0.667
3.633 12 93.067 89.400 18 93.000 89.500 24 92.300 89.433 30 93.467 89.900 36 93.467 89.600
NOTE: X is the longitudinal distance from the south end of the specimen
176
5.2.4.3
Specimen
thickness
Allstructuralstudsweredeliveredbythemanufacturerwithazincoutercoating
applied for galvanic corrosion protection. The total zinc thickness (i.e., summation of
the zinc coating thicknesses applied to each side of the steel sheet) and the base metal
thickness(sheetthicknesswithtotalzinccoatingremoved)definedinFigure5.16were
measured for each specimen. The total zinc thickness was used to calculate the
centerline crosssection dimensions from the outtoout measurements (see Section
5.2.4.2), which were then input along with the base metal thickness into the nonlinear
finiteelementmodelsdiscussedinChapter7.Thebasemetalthicknesswasalsousedto
calculatethesteelyieldstressprovidedinSection5.2.6.
tzinc=t1+t2
t1
zinc (typ.)
base metal
t2
tbare
Figure5.16Basemetalandzincthicknessdefinitions
Totalzincthicknessandbasemetalthicknessweremeasuredforeachspecimen
from tensile coupons cut from the west flange, east flange, and web of an untested
section of structural stud. The thickness measurements were made to a precision of
0.0001 inches with a digital micrometer fitted with a thimble friction clutch. The
thickness was determined by averaging five measurements taken within the gauge
177
length of the tensile coupon (see Figure 5.27 for the definition of gauge length). The
base sheet metal thicknesses tbare,w (web), tbare,f1 (west flange), tbare,f2 (east flange) and
correspondingtotalzinccoatingthicknessestzinc,tzinc,f1,andtzinc,f2aresummarizedforeach
specimeninTable5.5.
Table5.5Specimenbaresteelandzinccoatingthicknesses
Web
Specimen
tbare,w
in.
tzinc,w
in.
362-1-24-NH
N/M
362-2-24-NH 0.0368
362-3-24-NH
362-1-24-H
0.0390
0.0030
362-2-24-H
0.0368
0.0057
362-3-24-H
0.0394
0.0027
362-1-48-NH 0.0392
0.0025
362-2-48-NH 0.0393
0.0025
362-3-48-NH 0.0389
0.0013
362-1-48-H
0.0391
0.0019
362-2-48-H
0.0390
N/M
362-3-48-H
0.0401
0.0000
600-1-24-NH
N/M
600-2-24-NH 0.0438
600-3-24-NH
600-1-24-H
0.0414
0.0042
600-2-24-H
0.0427
0.0039
600-3-24-H
0.0429
0.0031
600-1-48-NH 0.0434
0.0026
600-2-48-NH 0.0435
0.0017
600-3-48-NH 0.0436
0.0015
600-1-48-H
0.0429
0.0022
600-2-48-H
0.0429
N/M
600-3-48-H
0.0430
N/M
NOTE: N/M Not measured
West Flange
tbare,f1
tzinc,f1
in.
in.
East Flange
tbare,f2
tzinc,f2
in.
in.
0.0415
N/M
0.0372
N/M
0.0391
0.0390
0.0394
0.0393
0.0394
0.0391
0.0393
0.0391
0.0400
0.0034
0.0023
0.0018
0.0020
0.0022
0.0009
0.0017
N/M
0.0000
0.0391
0.0391
0.0394
0.0392
0.0393
0.0390
0.0394
0.0391
0.0397
0.0028
0.0034
0.0026
0.0020
0.0026
0.0017
0.0017
N/M
0.0010
0.0432
N/M
0.0438
N/M
0.0422
0.0384
0.0431
0.0436
0.0430
0.0432
0.0426
0.0428
0.0434
0.0044
0.0084
0.0026
0.0024
0.0024
0.0021
0.0023
N/M
N/M
0.0428
0.0424
0.0430
0.0434
0.0430
0.0433
0.0429
0.0431
0.0430
0.0030
0.0042
0.0036
0.0028
0.0023
0.0020
0.0021
N/M
N/M
The zinc coating was removed by immersing the tensile coupons in a ferric
chloride bath for 100 minutes. The immersion time was determined with a study of
coupon thickness variation over time for the 362224H web and the 600224H west
flangetensilecoupons.Thecouponswereremovedfromtheferricchloridebathevery
10 minutes, cleaned, and then measured. Figure 5.17 demonstrates that the coupon
thicknessconvergestoaconstantvalue,thebasemetalthickness,atapproximately100
minutes.
178
The average zinc coating thickness (i.e., average of tzinc, tzinc,f1, and tzinc,f2) for all
specimens was 0.0026 inches using the ferric chloride method described above.
Specimen coating thickness measurements were also made with a Positest DFT digital
thicknessgauge(www.defelsko.com)whichproducedanaveragecoatingthicknessfor
allspecimensof0.0016in.Atthemicroscopiclevel,thebondingofthezinctothesteel
substrateresultsinagradientfrompurezinctoamixtureofsteelandzinc(Porter1991).
This gradient complicates the identification of the nonstructural thickness of the
galvanic coating. The base thickness and coating thickness determined with the ferric
chloride method (as reported in Table 5.5) are used throughout this thesis. Accurate
identificationofthenonstructuralandstructuralcontributionsofthegalvaniccoatingis
warrantedasatopicoffutureresearch,especiallysincetheloaddeformationresponse
andultimatestrengtharesensitivetobasemetalthickness.
1.4
362-2-24-H Web Coupon

600-2-24-H West Flange Coupon
coupon thickness/initial thickness
1.2
0.8
1
0.6
0.8
0.4
0.6
0.4
0.2
0.2
0
After 100 minutes, zinc coating (tzinc=t1+t2) has

been removed with ferric chloride solution
0
0
0.2
20
0.4
40
0.6
60
time (minutes)
80
0.8
100
1
120
Figure5.17Removaloftensilecouponzinccoatingasafunctionoftime
179
5.2.4.4
Specimenendflatnessandlength
Aftereachspecimenwassawcutandmilledflat,theverticalheightgauge(with
a precision of 0.001 inches) shown in Figure 5.18 was used to measure the specimen
length and flatness. For each specimen, two independent length measurements were
takenateachroundedcornerlocationdescribedinFigure 5.19.Theheightgaugeand
specimenareplacedonthesamesteeltabletoensurethatallmeasurementsaremadein
thesamereferenceplane.Thesteeltablewascheckedforflatnesswithadialgaugeand
precisionstandbeforemeasurementsproceeded.LengthsLRT1,LRT2,LRB1,andLRB2as
wellastheaveragelengthLareprovidedforeachspecimeninTable5.6.Thespecimen
flatness,definedasthedifferencebetweenLRT1,LRT2,LRB1,andLRB2andtheaverage
lengthL,isreportedinTable5.7.Allbutfourspecimensmettheflatnesstoleranceof
0.001 inches, with intermediate length column 362248H having the maximum
deviationof+0.003inchesatLRT2.
North
Figure5.18Aheightgaugeisusedtomeasurespecimenlength
180
LRT1
LRT2
West
East
LRB2
LRB1
Figure5.19LengthsaremeasuredatthefourcornersoftheCsectioncolumn
Table5.6Measuredcolumnspecimenlength
Specimen
362-1-24-NH
362-2-24-NH
362-3-24-NH
362-1-24-H
362-2-24-H
362-3-24-H
362-1-48-NH
362-2-48-NH
362-3-48-NH
362-1-48-H
362-2-48-H
362-3-48-H
600-1-24-NH
600-2-24-NH
600-3-24-NH
600-1-24-H
600-2-24-H
600-3-24-H
600-1-48-NH
600-2-48-NH
600-3-48-NH
600-1-48-H
600-2-48-H
600-3-48-H
LRT1
in.
24.100
24.097
24.097
24.100
24.097
24.099
48.214
48.303
48.192
48.217
48.232
48.196
24.100
24.102
24.100
24.102
24.098
24.101
48.255
48.250
48.295
48.089
48.253
48.061
LRT2
in.
24.100
24.098
24.098
24.099
24.099
24.099
48.214
48.300
48.19
48.216
48.232
48.200
24.101
24.104
24.098
24.100
24.099
24.101
48.255
48.250
48.294
48.088
48.251
48.061
181
LRB1
in.
24.098
24.099
24.098
24.098
24.099
24.099
48.214
48.301
48.191
48.216
48.231
48.195
24.099
24.102
24.099
24.100
24.100
24.101
48.255
48.250
48.295
48.089
48.253
48.059
LRB2
in.
24.099
24.099
24.099
24.100
24.100
24.100
48.214
48.298
48.189
48.216
48.231
48.198
24.099
24.103
24.099
24.101
24.100
24.100
48.255
48.251
48.294
48.088
48.253
48.059
L (avg.)
in.
24.099
24.098
24.098
24.099
24.099
24.099
48.214
48.301
48.191
48.216
48.232
48.197
24.100
24.103
24.099
24.101
24.099
24.101
48.255
48.250
48.295
48.089
48.253
48.060
Table5.7Specimenendflatness
Flatness (Deviation from Average Length)
Specimen
LRT1
LRT2
LRB1
LRB2
in.
in.
in.
in.
362-1-24-NH
0.001
0.001
-0.001
0.000
362-2-24-NH
-0.001
0.000
0.001
0.001
362-3-24-NH
-0.001
0.000
0.000
0.001
362-1-24-H
0.001
0.000
-0.001
0.001
362-2-24-H
-0.002
0.000
0.000
0.001
362-3-24-H
0.000
0.000
0.000
0.001
362-1-48-NH
0.000
0.000
0.000
0.000
362-2-48-NH
0.002
-0.001
0.001
-0.002
362-3-48-NH
0.002
-0.001
0.001
-0.002
362-1-48-H
0.001
0.000
0.000
0.000
362-2-48-H
0.001
0.001
0.000
0.000
362-3-48-H
-0.001
0.003
-0.002
0.001
600-1-24-NH
0.000
0.001
-0.001
-0.001
600-2-24-NH
-0.001
0.001
-0.001
0.000
600-3-24-NH
0.001
-0.001
0.000
0.000
600-1-24-H
0.001
-0.001
-0.001
0.000
600-2-24-H
-0.001
0.000
0.001
0.001
600-3-24-H
0.000
0.000
0.000
-0.001
600-1-48-NH
0.000
0.000
0.000
0.000
600-2-48-NH
0.000
0.000
0.000
0.001
600-3-48-NH
0.001
-0.001
0.001
-0.001
600-1-48-H
0.001
0.000
0.001
0.000
600-2-48-H
0.001
-0.001
0.001
0.001
600-3-48-H
0.001
0.001
-0.001
-0.001
5.2.4.5
Locationanddimensionsofslottedholes
The length and width of the slotted holes, Lhole and hhole, were measured to a
precision of 0.001 inches with digital calipers. The eastwest locationsof theholes, W1
and W2, were measured by clamping aluminum plates to the outside surface of the
flangesandthenusingthecaliperextensiontomeasurethedistancefromtheedgeofthe
hole to the aluminum plate. (This process is similar to the crosssection measurement
proceduresdescribedinFigure5.13.)Theholesizeandweblocationdimensions,based
ontheaverageofthreeindependentmeasurements,areprovidedforeachspecimenin
Table5.8.
182
Table5.8Measuredslottedholedimensionsandlocations
Specimen
362-1-24-H
362-2-24-H
362-3-24-H
362-1-48-H
362-2-48-H
362-3-48-H
600-1-24-H
600-2-24-H
600-3-24-H
600-1-48-H
600-2-48-H
600-3-48-H
5.2.4.6
X
in.
L/2
L/2
L/2
(L-24)/2
(L-24)/2
(L-24)/2
L/2
L/2
L/2
(L-24)/2
(L-24)/2
(L-24)/2
W1
in.
0.946
1.146
0.935
1.252
1.126
0.982
2.147
2.365
2.347
2.161
2.166
2.371
W2
in.
1.141
0.967
1.114
0.974
1.016
1.112
2.361
2.155
2.166
2.375
2.351
2.162
L hole
in.
4.003
4.000
4.005
3.999
4.001
4.000
4.002
4.001
4.001
4.002
4.001
3.999
h hole
in.
1.492
1.502
1.493
1.500
1.496
1.493
1.498
1.491
1.493
1.494
1.499
1.497
X
in.
W1
in.
W2
in.
L hole
in.
h hole
in.
(L+24)/2 1.198 0.952 4.001 1.494

(L+24)/2 1.171 0.973 4.003 1.494
(L+24)/2 0.967 1.133 4.003 1.491
(L+24)/2 2.162 2.383 3.998 1.497

(L+24)/2 2.176 2.360 4.002 1.498
(L+24)/2 2.365 2.156 4.003 1.494
Webimperfections
Variationsinthespecimenwebsweremeasuredtoprovideabasisforthelocal
buckling initial imperfection magnitudes in the specimen nonlinear finite element
modelsconstructedinSection7.2.ThemeasurementsetupshowninFigure5.20usesa
dialgaugewithaprecisionof0.001inchesmountedtoalaboratorystandincontactwith
aflatsteeltable.Thespecimenwassupportedhorizontallyatbothendsbyamatching
pairofsteelbarsthatweregroundflatandparallel.Thebarswerealsoincontactwith
the steel table, ensuring that the specimen and the dial gauge were in the same
horizontalreferenceplane.Eachspecimenwebwasmarkedwithagridofmeasurement
pointsshowninFigure5.21.Thestandanddialgaugewereshiftedfromgridpointto
gridpointandelevationmeasurementswererecorded.Thevariationsfromtheaverage
elevation of the specimen web, based on an average of two measurements per grid
point,areprovidedforeachspecimeninTable5.9.
183

Figure5.20Adialgaugeandprecisionstandareusedtomeasureinitialwebimperfections
+ variation
West
1.2 inches (362 specimens)

2.3 inches (600 specimens)
Center
East
Section a-a
X
CL Web (typ.)
North
a
6 in. (typ.)
Plan view
(short and intermediate length web grid layouts)
Figure5.21Webimperfectionmeasurementgridandcoordinatesystem
184
Table5.9Initialwebimperfections(deviationsfromtheaverageelevationoftheweb)
Specimen
362-1-24-NH
362-2-24-NH
362-3-24-NH
362-1-24-H
362-2-24-H
362-3-24-H
362-1-48-NH
362-2-48-NH
362-3-48-NH
362-1-48-H
362-2-48-H
362-3-48-H
600-1-24-NH
600-2-24-NH
600-3-24-NH
600-1-24-H
600-2-24-H
600-3-24-H
600-1-48-NH
600-2-48-NH
600-3-48-NH
600-1-48-H
600-2-48-H
600-3-48-H
X Distance
West
Center
East
West
Center
East
West
Center
East
West
Center
East
West
Center
East
West
Center
East
West
Center
East
West
Center
East
West
Center
East
West
Center
East
West
Center
East
West
Center
East
West
Center
East
West
Center
East
West
Center
East
West
Center
East
West
Center
East
West
Center
East
West
Center
East
West
Center
East
West
Center
East
West
Center
East
West
Center
East
West
Center
East
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
in.
0
0.013
0.022
0.013
0.019
0.015
0.015
0.016
0.017
0.016
0.006
0.016
0.009
0.007
0.014
0.025
0.016
0.021
0.017
0.003
0.009
0.015
-0.008
-0.004
-0.005
0.006
-0.001
0.010
-0.006
-0.015
0.009
0.013
0.011
0.010
0.013
0.019
0.014
0.016
0.055
0.016
0.013
0.061
0.021
0.007
0.034
0.017
0.005
0.052
0.020
0.009
0.020
0.014
0.006
0.007
0.007
0.023
0.060
0.024
0.019
0.060
0.014
0.026
0.055
0.013
0.014
0.059
0.009
0.032
-0.033
0.011
0.017
0.028
0.018
6
-0.007
-0.005
-0.007
-0.006
-0.014
-0.009
-0.004
-0.015
-0.010
-0.008
-0.010
-0.008
-0.009
-0.014
-0.001
-0.009
-0.009
-0.002
-0.010
0.004
0.019
-0.023
-0.016
-0.008
-0.002
-0.003
0.007
-0.003
0.003
0.021
-0.007
-0.011
-0.002
-0.007
-0.005
-0.002
-0.012
0.005
-0.003
-0.019
0.004
0.002
-0.016
-0.023
-0.021
-0.015
0.003
-0.003
-0.014
-0.018
-0.015
-0.001
-0.009
-0.014
-0.003
0.016
0.006
-0.004
0.012
-0.001
-0.003
0.013
-0.003
-0.004
0.012
0.003
0.002
0.023
0.008
0.012
0.003
-0.010
12
-0.011
-0.022
-0.013
-0.010
-0.020
-0.015
-0.010
-0.023
-0.016
-0.014
Hole
-0.013
-0.020
Hole
-0.017
-0.020
Hole
-0.015
-0.014
-0.006
0.010
-0.021
-0.021
-0.011
0.003
0.002
0.006
-0.003
Hole
0.016
-0.003
Hole
-0.004
-0.012
Hole
-0.010
-0.029
-0.027
-0.014
-0.033
-0.030
-0.010
-0.018
-0.029
-0.028
-0.031
Hole
-0.018
-0.024
Hole
-0.027
-0.003
Hole
-0.022
-0.018
-0.010
-0.002
-0.016
-0.005
0.002
-0.021
-0.012
-0.010
-0.026
Hole
0.000
-0.020
Hole
0.004
0.012
Hole
-0.023
Local Variations in Web

18
24
30
-0.004
0.015
-0.013
0.015
-0.004
0.014
-0.006
0.015
-0.007
0.024
-0.008
0.014
-0.003
0.015
-0.003
0.025
-0.008
0.014
-0.001
0.016
-0.009
0.009
-0.001
0.015
-0.003
0.014
-0.007
0.010
-0.009
0.014
-0.010
0.016
-0.015
0.015
-0.002
0.015
-0.011
-0.009
-0.005
-0.006
-0.005
-0.007
0.005
0.004
-0.001
-0.013
-0.002
0.006
-0.015
0.000
0.006
-0.009
0.004
0.010
0.005
0.003
0.002
0.001
-0.002
-0.004
0.003
-0.001
-0.007
0.002
0.007
0.007
0.001
0.006
-0.002
0.010
0.007
-0.004
-0.001
0.000
-0.004
-0.006
-0.003
-0.010
-0.001
-0.003
-0.004
-0.006
-0.003
0.003
-0.010
-0.002
-0.003
-0.006
0.001
0.001
-0.024
0.013
-0.023
0.027
-0.014
0.009
-0.027
0.009
-0.024
0.031
-0.011
0.014
-0.003
0.010
0.006
0.057
-0.012
0.018
-0.019
0.011
-0.017
0.021
-0.012
0.006
-0.011
0.012
-0.001
0.051
-0.013
0.016
-0.006
0.014
-0.002
0.040
-0.016
0.004
-0.026
-0.026
-0.020
-0.018
-0.010
-0.006
-0.008
-0.001
0.003
-0.016
-0.020
-0.025
-0.004
-0.007
-0.009
0.005
0.000
-0.004
-0.021
-0.014
-0.011
-0.011
0.003
0.003
-0.010
0.002
0.008
-0.026
-0.026
-0.024
-0.007
0.002
-0.009
0.002
0.004
0.002
-0.028
-0.022
-0.012
-0.001
0.006
0.005
0.004
0.004
0.007
0.010
0.000
0.007
-0.004
-0.022
-0.014
-0.032
-0.048
-0.040
185
36
42
48
-0.005
-0.005
-0.002
0.015
0.011
0.013
0.000
-0.011
-0.013
0.007
Hole
-0.016
-0.009
Hole
-0.012
-0.003
Hole
-0.007
-0.002
-0.008
-0.005
0.021
0.012
0.016
0.011
-0.001
-0.011
0.010
-0.009
-0.022
-0.002
-0.004
-0.006
-0.004
-0.008
-0.002
0.018
0.013
0.004
0.023
-0.002
0.012
0.007
0.004
-0.001
0.000
-0.009
-0.015
0.016
0.022
0.012
0.017
0.015
0.012
-0.019
-0.009
0.004
-0.024
-0.009
0.003
-0.015
-0.008
-0.002
-0.025
Hole
0.001
-0.019
Hole
0.000
0.009
Hole
-0.045
-0.014
-0.005
0.001
-0.014
-0.006
0.002
-0.015
-0.006
-0.002
-0.007
-0.002
0.004
-0.022
0.002
-0.004
0.011
0.018
0.002
0.016
0.030
0.011
0.011
0.023
0.008
0.013
0.031
0.008
0.009
0.024
0.002
0.013
0.025
0.004
0.020
0.046
0.021
5.2.4.7
Specimenorientationinthetestingmachine
When placing the specimen in the testing machine, the southern end of the
specimenwasorientedatthebottomplatensuchthatthecenterofthecompressiveforce
was applied through the gross centroid of the Csection. The centerline of the web is
positionedinlinewiththecenterlineofthebottomplatenandoffsettowardsthebackof
thetestingmachineasdescribedinFigure5.22.Thecentroidlocationswerecalculated
using the centerline dimensions of a nominal SSMA 362S16233 and 600S16233 cross
section.
CL Platen and Column Web
0.380 in. (600S162-33)

0.502 in. (362S162-33)
Location of interior
web edge
CL Platen
Column specimen
Center of platen, center
of load, centroid of Cee
channel
FRONT OF MTS MACHINE
Plan View
(Bottom Platen)
Figure5.22Columnspecimenalignmentschematic
186
The actual cross section and thickness measurements produced centroid offsets
slightly different from the nominal offsets considered in the column tests. The
difference between the nominal and measured offsets, defined here as CG, are
providedinTable5.10.CGproducesendmomentsinthespecimensthatareseveral
ordersofmagnitudesmallerthantheappliedloadsinthisstudy.Forexample,theend
momentscreatedbyaCGof0.059inchesforspecimen600324NHarecalculatedas
2.0x106kipinchesatpeakload(Ptest=12.24kips)usingthestructuralanalysisprogram
MASTAN(ZiemianandMcGuire2005).TheassumedMASTANstructuralsystemin
Figure 5.23 demonstrates that relatively stiff compression platens and fixedfixed end
conditionseffectivelyeliminateendmomentsfromsmallloadeccentricities.
Table5.10Specimengrosscentroidandoffsetfromappliedloadduringtests
Specimen
362-1-24-NH
362-2-24-NH
362-3-24-NH
362-1-24-H
362-2-24-H
362-3-24-H
362-1-48-NH
362-2-48-NH
362-3-48-NH
362-1-48-H
362-2-48-H
362-3-48-H
600-1-24-NH
600-2-24-NH
600-3-24-NH
600-1-24-H
600-2-24-H
600-3-24-H
600-1-48-NH
600-2-48-NH
600-3-48-NH
600-1-48-H
600-2-48-H
600-3-48-H
Specimen
Measurements
tz
xcg
in.
in.
0.482
0.038
0.471
0.038
0.504
0.038
0.511
0.042
0.490
0.042
0.524
0.042
0.475
0.041
0.468
0.042
0.475
0.040
0.470
0.041
0.470
0.042
0.486
0.040
0.354
0.047
0.347
0.047
0.344
0.047
0.363
0.046
0.368
0.047
0.361
0.046
0.362
0.046
0.355
0.045
0.353
0.045
0.362
0.045
0.352
0.046
0.356
0.046
Centroid Shift
xcg - tz
used in tests
in.
in.
0.463
0.502
0.452
0.502
0.485
0.502
0.489
0.502
0.469
0.502
0.503
0.502
0.454
0.502
0.447
0.502
0.455
0.502
0.449
0.502
0.449
0.502
0.466
0.502
0.330
0.380
0.323
0.380
0.321
0.380
0.340
0.380
0.344
0.380
0.338
0.380
0.339
0.380
0.333
0.380
0.330
0.380
0.340
0.380
0.329
0.380
0.333
0.380
tz sheet thickness with zinc coating

CS difference measured and as tested centroid offsets
187
CS
in.
0.039
0.050
0.017
0.013
0.033
-0.001
0.048
0.055
0.047
0.053
0.053
0.036
0.050
0.057
0.059
0.040
0.036
0.042
0.041
0.047
0.050
0.040
0.051
0.047
CG
All translation
and rotational
DOF restrained
EI flexural rigidity
1000EI
0.1EI
No moment in
column for stiff
platen even
with load offset
EI
EI
Column specimen
(Centroid shown)
CL Applied
Load
Horizontal
translation and
rotational DOF
restrained
Platen
(typ.)
1000EI
Actuator load
0.1EI
Stiff platen
Structural System
Flexible platen
Moment Diagrams
Figure5.23Influenceofplatenbendingstiffnessonendmomentsforafixedfixedeccentriccolumn
Oncethespecimenisalignedonthebottomplaten,500lbsofcompressiveforce
wasappliedtothecolumnandweakaxisoutofstraightnessmeasurementsweretaken.
The distance from the front of the top and bottom platens to the interior web edge is
denotedasStopandSbottominFigure5.24.StopandSbottomareobtainedastheaverageofthree
independent measurements with digital calipers as shown in Figure 5.25 and then
corrected for a systematic platen offset (see Figure 5.24) and the initial web
imperfections in Table 5.9. The initial outofstraightness S provided in Table 5.11 is
188
calculatedfromStopandSbottomandimplementedasaninitialgeometricimperfectioninto
thenonlinearfiniteelementmodelsin7.2.
CL Load
Stop
a
CL Platen
Front of MTS
Machine
Stop
CL Platen
Column Specimen
(orientation exaggerated)
Platen Offset=0.084 in.
Section a-a
S (negative magnitude shown)

Sbottom
Side View
(Looking west)
Figure5.24Columnspecimenweakaxisoutofstraightnessschematic
189

Figure5.25Digitalcalipersareusedtomeasurethedistancefromthecolumnwebtoplatenedge
Table5.11Summaryofoutofstraightnesscalculations
Specimen
Sbottom
Stop
Platen Offset
Sbottom
Correction
Sbottom
Correction
Stop
Corrected for
Corrected for
Top platen edge is Corrected for
Web
Web
As
As
Web
Web
Initial out of
offset from bottom
top platen
Imperfection @
Imperfection
measured measured
Imperfection
Imperfection straightness
platen Edge
offset
X=0
@ X=L
@ X=0
@ X=L
362-1-24-NH
362-2-24-NH
362-3-24-NH
362-1-24-H
362-2-24-H
362-3-24-H
362-1-48-NH
362-2-48-NH
362-3-48-NH
362-1-48-H
362-2-48-H
362-3-48-H
600-1-24-NH
600-2-24-NH
600-3-24-NH
600-1-24-H
600-2-24-H
600-3-24-H
600-1-48-NH
600-2-48-NH
600-3-48-NH
600-1-48-H
600-2-48-H
600-3-48-H
in.
6.507
6.523
6.531
6.524
6.532
6.529
6.352
6.535
6.537
6.530
6.534
6.532
6.352
6.365
6.451
6.356
6.360
6.355
6.346
6.354
6.354
6.311
6.352
6.348
in.
6.622
6.612
6.585
6.613
6.578
6.629
6.393
6.649
6.614
6.554
6.617
6.616
6.472
6.560
6.494
6.486
6.399
6.403
6.436
6.488
6.463
6.458
6.422
6.430
in.
0.084
0.084
0.084
0.084
0.084
0.084
0.084
0.084
0.084
0.084
0.084
0.084
0.084
0.084
0.084
0.084
0.084
0.084
0.084
0.084
0.084
0.084
0.084
0.084
in.
6.591
6.607
6.615
6.608
6.616
6.613
6.436
6.619
6.621
6.614
6.618
6.616
6.436
6.449
6.535
6.440
6.444
6.439
6.430
6.438
6.438
6.395
6.436
6.432
5.2.5
190
in.
0.015
0.015
0.017
0.016
0.014
0.021
0.009
-0.004
-0.001
-0.015
0.011
0.019
0.055
0.061
0.034
0.052
0.020
0.007
0.060
0.060
0.055
0.059
-0.033
0.028
in.
6.577
6.593
6.598
6.592
6.602
6.592
6.427
6.623
6.622
6.629
6.607
6.598
6.381
6.388
6.501
6.388
6.424
6.432
6.370
6.377
6.383
6.336
6.469
6.404
in.
0.022
0.024
0.025
0.009
0.010
0.015
0.013
-0.002
0.004
-0.009
0.022
0.015
0.027
0.031
0.057
0.021
0.051
0.040
0.030
0.023
0.031
0.024
0.025
0.046
in.
6.600
6.588
6.560
6.604
6.568
6.615
6.380
6.651
6.610
6.563
6.594
6.601
6.444
6.529
6.437
6.466
6.348
6.363
6.406
6.465
6.432
6.433
6.396
6.384
in.
-0.024
0.004
0.038
-0.012
0.034
-0.023
0.047
-0.028
0.012
0.066
0.013
-0.003
-0.063
-0.141
0.063
-0.078
0.076
0.069
-0.036
-0.087
-0.049
-0.098
0.072
0.020
5.2.6 Materials testing

Tensile coupon tests were performed to obtain the steel stressstrain curve and
yieldstressfortheweb,westflange,andeastflangeofeachspecimeninthisstudy.The
tests were conducted in accordance with ASTM specification E 8M04, Standard Test
MethodsforTensionTestingofMetallicMaterials(Metric)(ASTM2004).
5.2.6.1
Tensilecouponpreparation
Tensilecouponswerealwaysobtainedfromthesame8ft.structuralstudwhich
producedthecolumnspecimen.Flatportionsofthewebandflangeswerefirstrough
cutwithametalbandsawasshowninFigure5.26,andthenfinishedtothedimensions
in Figure 5.27 with a CNC milling machine. The special jig in Figure 5.27 allowed for
threetensilecouponstobeproducedatonce.Thetensilecouponswerestrippedoftheir
zinc coating (see Section 5.2.4.3 for procedure) and then measured within the gauge
length for bare metal thickness, t, and minimum width, wmin. The minimum width is
determinedbytakingtheminimumoffiveindependentmeasurementswithinthegauge
lengthofthespecimenwithdigitalcalipers.
191

Figure5.26Tensilecouponsarefirstroughcutwithametalbansaw
0.38 in.
0.38 in.
1.97 in.
3.18 in.
1.97 in.
0.492 in. *
0.79 in.
R=0.55 in.
gauge length
1.97 in.
*nominal, actual dimension will vary slightly
Figure5.27TensilecoupondimensionsasenteredintheCNCmillingmachinecomputer
192

Figure5.28AcustomjigallowsthreetensilecouponstobemilledatonceintheCNCmachine
5.2.6.2
Tensiletestsetup
A screwdriven ATS 900 testing machine with a maximum capacity of 10 kips

was used to apply the tensile load. Tensile coupons were positioned in the machine
withfrictiongripsasshowninFigure5.29.Abubblelevelwithashort,straightedge
was used to ensure that each specimen was aligned vertically between the grips. An
MTS 634.11D54 extensometer measured engineering strain and an MTS load cell
measuredforceonthespecimen.Theextensometerwasplacedattheverticalmidlength
of the specimen, centered within the gauge length. The raw voltage data from the
extensometer and load cell were sent to a PC containing a National Instruments data
acquisitioncard.Thevoltagedatawasconvertedtotensileforceandengineeringstrain
usingtheconversionfactorsprovidedinTable5.12.Thedatawasplottedonthescreen
andrecordedtoafilewithacustomLabVIEWprogram(Labview2005).
193
Figure5.29ATSmachineusedtotesttensilecoupons
Table5.12Voltageconversionfactorsfortensilecoupontesting
Tensile Coupon Testing
Measurement
Source
Tensile Force
MTS Load Cell
Conversion
1 Volt = 1000 lbf
-5
Engineering Strain MTS Extensometer 1 Volt = 3.96x10 strain (in./in.)
5.2.6.3
Tensiletestresults
Two distinct steel stressstrain curves were observed in this study. Tensile
coupons from the 362S16233 structural studs demonstrate gradual yielding behavior,
while the tensile coupons from the 600S16233 studs demonstrated a sharp yielding
plateau.Theyieldstress,Fy,forthegraduallyyieldingspecimenswasdeterminedwith
the 0.2% strain offset method. The stressstrain curve for specimen 362348NH (East
Flange)demonstratestheoffsetmethodinFigure5.30.Theyieldstressforthesharply
yielding specimens was determined by averaging the stresses in the yield plateau.
ASTMdoesnotprovidespecificguidelinesonhowtoaveragetheplateaustresses.For
this autographic method, the averaging range is determined by using two strain offset
lines,oneat0.4%strainoffsetandtheotherat0.8%offsetasshownforspecimen60024
194
NH (West Flange) in Figure 5.31. The steel modulus of elasticity, E, was assumed as
29500ksiforallspecimenswhendeterminingtheyieldstress.Thetensilecouponyield
stresses and cross section dimensions are summarized in Table 5.13. The mean and
standard deviation for all 362S16233 and 600S16233 tensile coupons tested are
providedinTable5.14.
100
0.2% strain offset line (slope=29500 ksi)
90
Axial Tensile Stress (ksi)
80
0.8
70
0.6
60
50
0.4
40
30
YIELD STRESS (0.2% offset)=60.1 ksi
0.2
20
0
10
0
0.2
0.05
0.4
0.6
0.1
0.15
Engineering Strain,(in./in.)
0.8
0.2
1
0.25
Figure5.30Graduallyyieldingstressstraincurvewith0.2%strainoffsetmethod
100
90
Axial Tensile Stress (ksi)
80
0.8
70
0.6
60
50
0.4
40
30
YIELD STRESS (Autographic Method)=59.7 ksi
0.2
20
0
10
0
0.2
0.05
0.4
0.6
0.1
0.15
Engineering Strain (in./in.)
0.8
0.2
1
0.25
Figure5.31SharpyieldingstressstraincurveusinganautographicmethodfordeterminingFy
Table5.13Summaryofcolumnspecimensteelyieldstress
195
Specimen
tbase,w
in.
Web
wmin
in.
Fy
ksi
362-1-24-NH
0.4985
53.3
362-2-24-NH 0.0368
362-3-24-NH
362-1-24-H
0.0390
0.4945
55.9
362-2-24-H
0.0368
0.4886
52.9
362-3-24-H
0.0394
0.4945
55.6
362-1-48-NH 0.0392
0.4985
59.4
362-2-48-NH 0.0393
0.4990
59.2
362-3-48-NH 0.0389
0.4930
58.0
362-1-48-H
0.0391
0.4998
59.5
362-2-48-H
0.0390
0.4992
58.8
362-3-48-H
0.0401
0.4990
57.8
600-1-24-NH
0.4950
60.6
600-2-24-NH 0.0438
600-3-24-NH
600-1-24-H
0.0414
0.4899
61.9
600-2-24-H
0.0427
0.4964
57.8
600-3-24-H
0.0429
0.4966
59.7
600-1-48-NH 0.0434
0.4985
58.7
600-2-48-NH 0.0435
0.4985
N/C
600-3-48-NH 0.0436
0.4995
60.4
600-1-48-H
0.0429
0.4970
60.3
600-2-48-H
0.0429
0.4994
61.8
600-3-48-H
0.0430
0.4992
60.7
NOTE: N/C Tests results were not obtained
tbase,f1
in.
West Flange
wmin
in.
Fy
ksi
tbase,f2
in.
East Flange
wmin
in.
Fy
ksi
0.0415
0.4975
54.7
0.0372
0.4955
57.4
0.0391
0.0390
0.0394
0.0393
0.0394
0.0391
0.0393
0.0391
0.0400
0.4963
0.4950
0.4927
0.4965
0.4975
0.5000
0.4985
0.4961
0.4957
59.3
58.8
N/C
59.7
59.3
58.9
58.2
60.6
58.0
0.0391
0.0391
0.0394
0.0392
0.0393
0.0390
0.0394
0.0391
0.0397
0.4968
0.4945
0.4947
0.4975
0.4970
0.4930
0.4991
0.4975
0.4978
58.5
59.5
56.4
59.9
59.2
60.1
58.1
59.8
59.1
0.0432
0.4950
59.7
0.0438
0.5000
55.9
0.0422
0.0384
0.0431
0.0436
0.0430
0.0432
0.0426
0.0428
0.0434
0.4940
0.4874
0.4954
0.4955
0.4970
0.4955
0.4980
0.4962
0.4961
63.6
55.6
58.0
62.3
63.4
N/C
63.0
62.1
59.7
0.0428
0.0424
0.0430
0.0434
0.0430
0.0433
0.0429
0.0431
0.0430
0.4964
0.4938
0.4960
0.4965
0.4970
0.4965
0.4970
0.4977
0.4977
60.3
61.8
62.6
59.3
63.3
61.9
60.8
62.2
64.0
Table5.14Columnspecimensteelyieldstressstatistics
Stud Type
362S162-33
600S162-33
yield stress, Fy
mean
STDV
ksi
ksi
58.1
2.0
61.0
2.0
5.3 Elasticbucklingcalculations
Elastic buckling provides a means to categorize and potentially better understand
theloaddeformationresponseandultimatestrengthofthethinwalledcolumnsinthis
study. The local, distortional, and global elastic buckling modes and their associated
critical elastic buckling loads (Pcrl, Pcrd, Pcre) are presented here for each specimen.
Calculations are performed with a shell finite element eigenbuckling analysis as
opposedtoananalysisusingFSM(Schaferanddny2006)tocapturetheinfluenceof
theslottedwebholesandthetested(fixedfixed)boundaryconditions.
196
5.3.1 Finite element modeling assumptions
Eigenbuckling analysis in ABAQUS is performed for the 24 column specimens
(ABAQUS 2007a). All columns are modeled with S9R5 reduced integration ninenode
thin shell elements. Coldformed steel material properties are assumed as E=29500 ksi
and =0.30. The centerline Csection dimensions input into ABAQUS are calculated
using the outtoout dimensions and flange and lip angles at the midheight of each
column specimen as provided in Table 5.3 and Table 5.4. Each column specimen is
loadedwithasetofconsistentnodalloadsinABAQUStosimulateaconstantpressure
across the bearing edge of the specimen. The nodes on the loaded column face are
coupled together in the direction of loading with an ABAQUS pinned rigid body
constraint(seeFigure4.12).
5.3.2 Elastic buckling results

5.3.2.1
Buckledshapes/eigenmodes
The first (lowest buckling load) local (L) and distortional (D) buckled shapes for
specimenswithandwithoutslottedholesarecomparedinFigure5.32andFigure5.33.
TheLandDmodesforeachspecimenwereidentifiedvisuallybymanuallysearching
throughtheelasticbucklingmodesproducedintheeigenbuckinganalysis.Thenominal
crosssection halfwavelengths in Table 5.1 were compared to the halfwavelengths in
thefiniteelementmodeltoassistinthecategorization.Thelocalanddistortionalmodes
thatmostresembledtheFSMresultsforLandDmodeswereselected.Thismethodof
modal identification is neither exact nor ideal, especially when both local and
197
distortional buckling are present in the same eigenmode. Formal modal identification
hasrecentlybeendevelopedinthecontextofthefinitestripmethod(Schaferanddny
2006)andfutureworkisongoingtoextendthismethodtofiniteelementanalysesandto
problemssuchastheonesencounteredhere.
(a)
(b)
Holes cause
mixed
distortional-local
mode
Hole
terminates
web local
buckling
Hole
terminates
web local
buckling
Local Buckling (L)
Distortional Buckling (D)
Local Buckling
Figure5.32(a)Localanddistortionalelasticbuckledmodeshapesfor(a)short(L=48in.)362S16233
specimensand(b)intermediatelength(L=48in.)362S16233specimens.
(a)
(b)
Holes cause mixed

distortional local
mode
Local Buckling (L)
Hole changes
number of
half-waves
from 5 (NH)
to 6 (H)
Distortional Buckling (D)
Local Buckling
Holes
change
number of
half-waves
from 8 (NH)
to 12 (H)
Figure5.33Localanddistortionalelasticbuckledmodeshapesfor(a)short(L=48in.)600S16233specimens
and(b)intermediatelength(L=48in.)600S16233specimens.
5.3.2.2
Bucklingloads/eigenvalues
TheprimarygoalofthisresearchprogramistoextendtheDirectStrengthMethodto
coldformedsteelstructuralmemberswithholes.TheDirectStrengthMethod(DSM),a
design method for coldformed steel structural members, predicts column ultimate
strength by predicting the column failure mode and ultimate strength through
198
knowledge of the local (L), distortional (D), or global (G) elastic buckling modes. This
connection is made using the critical elastic buckling load, Pcr, and the slenderness,
definedwiththeratioofcolumnsquashloadPygtoPcrfortheL,D,andGmodes.Table
5.15summarizesPcrandPygforthespecimensevaluatedinthisstudy.Thesquashload
Pyg is calculated with the gross crosssectional area, and Pcr includes the effects of the
holesandthetested(fixedfixed)boundaryconditions.(Note,theimplicationsofusing
PygasopposedtoPy,netatthenetsectionarediscussedinChapter8.)
The influence of holes on Pcr is of interest in the context of DSM because elastic
buckling loads and slenderness are used to predict ultimate strength. To isolate the
influenceofholesonPcr,additionaleigenbucklinganalysesofthespecimenswithholes
(specimens labeled with an H) were performed, but with the holes removed (the
boundaryandloadingconditionswerenotmodifiedandthemeshusedinthemodels
wasidenticalexceptfortheremovedelementsattheholelocation).Thecomparisonof
Pcrforspecimenswithholes(H)andthenwithholesremoved(noH)isalsosummarized
inTable5.15.
199
Table5.15Criticalelasticbucklingloads,influenceofholesonelasticbuckling
Specimen
Name
Pyg
kips
ELASTIC BUCKLING
Pcrl
Pcre
kips
HOLE INFLUENCE*
Pcrd
kips
kips
Pcre/Pcre
noH
noH
Pcrl/Pcrl
Pcrd/Pcrd
362-1-24-NH
15.5
109.4
4.9
10.6
N/A
362-2-24-NH
15.6
112.5
4.8
10.2
362-3-24-NH
15.7
112.2
5.0
10.7
362-1-24-H
16.4
119.3
5.9
13.5
0.98
1.03
1.12
362-2-24-H
15.7
112.8
5.4
12.4
0.98
1.02
1.13
362-3-24-H
16.4
130.6
5.7
12.9
0.99
1.02
1.12
362-1-48-NH
16.9
30.5
5.2
9.7
N/A
362-2-48-NH
16.7
29.5
5.2
9.6
362-3-48-NH
16.6
29.6
5.1
9.5
362-1-48-H
16.6
30.0
5.3
9.4
0.94
1.03
0.98
362-2-48-H
16.8
29.7
5.2
9.3
0.94
1.03
0.98
362-3-48-H
16.8
36.2
5.7
9.6
0.95
1.03
0.98
600-1-24-NH
24.7
244.5
3.4
6.8
N/A
600-2-24-NH
24.5
234.9
3.4
6.7
600-3-24-NH
24.5
218.4
3.4
6.6
600-1-24-H
25.0
239.3
3.3
7.0
1.01
1.02
1.09
600-2-24-H
23.1
238.4
3.2
6.7
1.01
1.01
1.08
600-3-24-H
24.7
242.6
3.5
7.3
1.02
1.01
1.08
600-1-48-NH
25.1
61.8
3.5
5.2
N/A
600-2-48-NH
26.2
59.6
3.4
5.7
600-3-48-NH
25.4
60.2
3.4
5.7
600-1-48-H
25.2
56.3
3.4
5.1
0.87
1.02
1.02
600-2-48-H
25.5
53.0
3.4
5.0
0.87
1.02
1.02
600-3-48-H
25.6
55.8
3.4
5.0
0.86
1.02
1.02
* For specimens with holes (H), the holes are removed and elastic buckling calculated (noH).
The hole (H) and no hole (noH) finite element models are otherwise identical, isolating the influence of the holes.
5.3.2.3
noH
Modalinteractionatultimatestrength
An additional reason for the selection of these specimen crosssections, at these
lengths, beyond the reasons discussed in Section 5.2.1, is that the specimens provide
much needed experimental data on crosssections with potential modal interaction at
ultimate strength both with and without holes. Typically modal interaction is
understoodtobeaconcernwhentheelasticbucklingloadsofmultiplemodesareator
near the same value, and the ratio of any two elastic bucking loads (e.g., Pcrl/Pcrd) is
considered a useful parameter for study. However, for modes with different post
buckling strength and where material yielding is considered, a more pressing concern
may be the situation when both failure modes predict similar capacities. Which mode
doesthecolumnfailinifthepredictedcapacityinlocal(Pnl)anddistortional(Pnd)areat
or near the same level? What impact does a hole have on the failure mode that is
200
triggered?Inthespecimensselectedhere,theratioofPcrl/Pcrdvariesfromaminof0.44to
amaxof0.68,butisnevernear1.0.Therefore,bythistraditionalmeasurenomeaningful
interactionwouldbeanticipated.However,iftheDSMmethodologyisusedtopredict
thecapacities,asillustratedinFigure5.34,thepredictionsfortheratioofthetwolimit
statesPnl/Pndrangesfromaminof0.86toamaxof0.90inthe362S16233shortcolumns
and from a min of 1.0 to a max of 1.05 in the600S16233 short columns (the ratios are
similarforthelongcolumnspecimens).Thus,thesecrosssectionsprovideameansto
examine the potential for localdistortional modal interaction at ultimate strength, and
offer valuable data for determining any necessary modification to the DSM
methodologywhenholesarepresent.
1.2
Local DSM
Distortional DSM
Pn/Pyg
0.8
362S162-33 short columns
0.6
600S162-33 short
columns
L
D
0.4
Local(L)-Distortional(D) interaction is
expected since predicted strengths (Pn)
are of similar magnitudes
0.2
0.5
1
1.5
2
slenderness, or d
l
2.5
Figure5.34Local(L)anddistortional(D)DSMstrengthpredictionsaresimilarinmagnitudeforboth
362S16233and600S16233crosssections,indicatingthatLDmodalinteractionwilloccurduringthetested
responseofthecolumns.
201
5.3.3 Discussion of elastic buckling results

5.3.3.1
Localbuckling
Boundary conditions have little influence on the local buckling mode shapes
(comparedwithFSMLmodes),butthepresenceoftheslottedwebholescanchangethe
shape, halfwavelength, and buckling load of the first (lowest) local buckling mode
observed. In the 362S16233 specimens the web holes terminate local buckling in the
vicinityoftheholes,seeFigure5.32.Inthe600S16233specimensthewebholescause
anincreasednumberofhalfwavesalongthelengthtooccurinthelowestlocalmode,
see Figure 5.33. The presence of holes causes a slight increase in Pcrl (see Table 5.15)
which is consistent with the increased number of observed local buckling halfwaves.
ThemoreextensiveelasticbucklingstudiesChapter3andChapter4demonstratethata
holecanincreaseordecreasethenumberofbuckledhalfwaves(andthecriticalelastic
bucklingload)ofrectangularplatesandcoldformedsteelstructuralstuds.
5.3.3.2
Boundary conditions and the presence of holes have an influence on the observed
distortionalbucklingmodeshapes(comparedwithFSMDmodes)andbucklingloads.
The boundary conditions (fixedfixed) allow a smaller number of halfwaves to form
than predicted using the simply supported FSM D modes of Table 5.1. For example,
observe the restrained shape of the buckled distortional halfwave near the member
ends in Figure 5.32a. In longer specimens (see Figure 5.32b and Figure 5.33b), the
influence of the boundary conditions lessens and the halfwavelength of distortional
202
bucklingatmidheightapproachesthatoftheFSMDmodeofTable5.1.(Section4.2.6.2
explores the influence of fixedfixed boundary conditions on Pcrd using the column
experiment database.) The presence of the web holes complicates the predicted D
modes,seeFigure5.32andFigure5.33.LocalbucklingnowappearswithintheDmode
itself.ThehalfwavelengthoftheseinteractingLmodesissignificantlyshorterthanthe
lowest L modes observed. Further, and rather unintuitively, the buckling load, Pcrd,
actuallyincreaseswiththepresenceofholesintheshortcolumnspecimens(asmuchas
13%).However,thisincreaseislostatthelongerspecimenlengthwherethemaximum
changeinthebucklingloadis+/2%.Thisresultsuggeststhatintheshorterspecimens
theremovalofthematerialmostsusceptibletooutofplanebending,atthemiddepth
of the web, actually serves to stiffen the column (a localized increase in the transverse
bendingstiffnessofplateswithholeshasbeenobserved,seeFigure4.30).Thisinfluence
does not persist in the longer specimens suggesting that the increased stiffness is only
relevantwhentheDmodeisatarestrainedhalfwavelength.Thus,iftheDmodeisfree
to form (over a long enough unbraced length) the holes do not increase the elastic
bucklingload.
5.3.3.3
Globalbuckling
The global (Euler) buckled shapes for the intermediate 362S16233 and 600S16233
columns in Figure 5.35 occur as flexuraltorsional buckling, although local and

distortional deformation are both present in the mode shape for specimens with and
withoutholes,whichisanunexpectedresult.Theinteractionbetweentheglobal,local,
203
anddistortionalmodesmakestheidentificationoftheglobalmodedifficult.TheEuler
bucklingloadandmodeshapepredictedwithclassicalmethods(inCUTWP),whichdo
notallowcrosssectiondistortionandignoreholes,wereusedtodeterminetherangeof
bucklingloads(eigenvalues)tobevisuallysearched.ThereportedmodesinFigure5.35
are the ones closest to the expected buckling load exhibiting significant global
deformations. Additional eigenbuckling analyses of the 362S16233 and 600S16233
crosssectionswereperformedatalongercolumnlength(8ft.)andtheseanalysesshow
no local or distortional interaction with the global modes. Therefore, the observed
interaction is length dependent and not a fundamental feature of global buckling in
thesecrosssections.AnalternativehypothesisfortheunusualmodeshapesinFigure
5.35 is that several buckling mode shapes exist near the global critical elastic buckling
load,whichcausestheeigensolvertomisreporttheglobalmodeasalinearcombination
ofbuckledshapes.
As for the global buckling loads, the slotted holes have a small influence on the
globalbucklingloadfortheintermediatelength362S16233specimens,reducingPcrebya
maximum of 6%. However, Pcre for the intermediate length 600S16233 columns
decreasesbyamaximumof14%withthepresenceofthetwoslottedholes,whichisan
unexpected result attributed to the local and distortional modes mixing with global
buckling (i.e., Figure 5.35). Additional research work is ongoing to determine under
whatconditionsholesinfluencetheglobalcriticalelasticbucklingload.
204
Global (Euler) Buckling

Finite element eigenbuckling
analyses predict global
buckling interacting with local
and distortional buckling
362-1-48-NH
600-1-48-NH
362-1-48-H
600-1-48-H
CUTWP predictions using

classical stability theory
Figure5.35Comparisonofglobalmodeshapesforintermediatelength362S16233and600S16233
specimens.
5.4 Experimentresults
5.4.1 Ultimate strength
The peak tested compressive load for all column specimens and an average peak
loadforeachtestgroupareprovidedinTable5.16.Theslottedholesareshowntohave
onlyasmallinfluenceoncompressivestrengthinthisstudy,withthelargestreduction
being2.7%forthe362S16233shortcolumns.
205
Table5.16Specimenultimatestrengthresults
Specimen
362-1-24-NH
362-2-24-NH
362-3-24-NH
362-1-24-H
362-2-24-H
362-3-24-H
362-1-48-NH
362-2-48-NH
362-3-48-NH
362-1-48-H
362-2-48-H
362-3-48-H
600-1-24-NH
600-2-24-NH
600-3-24-NH
600-1-24-H
600-2-24-H
600-3-24-H
600-1-48-NH
600-2-48-NH
600-3-48-NH
600-1-48-H
600-2-48-H
600-3-48-H
Ptest
kips
10.48
10.51
10.15
10.00
10.38
9.94
9.09
9.49
9.48
8.95
9.18
9.37
11.93
11.95
12.24
12.14
11.62
11.79
11.15
11.44
11.29
11.16
11.70
11.16
Mean
kips
Std. Dev.
kips
10.4
0.2
10.1
0.2
9.4
0.2
9.2
0.2
12.0
0.2
11.9
0.3
11.3
0.1
11.3
0.3
5.4.2 Failure modes and post-peak ductility

5.4.2.1
Shortcolumns
Theloadingprogressionforthe362162S33shortcolumnsisdepictedinFigure5.36
(withoutahole)andFigure5.37(withahole).Bothcolumnsexhibitlocalbucklingofthe
webnearthesupportscombinedwithonedistortionalhalfwavealongthelength.This
distortional buckling pattern is consistent with that predicted by the elastic buckling
modeshapesofFigure5.32a.Forthecolumnwiththehole,localizedholedeformation
(Figure5.37,rightmostpicture)initiatesataloadofapproximately0.4Ptestandincreases
in magnitude as the test progresses. This observed deformation behavior is visually
consistent with the unstiffened strip approach discussed in Error! Reference source
otfound.,wherethestripofweboneithersideoftheholeisassumedtobehaveasan
unstiffenedelement.
206
Theinwardflangedeformationconcentratesattheholeafterpeakloadintheshort
362S16233 specimens with holes. It is hypothesized that the slotted hole reduces the
postpeakresistanceoftheweb,causingtheflangesandlipstocarrymoreofthecolumn
load.Thisreductioninpostpeakresistanceisquantifiedbyobservingthereductionin
areaundertheloaddisplacementcurveforthecolumnwiththeslottedhole,asshown
inFigure5.38.
Distortional buckling
in one half-wave at
peak load
(a) P=0 kips
(b) P=7.0 kips
(c) P=10.5 kips

(peak load)
(d) P=7.5 kips
Figure5.36Loaddisplacementprogressionforshortcolumnspecimen362224NH
207
Local buckling at hole

(unstiffened strip)
(a) P=0 kips
(b) P=10.4 kips

(peak load)
(c) P=7.0 kips
Figure5.37Loaddisplacementprogressionforshortcolumnspecimen362224H
14
Slotted hole has small

influence on peak load
362-2-24-NH
362-2-24-H
Column axial load (kips)
12
10
8
6
4
Slotted hole influences post-peak

load path and reduces column
ductility
2
0
0.05
0.1
0.15
Column axial displacement (inches)
0.2
Figure5.38Loaddisplacementcurvefora362S16233shortcolumnwith,withoutaslottedhole
208
Positiontransducersplacedatthemidheightoftheshortcolumnspecimenscapture
therateoflateralflangedisplacementassociatedwithdistortionalbuckling,D,asshown
in Figure 5.39. Figure 5.39 demonstrates that the initiation of web local buckling does
notinfluencetheaxialstiffnessofspecimen362224NH,butratherthatasofteningof
the loadaxial deformation curve coincides with the increased rate of lateral flange
movement (distortional buckling). This observation suggests that the loss in axial
stiffnessassociatedwithdistortionalbucklingplaysalargerrolethanweblocalbuckling
inthepeakloadresponseofthe362S16233shortcolumns.Theinfluenceoftheslotted
hole on lateral flange displacement is provided in Figure 5.40, where the postpeak
flangedisplacementratesaresignificantlyhigherforthe362S16233shortcolumnwith
holes.TheresultsofFigure5.40indicatethatholespotentiallyhaveasignificantimpact
on the collapse mechanisms triggered from distortional buckling. Lateral flange
displacementplotsareprovidedforallspecimensinAppendixF.
209
12
2
Local buckling
half-waves
first observed
1.5
8
Rate of flange distortion
increases as loaddisplacement curve softens
D (inches)
10
4
0.5
2
D =
+ west
west + east
+ east
2
0
0.2
0.05
0.1
0.15
Figure5.39Comparisonofloaddeformationresponseandlateralflangedisplacementsforspecimen3622
24NH
2
362-2-24-NH
362-2-24-H
1.8
1.6
Increased rate of flange distortion

is observed for column with a hole
after peak load is reached
D (inches)
1.4
1.2
1
H
0.8
Peak load
occurs here
0.6
NH
0.4
D =
0.2
0
+ west
+ east
0.05
0.1
0.15
west + east
2
0.2
Figure5.40Influenceofaslottedholeon362S16233shortcolumnlateralflangedisplacement
210
Figure5.41andFigure5.42depictthedeformationresponseofthe600S16233short
columns with and without a slotted hole. In both cases, local buckling at the loaded
ends combines with one distortional halfwave along the column length. The
distortionalbucklingpatternforthesespecimensisnotwhollyconsistentwiththeelastic
bucklingpredictionsofFigure5.33a,whichshowstwodistortionalhalfwaves;however,
specimens600224Hand600324Hdidbuckleintwohalfwaves,seeAppendixFfor
pictures.Theseresultssuggestthatgeometricimperfectionsalsohavearoletoplayin
the details of the buckling mode initiated in the loaded response. The deformation
response of the member with and without the hole is similar through the test
progression,suggestingthattheholehasasmallinfluenceoncompressivestrengthand
postpeakductilityfortheholewidthtowebwidthratiosconsideredhere.Figure5.43
confirmsthattheslottedholehasaminimaleffectonthepostpeakloadresponseofthe
column.
211
Distortional
buckling in one
half-wavelength
at peak load
Web local
buckling
(a) P=0 kips
(b) P=8.0 kips
(c) P=11.9 kips

(peak load)
(d) P=8.0 kips
Figure5.41Loaddisplacementprogressionforshortcolumnspecimen600124NH
Web local
buckling
initiates
(a) P=0 kips
Local buckling near

supports combines
with distortional
buckling (one halfwave) at peak load
(b) P=7.5 kips
(c) P=12.1 kips

(peak load)
Similar failure
mode to no hole
specimens
(d) P=8.0 kips
Figure5.42Loaddisplacementprogressionforshortcolumnspecimen600124H
212
14
600-1-24-NH
600-1-24-H
12
10
8
6
NH
Slotted hole has small

influence on post-peak
response and ductility
2
0
0.05
0.1
0.15
0.2
Figure5.43Comparisonofloaddisplacementresponseforshort600S16233columnspecimenswithand
withoutholes
5.4.2.2
Intermediatelengthcolumns
Figure5.44andFigure5.45summarizethedeformationresponseofthe362S16233
intermediatelengthcolumnswithandwithoutholes.Inbothcases,localwebbuckling
is first observed at approximately 0.45Ptest which is lower than, but the same order of
magnitudeas,thecalculatedlocalcriticalelasticbucklingloadPcrl.NoteinFigure5.45
thatthelocalbucklinghalfwavesaredampenedinthevicinityoftheholes,similarto
the elastic buckling prediction of Figure 5.32b. Further, the observed local buckling
wavesareathalfwavelengthsconsistentwiththelocalbucklingmodeinFigure5.32b,
notthoseshowninteractingwithdistortionalbuckling.(Thisobservationsupportsthe
ideathatthefundamentalelasticbucklingmodesL,D,andGarerepresentativeofthe
physicalbehaviorofthecolumnandthatthemixedmodesobservedinaneigenbuckling
analysis only exist numerically.) Three distortional buckling halfwaves become well
213
formedatapproximately0.70Ptest,overcomingthelocalhalfwavesinthewebexceptat
themidheightofthecolumn.Thisdistortionalbucklingpatternisconsistentwiththe
elasticbucklingpredictioninFigure5.32b.Figure5.46demonstratesthatthepresenceof
slottedholeshasonlyaminimalinfluenceonloadaxialdisplacementresponse.
All of the 362S16233 intermediate length columns failed soon after the peak load
with a sudden loss in loadcarrying capacity caused by global flexuraltorsional

buckling. Yielding of the column flanges reduces the torsional stiffness of the section,
and the friction end conditions could not restrain the twisting of the column. The
twistingofspecimen362348NHisquantifiedinFigure5.47asthedifferencebetween
the west and east midheight flange displacements, T, captured by the position
transducers.Thelateraldisplacementoftheflangetipsduetodistortionalbuckling(D),
also shown in Figure 5.47, is separated from the twisting effect by averaging the west
and east midheight flange displacements. Figure 5.47 shows that the crosssection is
bothopeningandtwisting,butitistheabruptincreasein Toccurringwellpastpeak
loadthatleadstothecollapseofthememberinthetest.
214
Web local buckling

with flange distortion
in three half-waves
Local buckling mixes

with distortional halfwave at peak load
Pure distortional
buckling dominates over
local buckling in this
half-wave
(a) P=0 kips
(b) P=8.0 kips
(c) P=9.5 kips

(peak load)
Figure5.44Loaddisplacementprogression,intermediatelengthcolumnspecimen362348NH
Hole dampens web local buckling by

creating two stiff web strips on either side of
hole that locally boost Pcrl above Pcrd
(a) P=0 kips
(b) P=8.0 kips
(c) P=9.4 kips

(peak load)
Figure5.45Loaddisplacementprogressionforintermediatelengthcolumnspecimen362348H
215
14
Columns fail abruptly

with a global torsional
buckling mode
12
362-3-48-NH
362-3-48-H
10
8
6
4
NH
2
H
0.05
0.1
0.15
0.2
Figure5.46Loaddisplacementcurve,362S16233intermediatecolumnwithandwithoutahole
2
1.5
Abrupt failure
D, T (inches)
1
0.5
T =
east west
2
-0.5
-1
+ east
+ west
-1.5
-2
Peak load
occurs here
D =
east + west
0.05
0.1
0.15
2
0.2
Figure5.47362S16233longcolumnmidheightflangedisplacementsshowtheglobaltorsionalfailure
mode
216
The loaddisplacement response for the intermediate length 600S16233 columns
with and without slotted holes is depicted in Figure 5.48 and Figure 5.49. Local
bucklingisobservedatapproximately0.45Ptestforbothsections.Theholesdonotrestrict
the local buckling halfwaves as was the case in the 362S16233 intermediate length
columns. This local buckling behavior is consistent with that observed in the elastic
bucklinganalysis,seeFigure5.33b.Threedistortionalhalfwavesformasthecolumns
(all3ofthe600S16233intermediatelengthspecimens)approachpeakload.Twoloud
soundsresonatefromthecolumnsnearpeakloadasthelocalwebbucklinghalfwaves
at the two column ends abruptly snap into one distortional halfwave per end. The
changefromlocaldominatedtodistortionaldominatedwebbucklingisreflectedastwo
drops in the loaddisplacement response near peak load for the 600S16233 column
withoutholes,asshowninFigure5.50.The600S16233columnwithslottedholesisnot
affectedbythisabruptmodeswitching,asitmaintainsweblocalbucklingwellbeyond
peak load. The observations suggest that in this case the holes are beneficial because
theymaintainthelocalbucklinghalfwavesthroughpeakload,allowingthecolumnto
relymoreonthepostpeakstrengthprovidedbythebuckledweb.Thismodeswitching
is a difficult challenge for numerical models and these results, repeated in 3 tests,
provides an important and challenging experimental benchmark for the numerical
modelingofthesemembersSection7.2.
217
Multiple local half-waves change to one

distortional half-wave with loud resonant
sound at peak load
(a) P=6.0 kips
(b) P=11.2 kips (c) P=11.2 kips (d) P=7.0 kips

(peak load)
Hole preserves web local

buckling through peak load
(a) P=0 kips
(b) P=11.2 kips

(peak load)
(c) P=10.0 kips
218
14
600-1-48-NH
600-1-48-H
12
10
8
Drops in load occur when

multiple web local buckling
half-waves change abruptly to
one distortional half-wave
(north end first, then south end)
6
4
2
0
0.05
0.1
0.15
0.2
Figure5.50Loaddisplacementcomparisonofintermediatelength600S16233specimenswithandwithout
holes
5.4.3 Discussion of hole influence on elastic buckling and

tested response
Bothlocalanddistortionalelasticbucklingwereobservedinthetestedresponseof
the specimens and contributed in different ways to the failure modes of the columns.
Local buckling initiated plastic folding in the web at peak load, and distortional
buckling was reflected as either opening (D) or closing (+D) of the crosssection and
yieldingoftheflangesandlipstiffeners.Allthreeoftheshort362S16233columnswith
holes exhibited a closed distortional buckling failure (+D), where the presence of the
slottedholeconcentratedtheplasticdeformationintheflangesandlipsadjacenttothe
hole.Thisresultwasdifferentfromtheshort362S16233columnswithoutholeswhere
mixed localdistortional failures were observed. The slotted holes also changed the
219
bucklinginfluenceatpeakloadintheintermediatelength600S16233specimens,where
the holes prevented local web buckling from switching to distortional buckling in all
threespecimentests.Thedeformationatpeakloadfortheintermediatelength362S162
33 and short 600S16233 specimens was less sensitive to the presence of slotted holes,
exhibiting mixed localdistortional failure modes consistent with DSM predictions (L
andDofsimilarmagnitudes)asdiscussedinSection5.3.2.3.
The visual observations in this study highlight the complex relationship between
elasticbucklingandcolumnfailureandthesensitivityoftheirinteractiontothechoice
of crosssection and column length. In the cases of the short 362S16233 and
intermediate length 600S16233 specimens, it is clearly demonstrated that holes can
influence column deformation and ductility by changing how elastic buckling modes,
localanddistortionalinthiscase,affecttheaxialstiffnessandplasticdeformationofthe
columnunderload.ThisdataisimportantinthecontextoftheDirectStrengthMethod,
especially for this current effort to extend DSM to members with holes, since elastic
bucklingisusedtopredictthefailuremode(local,distortional,orglobal)andultimate
strength.
5.4.4 Discussion of friction-bearing boundary conditions
The frictionbearing end conditions used in this testing are advantageous because
specimen alignment and preparation can be performed without welding or the use of
grout or hydrostone. The specimens were aligned by hand in the testing machine
without special equipment. However, preparing the specimen ends with a milling
machinecanbetimeconsuming.Further,smalldeviationsinflatnessmaysignificantly
220
impactthetestedresultsandfailuremodes;realcaremustbetakeninthespecimenend
preparation.Finally,lackofapositiveconnectionbetweenspecimenandplatenmakes
itdifficulttoexactlyknowtheboundaryconditions.
Inthisstudy,frictionbetweenthecolumnendsandtheplatenspreventedachange
inshapeofthecrosssectionuptopeakloadinallspecimens,butslippingofthecross
section was observed after peak load. This slipping was signaled by loud metalon
metalpoppingsoundsassociatedwithobservablechangesinthecrosssection(Dof
the flanges, see Figure 5.47 for definition) at the column ends. Also, uplift warping
deformations like those shown in Figure 5.51 occurred in the postpeak range for the
short600S16233columnsexperiencingdistortionaltypefailures.Distortionalbuckling
modes are anticipated to be sensitive to this uplift since they are highly sensitive to
warping deformations. The intermediate length 362S16233 columns experienced a
suddenglobalflexuraltorsionalfailureshortlyafterreachingpeakloadasthetwisting
of the columns overcame the friction between the column ends and the platens. The
frictionbearing end conditions did not allow a detailed study of the global flexural
torsionalpostpeakresponsefortheintermediatelength362S16233columnsandlikely
decreasedtheirultimatestrengths.
Overall, for short and intermediate length column testing focused on local and
distortional buckling modes, the advantages of the simple frictionbearing boundary

conditionsoutweighedthedisadvantages.Propercaremustbetakentoinsuretheends
are milled flat and the platens are level and parallel. For longer column tests, where
221
large torsional rotations must be restrained, the bearing conditions employed here are
notrecommendedforuse.
Flange-lip corner lifts off platen

when large deformations exist
past peak load
Figure5.51Short600S16233columnflangelipcornerliftsoffplatenduringpostpeakportionoftest
222
Chapter 6
Predicting residual stresses and plastic
strains in cold-formed steel members
Thincoldformedsteelmembersbeginasthick,molten,hotsteelslabs.Eachslabis
typically hotrolled, coldreduced, and annealed before coiling and shipping the thin
steel sheet to rollforming producers (US Steel 1985). Once at a plant, the sheet is
unwoundthroughaproductionlineandplasticallyfoldedtoformthefinalshapeofa
structuralmember,asshowninFigure6.1.Thismanufacturingprocessimpartsresidual
stresses and plastic strains through the sheet thickness. These residual stresses and
strains influence the loaddisplacement response and ultimate strength of coldformed
steelmembers.
In previous work a statistical approach was employed to draw conclusions on the
magnitudeanddistributionoflongitudinalresidualstressesusingadatasetofsurface
strainmeasurementscollectedbyresearchersbetween1975and1997(SchaferandPekz
1998). The measured surface strains are converted to residual stresses using Hookes
223
Law and then distributed through the thickness as membrane (constant) and bending
(linearvariation)components.Theseresidualstressdistributionsareaconvenientway
to express the measured residual surface strains, and areconvenient as well for use in
nonlinear finite element analyses, but they are not necessarily consistent with the
underlyingmechanics.
Figure6.1Coldformedsteelrollforming:(left)Sheetcoilentersrollformingline,(right)steelsheetiscold
formedintoCshapecrosssection(photoscourtesyofBradburyGroup).
Plastic bending, followed by elastic springback, creates a nonlinear through
thicknessresidualstressdistribution,inthedirectionofbending,asshowninFigure6.2
(Shanley 1957). The presence of nonlinear residual stress distributions in coldformed
steel members has been confirmed in experiments (Key and Hancock 1993) and in
nonlinearfiniteelementmodelingofpressbrakingsteelsheets(Quachetal.2006).A
closedform analytical prediction method for residual stresses and equivalent plastic
strains from coiling, uncoiling, and mechanical flattening of sheet steel has also been
proposed (Quach et al. 2004 ). The same plastic bending that creates these residual
stressesalsoinitiatesthecoldworkofformingeffect,whereplasticstrainsincreasethe
apparentyieldstressinthesteelsheet(andultimatestrengthinsomecases)(Yu2000).
224
Together,theseresidualstressesandplasticstrainscomprisetheinitialmaterialstateof
acoldformedsteelmember.
Elastic springback
Plastic bending
Elastic springback
Plastic bending
compression
+
tension
Plastic bending
Elastic springback
=
Nonlinear residual
stress distribution
Figure6.2Formingabend:plasticbendingandelasticspringbackofthinsheetsresultsinanonlinear
throughthicknessresidualstressdistribution.
A general method for predicting the manufacturing residual stresses and plastic
strains in coldformed steel members is proposed here. The procedure is founded on

common industry manufacturing practices and basic physical assumptions. The
primarymotivationforthedevelopmentofthismethodistodefinetheinitialstateofa
coldformedsteelmemberforuseinasubsequentnonlinearfiniteelementanalysis.The
derivation of the prediction method is provided for each manufacturing step, and the
predictions are evaluated with measured residual strains from existing experiments.
Theendresultofthemethodisintendedtobeaccessibletoawideaudienceincluding
manufacturers, designengineers, and the academic community. This method also has
the potential to compliment and improve Chapter A7.1.2 of the existing Specification
(AISIS100 2007), which currently allows for an increase in member strength from the
225
coldworkofformingeffectatcrosssectioncorners,butdoesnotdirectlyaccountforthe
influenceofthenonlinearthroughthicknesscornerresidualstressesortheinfluenceof
plastic strains and residual stresses from coiling, uncoiling, and flattening of the sheet
steel.
6.1 Stressstraincoordinatesystemandnotation
Thestressstraincoordinatesystemandgeometricnotationusedintheforthcoming
derivationsaredefinedinFigure6.3.Thexaxisisreferredtoasthetransversedirection
and the zaxis as the longitudinal direction of a structural member. Crosssection
elementsarereferredtoaseithercornersorflats.Thesignconventionforstressand
strainispositivefortensionandnegativeforcompression.
roller dies
rx
sheet steel coil
Forming direction
Elevation View
t
y
rz
z
y
x
z
Section A-A
Figure6.3Stressstraincoordinatesystemasrelatedtothecoilingandcoldformingprocesses.
226
6.2 Predictionmethodassumptions
Thefollowingassumptionsareemployedtodevelopthispredictionmethod:
a. Plane sections remain plane before and after coldforming of the sheet steel. This
assumptionpermitstheuseofbeammechanicstoderivepredictionequations.
b. The sheet thickness t remains constant before and after coldforming of the sheet
steel. A constant sheet thickness is expected after coldbending if the bending is
performed without applied tension (Hill 1950). Crosssection measurements
demonstratemodestsheetthinningatthecorners,wheretinthecornersistypicallyfive
percent less than in the flange and web (Dat 1980). This thinning is ignored here to
simplify the derivations, although a reduced thickness based on the plastic strain
calculationsinSection6.4couldbeusedifahigherlevelofaccuracyisrequired.
c. Thesheetneutralaxisremainsconstantbeforeandaftercrosssectioncoldforming.
Theoreticalmodelsusedinmetalformingtheorydopredictasmallshiftinthethrough
thicknessneutralaxistowardstheinsideofthecornerasthesheetplastifies(Hill1950).
This shift is calculated as six percent of the sheet thickness, t, when assuming a
centerline corner radius, rz, of 2.5t. A neutral axis shift of similar magnitude has been
observedinthenonlinearfiniteelementmodelresultsforthinpressbrakedsteelsheets
(Quachetal.2006).Thissmallshiftisignoredheretosimplifythederivations.
227
d.Thesteelstressstraincurveisassumedaselasticperfectlyplastic whencalculating
residual stresses. More detailed stressstrain models that include hardening are
obviously possible, but a basic model is chosen to simplify the derivations. The
implication of this assumption is that the residual stresses may be underestimated,
especially in corner regions where the sheet has yielded completely through the
thickness.
e. Plane strain behavior is assumed to exist during coiling, uncoiling, and flattening
(x=0) andduringcrosssectioncoldforming (z=0).
f. The steel sheet is fed from the top of the coil into the rollforming bed as shown in
Figure6.4a.Thisassumptionisconsistentwithmeasuredbendingresidualstressdata
(see Section 6.6) and manufacturing setups suggested by rollforming equipment
suppliers(Figure6.1).TheauthordidobservethealternativesetupinFigure6.4b(sheet
steel unrolling from the bottom of the coil) at a rollforming plant, suggesting that the
directionofuncoilingisasourceofvariabilityinmeasuredresidualstressdata.
228
Roll-forming bed
(a)
Sheet has residual

CONCAVE curvature
coming off the coil
(b)
Sheet has residual

CONVEX curvature
coming off the coil
Figure6.4Rollformingsetupwithsheetcoilfedfromthe(a)topofthecoiland(b)bottomofcoil.The
orientationofthecoilwithreferencetotherollformingbedinfluencesthedirectionofthecoilingresidual
stresses.
g. Membrane residual stresses are zero. Membrane residual stresses have been
measuredbyseveralresearchers(Ingvarsson1975;Dat1980;WengandPekz1990;De
BatistaandRodrigues1992;Kwon1992;Bernard1993;KeyandHancock1993),although
themagnitudesaresmallrelativetobendingresidualstresses(seeTable6.1).Membrane
residual stresses are experimentally determined by averaging the measured surface
strains on the two faces of a thin steel sheet. Given the variability inherent in these
measurementsitisdifficulttoknowiftheresultingmembranestresses(strains)arereal
orsimplyunavoidableexperimentalerror.
229
6.3 Derivation of the residual stress prediction

method
The prediction method proposed here assumes that two manufacturing processes
contributetothethroughthicknessresidualstressesincoldformedsteelmembers:(1)
sheet coiling, uncoiling, and flattening, and (2) crosssection rollforming. Algebraic
equations for predicting the throughthickness residual stress and effective plastic
strains in corners and flats are derived here and then summarized in flowcharts in
Figure6.13andFigure6.17.
6.3.1 Residual stresses from sheet coiling, uncoiling, and

flattening
Coilingthesheetsteelafterannealingandgalvanizing,butpriortoshipment,may
yield the steel if the virgin yield strain, yield, is exceeded. If plastic deformation does
occur, a residual curvature will exist in the sheet as it is uncoiled. This residual
curvatureislockedintoastructuralmemberresultinginlongitudinalresidualstresses
as the sheet is flattened by the rollformers. This process of coiling, uncoiling with
residualcurvature,andflatteningisdescribedinFigure6.5.
230
DETAIL A
Uncoiled with residual

curvature
Coiling
Flattened as sheet enters

the roll-formers
Change in curvature locks in

bending residual stresses in
final member
DETAIL A
Figure6.5Coilingofthesteelsheetmayresultinresidualcurvaturewhichresultsinbendingresidual
stressesasthesheetisflattened.
6.3.1.1
Coiling
Thethroughthicknessstraininducedfromcoilingisrelatedtotheradiallocationof
thesheetinthecoilrx,withthewellknownrelationshipfrombeammechanics:
z
y
1
.
rx
(6.1)
zistheengineeringstrainthroughthethicknessyinthecoiling(longitudinal)direction
z.yvariesfromt/2tot/2,wheretisthesheetthickness.Theradiusassociatedwiththe
elasticplasticthresholdinitiatingthroughthicknessyieldingfromcoiling,rep,isderived
bysubstitutingz=yieldandy=t/2(outerfiberstrain)intoEq.(6.1):
rep =
t
2 yield
(6.2)
When the coil radius rx is greater than rep the sheet steel experiences only elastic
deformation on the coil. For sheetsteel rolled to a coil radius rx less than rep, through
thicknessyieldingwilloccurasshowninFigure6.6.
231
+yield
-yield
Figure6.6Longitudinalresidualstressdistributionfromcoiling.
Whenrx<rep,thedepthoftheelasticcorecisdefinedas:
6.3.1.2
c = 2rx yield t .
(6.3)
Uncoiling
As the yielded sheet is uncoiled in preparation for the rollforming line, the sheet
steelspringsbackelasticallyresultinginachangeinthethroughthicknessstress.This
stressdistributionisdeterminedbyfirstcalculatingtheplasticcoilingmoment
coil
x
t 2 1
2
= yield (rx yield ) ,
2 3
(6.4)
and then applying an opposing moment elastically to simulate the removal of the
imposedradialdisplacement
6.3.1.3
uncoil
z
12 M xcoil y
=
.
t3
(6.5)
Flattening
Afterthesheethasbeenunrolled,apermanentradiusofcurvaturewillstillexistifrx
waslessthanreponthecoil.Thispermanentradiusis
232
uncoil
rx
1
1 M xcoil
rx
EI
(6.6)
Steelsheetwithpermanentcurvaturefromcoilingispressedflatasthesheetentersthe
rollforming line. The longitudinal stresses resulting from flattening the sheet are
simply
zflatten = E
6.3.1.4
y
uncoil
x
(6.7)
Residualstressdistribution
The total throughthickness longitudinal residual stress distribution due to coiling,
uncoiling,andflatteningispresentedinFigure6.7.
y
+yield
uncoil
z
flatten
z
=
z
-yield
Coil
Uncoil
Flatten
Residual Stress
Figure6.7Predictedlongitudinalresidualstressdistributionfromcoiling,uncoiling,andflatteningofa
steelsheet.
Theresultingresidualstress,z, isselfequilibratingforaxialforcethroughthethickness
butcausesaresiduallongitudinalmoment.Section6.6comparesthestressescausedby
thismomentwithsurfacestrains(stresses)measuredinexperiments.
The longitudinal residual stresses also will create transverse stresses across the
width of the coil, assuming plane strain conditions for an infinitely wide sheet.
233
Supportingtheplanestrainassumptionistheobservationthatwhiletheactualwidthof
the sheet is finite, it remains several orders of magnitude greater than the sheet
thickness. Under this assumption, and further assuming only elastic stresses, the
transversestressesare:
x = ( coil + uncoil +
z
flatten
z
) .
(6.8)
Poissons ratio, , is assumed here as 0.30 for steel deformed elastically. The through
thickness deformation from the uncoiling and flattening components will occur
elastically, and the coiling component will be at least partially elastic through the
thicknessfortherangeofsheetthicknessescommoninindustry.
6.3.2 Residual stresses from cross-section roll-forming
A set of algebraic equations is derived here to predict the transverse and
longitudinal residual stresses created by rollforming a crosssection. Rollforming

residualstressesarecumulativewiththecoilingresidualstressesderivedinSection6.3.1
andprovideacompletepredictionoftheinitialstressstateofthemembercrosssection.
Therollformingresidualstressesareassumedtoexistonlyatthelocationoftheformed
corners, between the roller die reactions, as shown in Figure 6.8. Some yielding is
expected to occur outside of the roller reactions as the stress distribution transitions
fromfullyplastictofullyelastic;however,thistransitionareaisnotconsideredhereto
simplifythederivation.
234
Assume bend is fully plastic

between roller dies
Roller Die (Typ.)
Figure6.8Coldformingofasteelsheet.
The engineering strain in the steel sheet, x, and the bend radius, rz, are related for
bothsmallandlargedeformationswiththestraincurvaturerelationship
1 x
=
rz
y
(6.9)
Thisgeometricrelationshipisvalidforelasticandplasticbendingofthesteelsheet.For
the small bend radii common in the coldformed steel industry (rz =2t to 8t), the steel
sheetyieldsthroughitsthicknessduringthecoldformingprocess.Thesteelsheetwill
reachthefullyplasticstressstateshowninFigure6.9asthecornerapproachesitsfinal
manufacturedradius.
y
-yield
+yield
Figure6.9Fullyplastictransversestressstatefromcoldforming.
235
After the sheet becomes fully plastic through its thickness, the engineering strain
continuestoincreaseastheradiusdecreases.Whenthefinalbendradiusisreachedand
theimposedradialdisplacementisremoved,anelasticspringbackoccursthatelastically
unloadsthecorner(seeFigure6.2).Thechangeinstressthroughthethicknessfromthis
elasticreboundisderivedwiththeplasticmomentforcecoupleshowninFigure6.10.
y
+yield
Fp
t/2
Fp
-yield
Figure6.10Forcecouple(Fpt)appliedtosimulatetheelasticspringbackofthesteelsheetafterthe
imposedradialdeformationisremoved.
Theplasticmomentiscalculatedwiththeequation
bend
z
2
t yield 1 t t yield t
= FP =
=
2
2
2
4
(6.10)
whichisthenappliedelasticallythroughthethicknesstosimulatethestressdistribution
fromelasticreboundofthesheetsteel:
xrebound
yield t 2
y
3 y
M zbend y 4
=
=
= yield .
1
t
I
1 t 3
12
The final transverse stress state is the summation of the fully plastic stress
(6.11)
distributionthroughthethicknessandtheunloadingstressfromtheelasticspringback
ofthecornerasshowninFigure6.11,wherexisthetransverseresidualstressthrough
236
thethicknessfromthecoldformingofthecorner.Thisstressisnonlinearthroughthe
thickness and is selfequilibrating, meaning that axial and bending sectional forces are
absentinthexdirectionafterforming.
y
-yield
+1.5yield
Plastic Bending
xbend
+yield
+0.5yield
-yield
+yield
-0.5yield
-1.5yield
Elastic Springback
Transverse Residual Stress
xrebound
Figure6.11Coldformingofasteelsheetoccursasplasticbendingandelasticspringback,resultinginaself
equilibratingtransverseresidualstress.
Thetransverseresidualstresseswillcreatestressinthelongitudinaldirectionduetothe
assumedplanestrainconditions(seeSection6.2):
z = x .
(6.12)
ThePoissonsratio,,isassumedas0.30forsteeldeformedelasticallyand0.50forfully
plastic deformation. The longitudinal residual stresses through the thickness, z, are
determinedbasedontheseassumptionsasshowninFigure6.12.Longitudinalresidual
stress,z, isselfequilibratingforaxialforcethroughthethicknessbutcausesaresidual
longitudinal moment. This moment is hypothesized to contribute to the observed
longitudinal residual strains measured in experiments (refer to Section 6.6 for a
comparisonofthispredictiontoactualmeasurements).
237
-yield
0.50
+
+yield
Plastic Bending
plastic xbend
-0.05yield
+1.5yield
0.30
=
-1.5yield
Elastic Springback
elastic xrebound
-0.50yield
+0.50yield
z
+0.05yield
Longitudinal Residual Stress
Figure6.12Plasticbendingandelasticspringbackfromcoldforminginthetransversedirectionresultin
longitudinalresidualstressesbecauseoftheplanestrainconditions.
A flowchart summarizing the proposed prediction method for residual stresses in
rollformed members is provided in Figure 6.13. Figure 6.13 explicitly demonstrates

how coiling, uncoiling, flattening, and rollforming contribute to the residual stresses
lockedintothecrosssectionduringmanufacturing.
238
Start
Flat
Corner
Flat or Corner?
End
No
t
2 yield
t
rx >
2 yield
rx
Yielding on the Coil?
No residual
stresses!
Yes, yields on coil

No, remains elastic
c
t
y
2
2
c
c
< y<
2
2
t
c
y
2
2
+ yield
Yes
y
E
rx
zcoil =
Sheet Coiling
yield
c = 2rx yield
Sheet Coiling
M xcoil y
I
zuncoil =
t 2 1
2
1
M xcoil = yield (rx yield ) I = 1 t 3
12
2 3
xuncoil = elastic zuncoil
zflatten = E
uncoil
End
Yes
elastic = 0.30
xcoil = elastic zcoil
Sheet Uncoiling
rx
Sheet Flattening
y
rxuncoil
1
1 M xcoil
rx
EI
Sheet Flattening
elastic = 0.30
yield
0 y
t
2
+ yield t y 0
(For rz<8t)
zbend = plastic xbend
xrebound =
Sheet Uncoiling
elastic = 0.30
xflatten = elastic zflatten
bend
x
No
3 yield y
t
zrebound = elastic xrebound
Corner Bending
plastic = 0.50
t
t
y
2
2
Corner Rebound
End
elastic = 0.30
Figure6.13Flowchartsummarizingthepredictionmethodforresidualstressesinrollformedmembers.
239
6.4 Derivation of effective plastic strain prediction

method
In the method proposed here, plastic strains occur from sheet coiling and cold
forming, and together with residual stresses describe the initial material state of the
member.Thegeneralstateofplasticstrainatapointcanbequantifiedbyusingthevon
Misesyieldcriterionextendedtoplasticdeformations(ChenandHan1988):
p =
2
3
2
1
+ 2 + 3
2
(6.13)
wherepistheeffectiveplasticstrain,and1, 2,and 3aretheprincipalstrains.Allofthe

strainsaretruestrains,whichmaybecalculatedfromtheengineeringstrainsvia:
1 = ln(1 + x ) , 2 = ln(1 + y ) , 3 = ln(1 + z ) ,
(6.14)
where x, y, zareintheCartesiancoordinatesystem(Figure6.3)andx,y,ziscoincident
withtheprincipaldirections.Truestrainsareemployedinsteadofengineeringstrainsto
accommodate the large deformations from plastic bending. Also, from a practical
standpoint,nonlinearFEcodessuchasABAQUS(ABAQUS2007a)requiretheengineer
toprovidetruestress,truestraininformation(aslargedeformationtheoryisemployed).
The steel sheet is assumed to remain incompressible while experiencing plastic
deformations,thereforewhencalculatingp
1 + 2 + 3 = 0 .
240
(6.15)
6.4.1 Effective plastic strain from sheet coiling
Engineeringplasticstrains,asshowninFigure6.14,accumulateduringthecoilingof
sheetsteelifthecoilingradiusrxislessthantheelasticplasticthresholdrep.
y
zp
Figure6.14Plasticstraindistributionfromsheetcoilingwitharadiuslessthanelasticplasticthresholdrep.
Theengineeringplasticstraindistributionfromcoilingis:
zp =
zp =
y
c
yield , y
rx
2
y
c
yield , y
rx
2
(6.16)
zp = 0 otherwise ,
wheretheelasticcore,c,isdefinedinEq.(6.3).Planestrainconditionsresultin1=0,and
2=3 via the incompressibility assumption of Eq. (6.15). Further, the Cartesian
coordinate system is coincident with the principal axes, resulting in the following true
principalplasticstrains:
1 = 0 , 2 = ln(1 + zp ) , 3 = ln(1 + zp ) .
(6.17)
Substituting the principal strains into Eq. (6.13) and simplifying leads to the through
thicknesseffectiveplasticstrainfromcoiling
241
pcoiling =
2
ln (1 + zp ).
3
(6.18)
This plastic strain distribution, depicted in Figure 6.15, will exist at all locations in the
crosssection(cornersandflats)whenrxislessthantheelasticplasticthresholdrep.
pcoiling
y
Figure6.15Effectiveplasticstraininacoldformedsteelmemberfromsheetcoilingwhentheradiusrxis
lessthantheelasticplasticthresholdrep.
Theplasticstrainfromcoiling,pcoiling,willgenerallybemuchsmallerinmagnitudethan
theplasticstrainfromcrosssectioncoldforming,pbend,asdiscussedinfollowingsection.
6.4.2 Effective plastic strain from cross-section coldforming
Largetransverseplasticstrainsoccurthroughthethicknessofathinsteelsheetwhen
the sheet is permanently bent. The engineering plastic strain distribution from cold
formingisdescribedvia
xp =
y
rz
(6.19)
which assumes that the elastic core at the center of the sheet is infinitesimally small.
Thisassumptionisconsistentwiththesmallbendradiicommoninindustry(see6.3.2).
242
PlanestrainconditionsandEq.(6.15)resultin3=0,2=1.Physicallytheseconditions
implythatthesheetwillexperiencesomethinningatthelocationofcoldforming(see
Section6.2),butthetendencytoplasticallyshortenlongitudinallywillberesistedbythe
adjacent undeformed portion of the crosssection. As before, the Cartesian coordinate
systemiscoincidentwiththeprincipalaxes,resultinginthefollowingplasticprincipal
strains:
1 = ln(1 + xp ) , 2 = ln(1 + xp ) , 3 = 0 .
(6.20)
Substituting for the principal strains and simplifying, the effective plastic strain at a
coldformedcorneris:
bend
=
p
2
ln (1 + xp )
3
(6.21)
ThiseffectiveplasticstraindistributionisshowninFigure 6.16.Thedistributionexists
only at the coldbent locations in a crosssection and should be added to the coiling
plasticstraindistributioninFigure6.15.
y
Figure6.16EffectivevonMisestrueplasticstrainatthelocationofcoldformingofasteelsheet.
A flowchart summarizing the prediction method for effective plastic strains in roll
formedmembersisprovidedinFigure6.17.
243
Start
Flat
Corner
Flat or Corner?
End
No
No equivalent
plastic strains!
t
2 yield
t
rx >
2 yield
rx
No
No, remains elastic
Yes
y
yield
rx
Sheet Coiling
zp =
c
2
c
y
2
y
y
yield
rx
Sheet Coiling
c = 2rx yield
otherwise
bend
=
p
Yes
2
ln (1 + zp )
3
pcoiling =
End
Yes, yields on coil
2
ln (1 + xp )
3
xp =
t
t
y
2
2
Corner Bending
y
rz
End
Figure6.17Flowchartsummarizingthepredictionmethodforeffectiveplasticstrainsinrollformed
members
6.5 Employingthepredictionmethodinpractice:
quantifyingthecoilradiusinfluence
The residual stress and plastic strain distributions derived for crosssection cold
forming(Sections6.3.2and6.4.2)arestraightforwardtocalculateiftheyieldstress,yield,
andthickness,t,ofthesheetsteelareknown.Thecoilingresidualstressesandplastic
strains are more difficult to calculate because the coil radius coinciding with the as
formed member, i.e., the radial location of the sheet, rx, is almost always unknown in
244
practice.However,rxcanbederivedinanaveragesensethough,sincetherangeofinner
and outer coil radii are known and the probability that a structural member will be
manufacturedfromacertainrxcanbequantified.
The relationship between coil radius, rx, and corresponding linear location S of the
sheetwithinthecoilcanbedescribedusingArchimedesspiral(CRC2003)
S=
(r
t
2
rinner .
(6.22)
Thespiralmaintainsaconstantpitchwithvaryingradii,wherethepitchisthethickness
ofthesteelsheet,t,asshowninFigure6.18,Listhetotallengthofsheetinthecoil,and
rinner and router are the inside and outside coil radius, respectively. Asshipped outer coil
radii range from 24 in. to 36 in. and inner coil radii range from 10 in. to 12 in. These
rangesweredeterminedbytheauthorduringavisittoalocalrollformingplant.
S
rx
t
Start
S =0
End
S=L
Figure6.18Coilcoordinatesystemandnotation.
Archimedes spiral is used to describe the probability that the steel sheet will come
fromacertainrangeofradiallocationsinthecoil.Thecumulativedistributionfunction
245
(CDF),FR(rx)=probabilitythattheradiusislessthanrx,isobtainedbynormalizingSby
L
2
rx2 rinner
S
= FR (rx ) = 2
.
2
L
router rinner
(6.23)
Theprobabilitydensityfunction(PDF)ofrxiscalculatedbytakingthederivativeofFR(rx)
f R (rx ) =
2rx
dFR (rx )
= 2
.
2
drx
router rinner
(6.24)
Themeanvalueoftheradiallocationforagiveninnerandoutercoilradiiis
rx =
router
rinner
r r
2
f R (rx )rx drx = rinner + router inner outer
3
rinner + router
(6.25)
Thevarianceoftheradiallocationis
s R2 =
router
rinner
f R ( rx )(rx rx ) 2 dr =
(router rinner ) .
1 2
2
router + 4router rinner + rinner
18
(router + rinner )2
(6.26)
Thesestatisticsforrxcanthenbeusedwiththepredictionmethodforcoiling,uncoiling,
andflatteningresidualstressesandplasticstrainsdescribedinSections6.3.1and6.4.1.
Figure6.19summarizestheinfluenceofsheetthicknessandvirginyieldstressonthe
longitudinalresidualstressdistributionsinflatsandcorners.(Themethodproposedin
this chapter provides residual stresses and strains for the entire member, only the
longitudinal residual stresses are shown in Figure 6.19.) The solid lines in Figure 6.19
arecalculatedusingthemeanvalue, rx =18.7in,fromEq.(6.25)assumingrinner=12in.and
router=24 in. The distributions with the dashed lines are calculated with rx sR, where
sR=3.4 in. is calculated with Eq. (6.26). The residual stresses are nonlinear through the
thicknessandhavedifferentshapesforflatsandcorners.Thestressmagnitudesatthe
246
outer fibers increase for thicker sheets and lower yield stresses. The accuracy of the
linearbendingresidual stressmodelcommonlyemployedinfiniteelementanalysesis
perhaps sufficient when yield stress is low and thickness is high (relatively), but for
typicalthicknesses(0.0346in.to0.0713in.)andyieldstress(50ksi)theassumptionofa
linear longitudinal stress distribution is not consistent with the mechanicsbased
predictionsinFigure6.19.
Longitudinal Residual Stresses

yield=50 ksi
yield=30 ksi
t/2
flat
corner
flat
yield=80 ksi
corner
flat
corner
t=0.0346 in.
-t/2
-yield 0
yield
t/2
t=0.0713 in.
-t/2
-yield 0
yield
t/2
t=0.1017
-t/2
-yield 0
yield
Positive stress is tension, negative stress is compression

Figure6.19Influenceofsheetthicknessandyieldstressonthroughthicknesslongitudinalresidualstresses
(zdirection,solidlinesarepredictionsformeancoilradius,dashedlinesformeanradius+/onestandard
deviation).
247
6.6 Comparisonofpredictionmethodtomeasured
residualstresses
The flat and corner residual surface strain measurements from 18 rollformed
specimens are used to evaluate the proposed residual stress prediction method. The
prediction methodprovidesthecompletethroughthicknesslongitudinalstrain(stress)
distributioniftheradiallocationinthecoilfromwhichthespecimenoriginatedinthe
coil, rx, is known. Since the radial coil location of the 18 specimens is unknown, rx is
statisticallyestimatedforeachspecimenusingthecoilradiusthatbestfitsthepredicted
surface strains to the measured surface strains from a specimen crosssection (for both
corners and flats). Once the best fit radial locations have been calculated, they are
examined to determine if their magnitude is rational when compared to typical inner
andouterdimensionsofasheetcoil.Althoughthiscomparisononlyprovidesapartial
evaluation of the prediction method, it is as far as one can go with the available data.
Qualitatively the prediction method is consistent with the more detailed through
thicknessfindings(KeyandHancock1993;Quachetal.2006).
6.6.1 Measurement statistics
The mean and standard deviation of the residual stresses for the 18 rollformed
specimens used in this comparison are provided in Table 6.1. Positive membrane
stresses are tensile stresses and positive bending stresses cause tension at y=t/2 (see
Figure 6.3 for coordinate system). The statistics demonstrate that both membrane and
bending residual stress measurements are highly variable and that the membrane
248
stresses are small relative to the steel yield stresses. Details on the residual stress
measurements for each of the 18 specimens are described in a previous research
progressreport(MoenandSchafer2007b).
Table6.1Statisticsoftheresidualstressesinrollformedmembers
Element
Corners
Flats
Residual stress as %yield

Membrane
Bending
Mean
STDEV
Mean
STDEV
5.7
10.1
32.0
23.8
1.8
10.7
25.2
20.7
No. of
Samples
23
120
6.6.2 Mean-squared error (MSE) estimate of radial location
Toexplorethevalidityofthepredictionmethod,theflatandcornerresidualstress
measurements from the 18 specimens are used to estimate the radial location rx from
which each specimen originated. These estimated radial locations are then used to
calculate the difference between the predicted and measured longitudinal residual
stresses.
6.6.2.1
MSEminimization
The location of the specimen in the coil, rx, is estimated by minimizing the sum of the
meansquared errors (MSE) for the p=1,2,nq measurements taken around the cross
sectionoftheq=1,2,,18specimens
2
measured
predicted
pq
pq
.
rx ,q = arg min
p =1
yield , pq
nq
Bothcornerandflatmeasurementsareincludedintheminimization.
249
(6.27)
6.6.2.2
Bending component of longitudinal residual stress
distribution
Thebendingcomponentofthepredictedresidualstressdistributionmustbeisolated
tocompare withthemeasuredvalues.Thetotalpredictedlongitudinal residualstress

distribution in the flats and corners of each crosssection is integrated to calculate the
sectionalmomentthroughthethickness
t
2
t
M x = z ydy .
(6.28)
Mxisthenconvertedintoapredictedouterfiberbendingresidualstresswhichcanthen
bedirectlycomparedtothemeasurements
6.6.2.3
pqpredicted
t
M x
2
=
.
I
(6.29)
EstimatedcoilradiiusingMSE
Figure 6.20 demonstrates the meansquared error results for de M. Batista and
Rodrigues Specimen CP1 (De Batista and Rodrigues 1992). The radial location that
minimizesthepredictionerroris1.60rinnerinthiscase,andissummarizedinTable6.2for
all18rollformedspecimensconsidered.Theestimatedradiallocationsfallwithinthe
rangeofinnerandoutercoilradiiassumedintheprediction(rinner to2.40rinner)exceptfor
Dat RFC13 which is slightly outside the range at 2.45rinner. The MSE radial location
cannotbedeterminedinthethreeBernardspecimens(Bernard1993)sincethebending
250
residual stresses in the flats are predicted to be zero. These three specimens are cold
formedsteeldeckingwithathinsheetthicknesstrangingfrom0.022in.0.0400in.anda
relativelyhighyieldstressyieldrangingfrom87ksito94ksi.Inthiscase,thecoilingand
uncoiling of the steel sheet will occur elastically as demonstrated in Figure 6.19.
MeasuredbendingresidualstressmagnitudesintheflatsoftheBernardspecimensare
onaverage0.03yieldwhichisconsistentwiththepredictionmethod.
8
7
6
MSE
5
4
3
2
1
0
1.5
2.5
rx /rinner
3.5
Figure6.20ThemeansquarederrorofthepredictedandmeasuredbendingresidualstressesfordeM.
BatistaandRodrigues(DeBatistaandRodrigues1992),SpecimenCP1isminimizedwhenrx=1.60rinner.
251

Table6.2Radiallocationinthecoilthatminimizesthesumofthemeansquarepredictionerrorforroll
formedmembers
Researcher
rx estimate
Specimen
de M. Batista and Rodrigues (1992)

de M. Batista and Rodrigues (1992)
Weng and Pekz (1990)
Dat (1980)
Dat (1980)
Bernard (1993)
Bernard (1993)
Bernard (1993)
Abdel-Rahman and Siva (1997)
rinner=10 in., router=24 in.
CP2
CP1
RFC13
RFC14
R13
R14
P3300
P4100
DC-12
DC-14
RFC14
RFC13
Bondek 1
Bondek 2
Condeck HP
Type A - Spec 1
Type A - Spec 2
Type B - Spec 1
N/A coiling residual stresses are predicted as zero
in.
12.0
16.0
18.0
11.0
14.5
13.0
19.5
15.0
23.0
16.0
20.0
24.5
N/A
N/A
N/A
16
16
13
6.6.3 Statistical variations between measurements and

predictions
The predicted radial locations in Table 6.2 are now used to calculate the statistical
variations between the experiments and predictions. The bending residual stresses in
the 18 rollformed members are calculated using the MSEpredicted radial location rx
with the residual stress prediction method summarized in Figure 6.13. The bending
component of the residual stress prediction is then obtained with Eq. (6.29). The
difference between the predicted and measured residual bending stresses, epq, for the
p=1,2,,nmeasurementstakenaroundthecrosssectionoftheq=1,2,,18specimensis
calculatedas
measured
predicted
pq
pq
e pq =
yield , pq
252
(6.30)
The error histogram for the flat crosssectional elements in Figure 6.21a demonstrates
that the mean difference e is near zero with a standard deviation se=0.15yield. The
scattergram in Figure 6.21b demonstrates the strength of the correlation between the
measurements and predictions in the flats; the solid regression line passes nearly
through zero (yintercept=0.05yield) and has nearly a unit slope (m=0.92). Also, the
majorityofthedatalieswithinonestandarddeviationoftheestimate,denotedasthe
dashedlinesinthefigure.Itisconcludedthatthepredictionmethodisconsistentwith
themeasureddataintheflats.
The corner element error histogram in Figure 6.22a shows a negative bias of e=
0.16yield meaning that the predicted residual stresses are generally higher than the
measured values. The standard deviation of the error is large (se=0.19yield) but is less
than the standard deviation of the corner residual stress measurements in Table 6.1
(sm=0.24yield). This demonstrates a greater match between the measurements and
predictions,althoughmorecornerresidualstressmeasurementsareneededtoimprove
the strength of this comparison. The scattergram in Figure 6.22b highlights the
variability in the measured corner data, especially in the region corresponding to
predicted=0.4yield,wherebendingresidualstresses(strains)varyfrom0to0.7yield.
253
40
e = 0.03 yield
30
e = 0.15 yield
0.7
0.6
0.5
/ yield
25
measured
20
15
0.4
0.3
0.2
Observations
0.8
35
0.1
10
5
0
-1
-0.1
-0.5
0
(
measured
0.5
-0.2
0.1
0.2
predicted
)/ yield
0.3
0.4
predicted
(a)
0.5
0.6
0.7
0.8
/ yield
(b)
Figure6.21(a)Histogramand(b)scattergramofbendingresidualstresspredictionerror(flatcross
sectionalelements)for18rollformedspecimens.
40
0.8
0.7
e = 0.16 yield
25
e = 0.19 yield
0.6
0.5
/ yield
30
measured
20
15
0.4
0.3
0.2
Observations
35
0.1
10
5
0
-1
-0.1
-0.5
0
(
measured
0.5
-0.2
predicted
)/ yield
(a)
0.1
0.2
0.3
0.4
predicted
0.5
0.6
0.7
0.8
/ yield
(b)
Figure6.22(a)Histogramand(b)scattergramofbendingresidualstresspredictionerror(cornercross
sectionalelements)for18rollformedspecimens.
6.7 Discussion
The residual stresses and strains predicted with this method (Section 6.3 for stress,
Section 6.4 for strain) form the initial material state in thecrosssection. In design, this
initialmaterialstateissometimesconsideredthroughthesocalledcoldworkofforming
254
effect, where the yield stress of the material is increased above the virgin yield stress,
yield, to account for the working of the corners. For onedimensional stressstrain this
concept is expressed as shown in Figure 6.23, where working the corners results in a
residual plastic strain, p, such that when the section is reloaded the stress at which
yieldingreinitiates,ey,isgreaterthanthevirginyieldstress,yield.Ifnoresidualstresses
existed the apparent increase in the yield stress from yield to ey can be significant.
However, as Figure 6.19 illustrates, working the corners also contributes to residual
stresses,ers,andtheseresidualstressesmaydecreasetheapparentyieldstress.
Thepredictionmethodpresentedhereinprovidesamorenuancedunderstandingof
thecoldworkofformingeffects.Theresidualplasticstrainsmayincreasetheapparent
yield stress, but those strains vary through the thickness and have contributions from
both transverse and longitudinal strains. Further, residual stresses follow their own
relatively complicated distribution through the thickness. In a multiaxial stress state
using the von Mises yield criterion, Figure 6.23 is enforced for the effective stress
effectivestrainpairfor everypoint inthecrosssection.Asaresult,theapparentyield
stressuponloadingvariesthroughthethicknessandisinfluencedbyboththeresidual
stresses and strains. Even under simple loading conditions (e.g., compression) a cold
formed member undergoes plate bending well in advance of collapse, so the strains
demandedofthematerialalsovarythroughthethicknessandaroundthecrosssection.
Whileitisindeedpossibletomodelsucheffectsinafiniteelementanalysis,assuming
these effects can be collapsed into a generic increase in the yield stress for the entire
sectionasiscurrentlydoneindesignwouldseemtobeanoversimplification.
255

Virgin steel
ey
yield
Coldformed
steel
ers
Apparent
yield
stress
Effective residual
stress
yield + p
Figure6.23Definitionofapparentyieldstress,effectiveresidualstress,andeffectiveplasticstrainasrelated
toauniaxialtensilecoupontest.
Implementationoftheresidualstressesandinitialplasticstrainsintoacommercial
finite element program such as ABAQUS, where the member is modeled using shell
elements, is relatively straightforward. The number of throughthickness section
(integration) points must be increased to resolve the nonlinear throughthickness
residualstressandstraindistributions.Theresidualstressesandstrainspredictedherein
can be relatively large. Further, conventional loading (e.g., compression, majoraxis
bending)maycauseloadingorunloadingoftheseinitialstressesatagivenpointinthe
crosssection. As a result, the hardening rule: isotropic, kinematic, or mixed can have
practicaldifferencesintheobservedresponseevenwhentheappliedloadsthemselves
arenotreversing.
Forthissituation,kinematichardening,whichapproximatestheBauschingereffect,
providesamoreconservativemodeloftheanticipatedmaterialbehaviorthanisotropic
hardening. However, to model kinematic hardening the location of the center of the
yieldsurfaceinstressspace(alsoknownasthebackstress)mustbedeterminedforeach
256
point in the crosssection at the end of the manufacturing process. This location is a
functionoftheextentofyielding,intheexampleofFigure6.23,thebackstresswouldbe
the 1, 2, 3 triad that results in the effective stress increasing from yield to ey.
Unfortunately, the elasticperfectly plastic assumption used to predict residual stresses
herein does not directly allow for the calculation of the backstress. However, the
effective plastic strain may be used to approximate the backstress as provided in
AppendixG.Furtherexaminationofthepredictedresidualstressandstrainsandtheir
impact on the peak strength and collapse response of coldformed steel members in
nonlinearfiniteelementanalysisiscurrentlyunderway,includingtheworkpresentedin
Section7.2.
6.8 Acknowledgements
The development of this residual stress prediction method would not have been
possible without accurate information about the manufacturing process of sheet steel
coilsandcoldformedsteelmembers.ThankstoClarkWesternBuildingSystems,Mittal
Steel USA, and the ColdFormed Steel Engineers Institute (CFSEI) for their important
contributions to this research, especially Bill Craig, Ken Curtis, Tom Lemler, Joe
Wellinghoff,EzioDefrancesco,JeanFraser,NarayanPottore,andDonAllen.
257
Chapter 7
Nonlinear finite element modeling of
cold-formed steel structural members
Commercial finite element programs provide a means for realistic collapse

simulation of coldformed steel structural members. Thin shell finite element
formulationsprovidedinABAQUS(e.g.,theS9R5elementdiscussedinChapter2)are
designed to capture the sharp folds and throughthickness yielding characteristic of
coldformed steel beams and columns at their ultimate limit state. Robust solution
algorithms are available to predict unstable, geometrically nonlinear collapse. The
abilitytodefinetheinitialstateofamember,includinggeometricimperfectionsandthe
effectsofresidualstressesandinitialplasticstrainsfromthemanufacturingprocess,is
alsofeasible.Caremustbetakenthoughwithcomputationalresultssincetheyareoften
sensitivetomodelinginputsandassumptions.Itisprudenttostudythesesensitivities
andvalidateaspecificmodelingprotocolwithknownexperimentresultsbeforetrusting
theprotocoltoconsistentlyproducephysicallyrealisticresults.
258
This chapter begins with preliminary nonlinear finite element studies of stiffened
elements(i.e.,asimplysupportedplate,seeFigure3.1fordefinition)withandwithout
holes, which are designed to gain experience with available ABAQUS nonlinear finite
elementsolutionmethods.Theinfluenceofimperfectionsonstiffenedelementsisalso
evaluated, and the throughthickness yielding patterns of a stiffened element (i.e.,
effectivewidth)withandwithoutaholearecompared.Theconclusionsreachedfrom
thispreliminaryworkareusedtoguidethedevelopmentandvalidationofanonlinear
finite element modeling protocol which is needed in Chapter 8 to explore the Direct
StrengthMethodformemberswithholes.
7.1 PreliminarynonlinearFEstudies
Exploratory nonlinear finite element studies are conducted in this section to gain
experience with ABAQUS input parameters and solution controls. All studies are
focused on the simulation of a stiffened element loaded unixaxially to collapse, and
specificattentionispaidtothemodelingofastiffenedelementwithahole.Experience
gainedfromsolvingthishighlynonlinearproblemwillbevaluablewhenimplementing
the larger simulation studies on full coldformed steel members with holes in Section
7.2.
7.1.1 Finite element modeling definitions
The stiffened element is modeled with ABAQUS S9R5 thin shell finite elements,
where the plate dimensions are h=3.4 in. and L=27.2 in. (see Figure 3.2 for plate
dimension definitions) and the plate thickness t is 0.0346 in. (These dimensions are
259
specifically chosen to be consistent with the flat web width and thickness of an SSMA
362S16233 structural stud.) Coldformed steel material properties are assumed as
E=29500 ksi and =0.30. Material nonlinearity is simulated in ABAQUS with classical
metal plasticity theory, including the assumption of a von Mises yield surface and
isotropic hardening behavior. The nonlinear plastic portion of the true stressstrain
curveshowninFigure7.1wasobtainedfromatensilecoupontest(Yu2005)andinput
intoABAQUStodefinethelimitsofthevonMisesyieldsurfaceasafunctionofplastic
strain.
70
60
true plastic stress, ksi
50
40
True stress True strain

ksi
33
0.0000
34.31
0.0016
42.14
0.0259
51.35
0.0700
60.86
0.1407
65.5
0.1988
30
20
10
0.02
0.04
0.06
0.08
0.1
0.12
true strain
0.14
0.16
0.18
0.2
Figure7.1Truestressstraincurvederivedfromatensilecoupontest(Yu2005)
TheboundaryconditionsofthestiffenedelementaresummarizedinFigure7.2.
Theplateissimplysupportedaroundtheperimeterwithsidesfreetowave.Thenodes
attheloadededgesoftheplatearecoupledtodisplacetogetherlongitudinally(inthe1
260
direction), which prevents local failure modes of the plate at the loaded edges. The
nodalcouplingisprovidedwithanequationconstraintinABAQUS.
Loaded edge coupled to move
together in 1 using equation
constraint (all u are equal)
Transverse midline
restrained in 3 (w=0)
Perimeter restrained in 2 (v=0),

unloaded edges free to wave
Longitudinal midline
restrained in 1 (u=0)
2
together in 1 using equation
constraint (all u are equal)
Figure7.2Simplysupportedboundaryconditionswithequationconstraintcouplingatloadededges
Two types of loading conditions, uniform load and uniform displacement, are
considered as shown in Figure 7.3. The uniform compressive load is applied as
consistent nodal loads on the plate edge. The magnitude of the uniform load is
representedbytheparameter,whichisanaccumulationofloadstepsautomatically
determined by ABAQUS. is large when the NewtonRaphson algorithm converges
quickly(alongthelinearbranchoftheloaddisplacementcurve)andadjuststosmaller
incrementsasequilibriumbecomesmoredifficulttoachieve(nearthepeakoftheload
displacementcurve).Fortheuniformdisplacementcase,thetotaldisplacementofthe
plate edges is applied over 100 steps, where the maximum displacement increment at
eachstepissetto=0.0145t.
261
P
h
h
(b)
(a)
Load scaling factor

P Unit force
h Loaded width of the plate
Displacement of plate edge
Figure7.3Applicationof(a)uniformloadand(b)uniformdisplacementtoastiffenedelement
Initial geometric imperfections are imposed based on the fundamental elastic

buckling mode of the stiffened element (see buckled shape in Figure 7.3). The
magnitudeoftheimperfectionsischosenbasedonaprobabilistictreatmentdeveloped
for coldformed steel members (Schafer and Pekz 1998). Since the stiffened element
considered here is chosen to be consistent with the web of a structural stud, a Type 1
(local buckling) imperfection is assumed as shown in Figure 7.4. The maximum
magnitudeoftheimperfectionfieldisselectedsuchthatthereisa50percentchancethat
a randomly occurring imperfection in the plate, , will have a magnitude less than d1,
i.e.,P(<d1)=0.50.Forthisprobabilityofoccurrence,theinitialimperfectionfieldofthe
stiffenedelementisscaledtod1/t=0.34.
262
Cross section with initial

geometric imperfections
d1
Figure7.4Type1imperfection(SchaferandPekz1998)
7.1.2 ABAQUS nonlinear solution methods
Twononlinearsolutionmethods,themodifiedRiksmethodandaNewtonRaphson
techniquewithartificialdamping,areavailableinABAQUStosolvedifficultnonlinear
problems.ThemodifiedRiksMethod(i.e.,*STATIC,RIKSinABAQUS),wasdeveloped
in the early 1980s and enforces an arc length constraint on the NewtonRaphson
incremental solution to assist in the identification of the equilibrium path at highly
nonlinearpointsalongtheloaddeflectioncurve.Thismethodisdiscussedextensively
in several publications (Crisfield 1981; Powell and Simons 1981; Ramm 1981; Schafer
1997; ABAQUS 2007a). Another solution option is a NewtonRaphson technique (i.e.,
*STATIC,STABILIZEinABAQUS)whichaddsartificialmassproportionaldampingas
localinstabilitiesdevelop(thatis,whenchangesinnodaldisplacementsincreaserapidly
over a solution increment) to maintain equilibrium (Yu 2005; ABAQUS 2007a). Local
instabilitiesnearpeakloadarecommonincoldformedsteelmembers,suchaswhena
thinplatedevelopsatafoldlinepriortocollapse.
263
Inthisstudy,thestiffenedelementdescribedinSection7.1.1isloadedtocollapse
in ABAQUS employing the modified Riks method with uniform loads applied
uniaxially (see Figure 7.3a) and then with the artificial damping solution method
employing uniform displacements (see Figure 7.3b). (Either method is capable of
solving problems with applied loads or applied displacements.) The goal of this
preliminary study is to gain experience with the solution controls for each method.
Additional background information pertaining to the ABAQUS implementation of the
artificial damping method is also discussed to provide specific guidance (and raise
futureresearchquestions)onitsproperuse.
7.1.2.1
ModifiedRikssolution
The loaddisplacement curves and deformed shapes (at peak load) of the
stiffened element solved with the modified Riks method are provided in Figure 7.5.
Different postpeak equilibrium paths were obtained by varying max, the maximum
loadincrementlimitfortheABAQUSautomaticstepselectionalgorithm.Theexistence
of multiple solutions is consistent with a plate containing periodic geometric
imperfections,sinceeachhalfwavelengthhasanequalchanceofdeforminglocallyinto
aplasticfailurezone.
Althoughtherewereseveraldifferentpostpeakpathsobserveddependingupon
the choice of max, the primary failure mechanism for the plate was a sharp yieldline
fold occurring transversely across the plate. Figure 7.6 demonstrates that this folding
occursatthecrestofthebuckledhalfwaveoftheinitialgeometricimperfectionfield;in
264
this case the failure mechanism of the plate is linked to the initial imperfection shape.
The quantity and location of the plastic folds influenced the overall ductility of the
stiffenedelement(i.e.,theareaundertheloaddisplacementcurve).Asthenumberof
folds increase, the postpeak strength and ductility of the plate increase. The peak
compressiveloadofthestiffenedelementwasnotsensitivetochangesinmax.
Failure modes and

associated Riks step sizes
0.9
0.8
0.8
max = 0.70
0.7
0.6
P/Py
0.6
max = 0.10, 0.20, 0.25,

0.30, 0.40, 0.60
0.5
0.4
0.4
0.3
max = 0.50
0.2
0.2
P
h
max = 0.35
0.1
0
Initial imperfection shape

(scale exaggerated)
0.2
0
0.25
0.5
0.4 h
0.75
0.6
1
0.8
1.25
1.5
max = 0.05,0.15
1.75
/t
Figure7.5ModifiedRiksmethodloaddisplacementsolutionsandfailuremodes
ultimate limit state

fold line (typ.)
initial state with

assumed
imperfections
(imperfection and deformation magnitudes not to scale)
Figure7.6Correlationbetweeninitialimperfectionshapeandfoldlinelocationsatfailure
265
7.1.2.2
7.1.2.2.1
Artificialdampingsolution
Backgroundonartificialdampingsolutionmethod
Artificial mass proportional damping is employed in ABAQUS to alleviate local
instabilities in the *STATIC, STABILIZE solution method. The global equilibrium

equationsateachdisplacementstepcanbewrittenas:
P F D = 0
(7.1)
where P is the vector of applied external forces, F is the vector of calculated internal
forces, and D is the vector of viscous damping forces. The damping force vector D is
calculatedateachstepbasedonthefollowingrelationship:
D = (cM )v
(7.2)
where c is the damping ratio, M is an artificial mass matrix calculated with a unit
materialdensity,andvrepresentsthechangeinnodaldisplacementsdividedbythesize
of the time step selected by ABAQUS. v is called the nodal velocity in ABAQUS
since the dimensions are length/time, which makes v sensitive to the definition of
time.Inthisstudy,thetotaltimeisselectedasoneunitandthemaximumtime
stepallowedis0.01units.Ifthetotaltimeischosenas100unitsandthemaximum
time step as 1 unit, it seems that the magnitude of the damping forces D would
change.Followingthesameargument,themagnitudeofvisimpactedbythechoiceof
unitsfortheproblem(feet,inches,meters,mm)sincevhasdimensionsoflength/time
units.Futureworkisneededtoevaluatetheinfluenceoftimeandlengthunitsonthe
calculationofthenodalvelocityv.Theevaluationoftheartificialdampingsolution
266
sensitivitytothemagnitudeanddistributionofmassinamemberisalsoanotherfuture
researchtopic.
When the solution is stable, changes in nodal displacements are small and
viscousdampingisnegligible.Whenlargechangesindisplacementsoccurbetweentwo
consecutiveloadsteps(asinthecaseofalocalinstability),dampingforcesareappliedto
helpmakeupthedifferencebetweenPandF.vmayonlybehighforcertainlocations
inthemember,andthereforedampingwillonlybeappliedthere.ABAQUSprovides
both automatic and manual options for selection the damping factor c; if c is chosen
manually,ABAQUSrecommendsthatitshouldbechosenasasmallnumbersincelarge
damping forces can add too much artificial stiffness to the system, producing an
unreasonable solution. When the automatic option is selected, ABAQUS finds c such
thatthedissipatedenergytototalstrainenergyratioafterthefirstloadstepisequalto
2.0x104.
7.1.2.3
Artificialdampingresults
TheartificialdampingsolutionsforthestiffenedelementarepresentedinFigure
7.7. Loaddeformation results pertaining to both manually and ABAQUSselected
damping factors are plotted, demonstrating that the magnitude of the damping
parameter c influences the postpeak response and causes the prediction of several
differentloadpaths(inasimilarwaytohowmaxaffectedthemodifiedRikssolutions).
Peakloadisnotsensitivetothequantityofdampinginthiscase,andisconsistentwith
themagnitudepredictedwiththemodifiedRiksmethod.
267
0.9
0.8
ABAQUS chooses this

path (c=0.0162)
0.8
Failure modes and

associated damping
factor c
0.7
0.6
P/Py
0.6
c=0.005
0.5
0.4
0.4
0.3
0.2
0.2
c=0.000, 0.0162
c=0.050
0.1
0

(scale exaggerated)
P
h
Displacement
0.2
Control
0.25
0.5
0.4
P
h
0.75
0.6
1
0.8
1.25
1.5
1
1.75
/t
Figure7.7Artificialdampingloaddisplacementsolutionsandfailuremodes
7.1.3 ABAQUS nonlinear solution controls

Section 7.1.2 summarizes the preliminary experiences gained using ABAQUS
nonlinear solution methods to determine the ultimate strength of stiffened elements.
Equilibriumpathsandfailuremodescanbesensitivetosolutioncontrols,althoughthe
peakresistingloadoftheplatewasconsistentlypredicted.Thenonlinearsolutionofa
stiffened element with a slotted hole is attempted with the modified Riks method
(*STATIC, RIKS), the default NewtonRaphson solution algorithm (*STATIC), and
NewtonRaphson with artificial damping (*STATIC, STABILIZE) solution methods
availableinABAQUS.Thegoalofthestudyistodetermineasetofsolutionscontrols
(loadstepsize,dampingfactor,convergencelimits)thatiscapableofcapturingthepost
peakloaddisplacementresponseofastiffenedelementwithahole.
268
7.1.3.1
Problemdescription
The stiffened element described in Section 7.1.1 is considered in this study. A

single slotted hole is placed at the midlength of the plate and centered between the
unloadededges.Theslottedholehasdimensionsofhhole=1.5in.,Lhole=4in.,andrhole=0.75in.
(seeFigure3.2forholedimensiondefinitions).
The boundary conditions of the stiffened element were initially assumed to be
simplysupported with both transverse and longitudinal plate midlines restrained and
theloadededgenodescoupledwithconstraintequationsasdescribedinFigure7.2.It
wasoftenobservedthattheconstraintsusedtoenforceuniformdisplacementsin1(u)at
the loaded edges and the transverse midline restraint in 2 (w=0) were contributing to
solutionconvergenceproblems,andthereforeanalternativesetofboundaryconditions
wasdevelopedasshowninFigure7.8.
Restraint at center of
loaded edge in 3 (w=0)
together in 1 using rigid body
formulation in ABAQUS
(all u are equal)
Perimeter restrained in 2 (v=0),
unloaded edges free to wave
Longitudinal midline
restrained in 1 (u=0)
Restraint at center of
loaded edge in 3 (w=0)
2
together in 1 using rigid body
formulation in ABAQUS
(all u are equal)
Figure7.8Stiffenedelementboundaryconditionswithrigidbodycouplingatloadededges
Acomparisonofthegeometricimperfectionsassumedforthestiffenedelement
withandwithouttheholeisprovidedinFigure7.9.d1/t=0.34isusedtoscaletheinitial
269
imperfection field of the plate. This magnitude corresponds to a probability of

occurrenceofP(<d1)=0.50(seeSection7.1.1fordetails).Thepeakloadofthestiffened
element is sensitive to initial geometric imperfections, and therefore it is important to
considerthesameimperfectionshapewhencomparingtheloaddisplacementresponses
ofthestiffenedelementwithandwithoutahole.Theimperfectionshapeisimposedon
thestiffenedelementwiththeslottedholebymappingthebuckledmodeshapetonodal
coordinatesusingcustomMATLABcode(Mathworks2006).
fundamental
buckling mode of
plate
initial geometric
imperfection fields
fundamental buckling
mode mapped to plate
with slotted hole
Figure7.9Initialgeometricimperfectionfieldusedforthestiffenedelementwithandwithoutahole
EightexploratoryABAQUSmodelsareevaluatedinthisstudy,eachsolvingthe
samenonlinearproblemofastiffenedelementwithaslottedholecompresseduniaxially
until failure as shown in Figure 7.10. Each model employs a different combination of
ABAQUSsolutioncontrolsandboundaryconditionsassummarizedinTable7.1.
270
Figure7.10Deformationatultimateloadofastiffenedelementwithaholeloadedincompression.Thecommonfailuremechanismismaterialyieldingadjacent
totheholefollowedbyplatefolding.
Table7.1Summaryofnonlinearfiniteelementmodelsandassociatedsolutioncontrols
*STATIC
Model
ABAQUS Method
initial
step size
total
time
min step
size
RIKS1
RIKS2
RIKS3
STATIC1
STATIC2
STAB1
STAB2
STAB3
*STATIC, RIKS
*STATIC, RIKS
*STATIC, RIKS
*STATIC
*STATIC
*STATIC, STABILIZE
*STATIC, STABILIZE
*STATIC, STABILIZE
------0.01
0.01
0.01
0.01
0.01
------1
1
1
1
1
------1.00E-20
1.00E-20
1.00E-20
1.00E-20
1.00E-20
max
step
size
------0.01
0.01
0.01
0.01
0.01
*STATIC, RIKS
Solution Controls
initial
residual
damping
iteration limits
max step
step
limits
factor
size
size
Io
Ir
Ic
c
Rn
0.05
0.2
4 (D) 8 (D) 16 (D) 0.005 (D)
--(D)
(D)
4 (D) 8 (D) 16 (D) 0.005 (D)
--0.05
0.05
8
16
33
0.005 (D)
------8
16
33
0.005 (D)
------8
16
33
0.005 (D)
------8
16
33
0.005 (D)
0.0162
----8
16
33
0.1
0.0162
----8
16
33
0.005 (D)
0.0162
271
line
search
NO
NO
NO
NO
YES
NO
NO
NO
Loading Type and Boundary Conditions

uniform load, equation constraint
imposed displacement, rigid body constraint
uniform load, rigid body constraint
7.1.3.2
ModifiedRiksmethodsolutioncontrols
The RIKS1 and RIKS2 finite element models are loaded with a uniformly
distributed load at both ends as shown in Figure 7.3(a), where equation constraints
couple the loaded edge nodes (see Figure 7.2). The initial and maximum load step
magnitudesaredefinedforRIKS1basedonexperiencegainedfromthestudyinSection
7.1.2.1. The RIKS2 model allows ABAQUS to select all load stepping parameters
automatically.
1
RIKS1
RIKS2
0.8
0.6
cannot move
past peak load
0.4
P/Py,g
0.2
compression
tension
-0.2
-0.4
-0.6
-0.8
-1
-2

(scale exaggerated)
P
b
-1.5
-1
-0.5
/t
0.5
1.5
P
b
Figure7.11LoaddisplacementcurvefortheRIKS1andRIKS2modelsshowingdirectionreversalalong
loadpath
The loaddisplacement responses from the RIKS1 and RIKS2 models are
compared in Figure 7.11. For both models, ABAQUS does not capture the peak load
andrevertsbackalongtheequilibriumpathuntiltheplateisloadedtofailureintension!
TheABAQUSTheoryManualstatesthatthistypeofdirectionswitchispossiblewhen
272
the equilibrium path exhibits very high curvature (ABAQUS 2007a). The ABAQUS
messagefiles(.msg)for thesemodelsreportthatthemomentresidualsaretoohighat
theloadededgenodesandatnodesalongthetransversemidlineoftheplate,suggesting
that these boundary conditions are contributing to the convergence difficulties for the
solution.
TheRIKS3modelisloadedwithimposeddisplacementsatbothendsasshown
inFigure7.3(b),wherethemidlineconstraintisremovedandtheloadededgenodesare
coupledwitharigidbodyconstraintinsteadofanequationconstraintinABAQUS(see
Figure 7.8). According the ABAQUS Analysis Users Manual, only the reference node
governing the motion of the rigid body is involved in element level calculations. This
improves computational efficiency and releases the solution algorithm from the force
andmomentresidualminimizationconstraintsforallnodesintherigidbodyexceptthe
referencenode.
The solution results from the stiffened elements loaded with consistent nodal
loads(RIKS1,RIKS2)andimposeddisplacements(RIKS3)arecomparedinFigure7.12.
Before yielding occurs, the three models produce nearly identical loaddisplacement
results. As yielding initiates, the RIKS3 model predicts a peak load and postpeak
response for the stiffened element. This comparison demonstrates that imposed
displacements and rigid body constraints (in contrast to applied loads and equation
constraints)improvethechancesforconvergenceinthiscase.
273
RIKS1, RIKS2
RIKS3
P/Py,g or P/Py,g
0.8
0.6
0.39
0.8
0.4
0.37
0.2
0.6
0.35
0.24
0
-0.2
-0.4
Riks Method finds post-peak path

when plate is compressed with
imposed displacements
0.29
0.34
0.4
Riks Method retreats back along
elastic load path when plate is
loaded with consistent nodal
loads
0.2
-0.6
-0.8
-1
-2
P
h
0
-1.5
0.2
-1
0.4
-0.5
0.6
0
0.5
/t
0.8
1
1.5
P
1h
2
Figure7.12RIKS1andRIKS2modelsexperienceconvergenceproblemsandreturnalongtheloadingpath,
theRIKS3modelsuccessfullypredictsapeakloadandfindsapostpeakloadpath
7.1.3.3
NewtonRaphsonmethod
TheSTATIC1andSTATIC2modelsemploytheNewtonRaphsonalgorithmwith
uniform displacements at the loaded edges imposed with equation constraints (see
Figure 7.2). The stepping parameters are chosen to ensure at least 100 increments are
achieved before completion of the simulation. The number of convergence criteria
iterationsisalsomodifiedbydoublingtheABAQUSparametersIo,Ir,andIcfromtheir
defaultvalues(seeTable7.1).Iorepresentsthenumberofequilibriumiterationsbeforea
check is performed to ensure that the magnitudes of the moment and force residual
vectors are decreasing. After Io iterations, if the residuals are not decreasing between
twoconsecutiveequilibriumiterationsthenthelengthoftheincrementstepisreduced
and the equilibrium search is restarted. Ir represents the number of equilibrium
274
iterationsafterwhichthelogarithmicrateofconvergencecheckbegins.Icrepresentsthe
maximumnumberofequilibriumiterationswithinatimeincrementstep. Alinesearch
algorithm is also employed in the STATIC2 model to improve the convergence of the
NewtonRaphson algorithm when nodal force and moment residuals are large. This
algorithmfindsthesolutioncorrectionvectorwhichminimizestheoutofbalanceforces
inthestructuralsystem(ABAQUS2007a).
0.37
STATIC1
STATIC2
P/Py,g
0.9
0.8
0.365
STATIC2 (with line search algorithm)

finds post-peak equilibrium path
before terminating
0.7
P/Py,g
0.6
0.36
0.26
0.27
/t
0.5
0.28
0.4
0.2
0.1
0

(scale exaggerated)
P
h
0.3
P
h
Displacement control
0.25
0.5
0.75
/t
1.25
1.5
1.75
Figure7.13STATIC1andSTATIC2loaddisplacementcurvesdemonstrateconvergencedifficultiesnearthe
peakload.
Figure7.13comparestheSTATIC1andSTATIC2loaddisplacementcurves.The
STATIC1 model finds the peak load but then terminates due to moment residual
convergenceissuesasitattemptstopredictthefirststepofthepostpeakresponse.The
ABAQUS message (.msg) file for this model states that themoment residuals at nodes
along the loaded edges, along the transverse midline of the plate, and at some nodes
275
near the hole are increasing and convergence is judged unlikely. The solution is
terminated after the automatic time stepping procedure requires a smaller time step
thantheminimumsetinthismodel(1x1020).TheSTATIC2modelwiththelinesearch
algorithm also finds the peak load of the stiffened element and is able to track onto a
postpeak equilibrium path before terminating from the same convergence problems
experienced by the STATIC1 model. The success of the line search algorithm in
identifying a postpeak equilibrium path highlights its potential for solving nonlinear
problems, although a significant increase in computational effort (almost twice the
wallclocktime)wasalsonoted.
7.1.3.4
NewtonRaphson
with
artificial
damping
The STAB1 and STAB2 models solve the stiffened element problem using a
displacement control Newton Raphson algorithm coupled with the automatic artificial
damping discussed in Section 7.1.2.2. The boundary conditions are modified to those
summarized in Figure 7.8 because of the convergence issues observed with the
constraintequationsandtransversemidlinerestraints.AsinthecaseofSTATIC1and
STATIC2, the convergence iteration limits Io, Ir, and Ic are doubled from their default
values. In an attempt to alleviate the moment residual convergence issues from
previous runs, the Newton Raphson parameter R n is modified to relax the residual
requirementswhenthesolutionapproachesthepeakload.R nistheallowablelimiton
theratioofthelargestresidualforceormomentatanode(r max)tothelargestchangein
276
force or moment at a node averaged over each time step increment that has been
completed (q). The superscript indicates that R n can be defined for either a
displacement field u or a rotation field . The convergence limit can be written

mathematicallyas:
rmax
Rn q .
(7.3)
ThedefaultforR nof0.005isusedinSTAB1forbothuand fields,whereasinSTAB2

u
R n=0.005andR n=0.100.
1
0.38
STAB1
STAB2
P/Py,g
0.9
0.8
0.34
Highly nonlinear post-peak

equilibrium path found with
STAB1 and STAB2
0.7
0.3
0.25
P/Py,g
0.6
0.3
/t
0.5
0.35
0.4
0.2
0.1
0

(scale exaggerated)
P
h
0.3
P
h
0.25
0.5
0.75
1.25
/t
1.5
1.75
Figure7.14STAB1andSTAB2loaddisplacementcurvesdemonstrateahighlynonlinearpostpeak
equilibriumpath
The ABAQUS solutions from models STAB1 and STAB2 in Figure 7.14
demonstrate a highly nonlinear postpeak equilibrium path. Both models are able to
successfullypredictthepeakloadandthenmovetoasecondaryloadpath.Thesolution
terminates because the maximum number of NewtonRaphson iterations is reached.
ThemodificationofthemomentresiduallimitR nfrom0.005to0.100didnotinfluence
thesolutionresults.
277
The STAB3 finite element model employs a uniform loading with the Newton
Raphson algorithm and artificial damping to determine the nonlinear response of the
stiffenedelementwithaslottedhole.TheSTAB3boundaryconditionsarethesameas
thosefortheSTAB1andSTAB2models,wheretheplateedgesareconstrainedtomove
totogetherasrigidbodies(seeFigure7.8).Figure7.15comparestheSTAB1andSTAB2
(displacement control) to the STAB3 (load control) results and shows that, prior to
yielding,thethreemodelspredictthesameresponse.Differencesintheloadpathsare
observedafteryieldingthough,especiallyintheSTAB3model,whichreachespeakload
and then carries this load with zero stiffness over a large deformation range. This
unstable postpeak behavior results from a complete loss of stiffness as the hole
collapsesunderloadcontrol.Thepeakloadspredictedforthestiffenedelementbythe
displacementcontrolSTAB3modelissevenpercenthigherthantheSTAB1andSTAB2
load control solutions, demonstrating that the peak load is sensitive to the loading
method(uniformloadoruniformdisplacements)inthiscase.
278
STAB1, STAB2
STAB3
0.9
0.4
0.8
0.8
0.35
0.7
0.6
P/Py,g
0.6
0.5
0.3
Higher peak load found with load

control compared to displacement
control (0.395Py,g versus 0.375 Py,g)
0.28 0.32
0.36
Load control solution demonstrates
complete loss of stiffness at peak load
(yielding and plate folding at hole)
0.4
0.4
0.3
P
h
0.2
0.2
0
0.1
0
0.2
0.25
0.5
0.4
0.75
0.6
1
/t
1.25
0.8
1.5
P 1
h
1.75
Figure7.15TheSTAB1andSTAB2models(artificialdamping,displacementcontrol)exhibitasharpdropin
loadasfoldingoftheplateinitiatesnearthehole.TheSTAB3model(artificialdamping,loadcontrol)finds
thecompressiveloadatwhichacompletelossofstiffnessoccurs.
7.1.4 Influence of a slotted hole on the ultimate strength of

a stiffened element (without geometric imperfections)
The solution controls from the previous section resulting in successful
simulations are now implemented to evaluate the influence of a slotted hole on the
ultimate strength and failure mode of a stiffened element. The loading and boundary
conditions,dimensions,materialproperties,andsolutioncontrolsarethesameasthose
usedfortheSTAB2modeldescribedinSection7.1.3.4andTable7.1.Initialgeometric
imperfectionsarenotconsidered.
279
Ultimate Limit State
0.9
0.8
0.8
0.7
Py,net=0.56*Py,g
0.6
P/Py,g
0.6
0.5
Pcr,no hole=0.33*Py,g
0.4
Pcr,hole=0.30*Py,g
0.4
0.3
0.2
0.2
0
0.1
0
Elastic Buckling
0
0.2
0.25
0.5
0.4
0.75
0.6
1
0.8
1.25
/t
1.5
1
1.75
Figure7.16Comparisonofultimatelimitstateandelasticbucklingplatebehavior,initialimperfectionsare
notconsideredintheseresults
Figure7.16demonstratesthattheslottedholereducesthestrengthofthestiffened
elementfrom1.0Py,g to0.58Py,g,wherePy,gistheresultantcompressiveforceonthe
stiffenedelementtocauseyieldingcalculatedwiththegrosscrosssectionalareaof
the plate. The predicted peak load of the stiffened element with the hole is
consistentwiththeloadatyieldingofthenetsection,Py,net=0.56Py,g.Thisobservation,
thatthestrengthofthestiffenedelementwiththeholeislimitedtoPy,net,highlights
an important consideration in the development of the Direct Strength Method in
Chapter 8. The hole also reduces the axial stiffness of thestiffened element in this
case, as demonstrated by the change in slope of the linear portion of the load
displacementcurveinFigure7.16.
280
7.1.5 Influence of geometric imperfection magnitudes on

the ultimate strength of a stiffened element with and
without a slotted hole
The ultimate strength of coldformed steel members is sensitive to initial
geometricimperfections.Inthisstudytheinfluenceofimperfectionmagnitudeonthe
ultimatestrengthofstiffenedelementswithandwithoutaslottedholeisevaluated.The
loadingandboundaryconditions,dimensions,materialproperties,andsolutioncontrols
are the same as those used for the STAB2 model discussed in Section 7.1.3.4 and
summarizedinTable7.1.Theimperfectionfieldisassumedasthefundamentalelastic
bucklingmodepicturedinFigure7.9.Localbuckling(Type1)imperfectionmagnitudes
correspondingtoP(<d1)=0.25,0.50,0.75,0.95,and0.99fromtheCDFinFigure7.32are
considered.
Figure7.17andFigure7.18presenttheloaddisplacementresultsforthestiffened
element without and with the hole and demonstrate that increasing imperfection
magnitudesreducespeakloadandchangepostpeakresponse.Theelasticstiffnessis
alsosoftened,whichcanbeobservedbycomparingtheimperfectionresultstothelinear
slopeoftheloaddisplacementcurvewithoutimperfections.Thissofteningresultsfrom
theinitialoutofplanedeformationswhichengagemoreofthebendingstiffnessofthe
plate and less of the axial stiffness as the plate is compressed. Outofplane
deformations (such as initial imperfections) increase the magnitude of the geometric
stiffness matrix which negates the initial elastic stiffness of the undeformed system.
Imperfections are observed to decrease strength but increase ductility of the stiffened
281
elementwithandwithoutahole.Theloaddisplacementresultsalsohighlightthatthe
holereducestheductilityofthestiffenedelement,whichisconsistentwiththecolumn
experimentresultsinChapter5.
1
0.9
no imperfections
d1/t=0.14
0.8
d1/t=0.34
d1/t=0.66
0.7
d1/t=1.35
d1/t=3.85
P/Py,g
0.6
0.5
0.4
0.3
P
h
0.2
0.1
0
Displacement
Control
0.25
0.5
0.75
P
h
1.25
1.5
1.75
/t
Figure7.17Loaddisplacementsensitivitytoimperfectionmagnitudeforaplatewithoutahole
1
0.9
no imperfections
d1/t=0.14
0.8
d1/t=0.34
d1/t=0.66
0.7
d1/t=1.35
P/Py,g
0.6
d1/t=3.85
0.5
0.4
P
h
0.3
0.2
0.1
0
P
h
0.25
0.5
0.75
/t
1.25
1.5
1.75
Figure7.18Loaddisplacementsensitivitytoimperfectionmagnitudeforaplatewithaslottedhole
282
7.1.6 Determination of unstiffened element effective

width using nonlinear finite element modeling
The effective width method provides an approximation to the complex non
uniform stress distribution in a thin buckled plate under compression. Initially
presentedinthe1930sbyvonKarmanandextendedtocoldformedsteelmembersby
Winterinthe1940s,themethodaccountsforthereductioninloadcarryingcapacityof
astiffenedelement(vonKarmanetal.1932;Winter1947).Theinabilityofthecenterof
theplatetocarrycompressiveloadiscausedbyoutofplanedeformationsintheshape
ofthefundamentalelasticbucklingmode.Thesedeformationsreducetheaxialstiffness,
concentratingthecompressiveforceattheedgesofaplate.Theultimateloadisreached
whentheseedgestresses,carriedbytheeffectivewidth,exceedtheyieldstressofthe
plate material. The effective width concept is the basis of most coldformed steel
designcodesaroundtheworldtoday.
In this study, a nonlinear finite element model is employed to calculate the
longitudinal stress distribution at failure for a stiffened element with and without a
slottedhole.Thedistributionofstressesforbothcasesiscompared,andthevariationin
effective width along the length of the stiffened element is determined. The stiffened
element is modeled with the same loading and boundary conditions, dimensions,
materialproperties,andsolutioncontrolsasthoseusedfortheSTAB2modeldiscussed
in Section 7.1.3.4 and described in Table 7.1. The initial imperfection geometry
correspondstothefundamentalelasticbucklingmodeoftheplatewithouttheholeas
described in Figure 7.9. d1/t=0.34 is used to scale the initial imperfection field of the
283
plate,whichcorrespondstoaprobabilityofoccurrenceofP(<d1)=0.50asdiscussedin
Section7.1.1.Theeffectivewidthiscalculatedbyfirstintegratingthelongitudinal(S11)
membrane stresses at crosssections along the length of the stiffened element and then
dividingtheresultingareasbytheyieldstressofthesteelasshowninFigure7.19.The
membranestressesarethelongitudinal(S11)stressesthatoccuratthemidplaneofthe
stiffenedelementasdefinedinFigure7.20.
distribute area (A) to
edges of plate
he/2
A/2
calculate area under

stress curve (A)
A/2
he/2
membrane stress (S11)
yield stress
Figure7.19Calculationofeffectivewidthatacrosssectionalongastiffenedelement
+S11
+S11
Plan view of element
Top
Membrane
stress
Midplane
Membrane
stress
Bottom
Elevation view of element
Figure7.20Definitionoflongitudinal(S11)membranestress
284
Figure 7.21(a) highlights the variation in membrane longitudinal stress (S11)

occurring at the failure load of the stiffened element. The highest stresses accumulate
along the edges of the plate and decrease toward the center of the plate. The largest
edgestressesoccuratthecrestsofthehalfwaveswherethegreystresscontoursindicate
yieldingoftheplate.ThecorrespondingeffectivewidthispresentedinFigure7.21(b).
The maximum effective width of 0.51 he/h occurs at the inflection point between half
waves, while the minimum effective width of 0.48 he/h occurs at the wave crests. The
predictedeffectivewidthforthisplateusingSectionB2.1oftheAISIspecificationis0.50
he/h(AISIS1002007).
+S11
S, S11
Mid, (fraction = 0.0)
(Ave. Crit.: 75%)
+5.059e-01
-2.286e+00
-5.078e+00
-7.871e+00
-1.066e+01
-1.345e+01
-1.625e+01
-1.904e+01
-2.183e+01
-2.462e+01
-2.742e+01
-3.021e+01
-3.300e+01
-3.578e+01
+S11
effective width
average
standard deviation
max
min
he/h
0.51
0.02
0.55
0.48
Elevation
(a) membrane stress in 1 direction (S11)
he/2
(b) variation in effective width along plate
Figure7.21(a)longitudinalmembranestressesand(b)effectivewidthofastiffenedelementatfailure
The failure mode of the stiffened element with the slotted hole is fundamentally
differentthanwithoutthehole.ThestressesinFigure7.22(a)demonstratethatyielding
occurs only at the location of the hole when the peak load is reached. Compressive
stressesarehighestattheedgeoftheplateandthentransitiontotensilestressesatthe
285
faceofthehole.TheeffectivewidthoftheyieldedportionoftheplateinFigure7.22(b)
islessthanthatfortheplatewithoutthehole,evenwiththebeneficialtensilestressesat
theface.Theaverageeffectivewidthis0.38he/h,whichis25percentlessthanthatofthe
stiffenedelementwithoutthehole.ThepredictedeffectivewidthusingSectionB2.2of
theAISISpecificationis0.30he/h.Theeffectivewidthsofthestiffenedelementwithand
withoutaslottedholearecomparedtogetherinFigure7.23.
+S11
S, S11
Mid, (fraction = 0.0)
(Ave. Crit.: 75%)
+1.151e+01
+7.804e+00
+4.095e+00
+3.852e-01
-3.324e+00
-7.034e+00
-1.074e+01
-1.445e+01
-1.816e+01
-2.187e+01
-2.558e+01
-2.929e+01
-3.300e+01
-3.615e+01
+S11
effective width
average
standard deviation
max
min
he/h
0.38
0.03
0.41
0.34
Elevation
(a) membrane stress in 1 direction (S11)
he/2
(b) variation in effective width along plate
Figure7.22(a)longitudinalmembranestressesand(b)effectivewidthofastiffenedelementwithaslotted
holeatfailure
he/2 (plate with hole)
he/2 (plate without a hole)
Figure7.23Effectivewidthcomparisonforaplatewithandwithoutaslottedhole
The longitudinal stresses (S11) in the top and bottom fibers of the stiffened
elementatfailurearedifferentfromthemembranestressesatthemidplane,suggesting
286
that the effective width of a stiffened element actually varies through its thickness.
Figure 7.24 and Figure 7.25 provide a comparison of this variation for a stiffened
elementwithandwithoutaslottedhole.Itisobservedthataplateismoreeffectiveon
thesurfacewheretheoutofplanedeformationcausescompression.Theeffectivewidth
isreducedwhentensileandcompressivestressesnegateeachother,asshowninthe2D
plotofextremefiberandmembranestressesatarepresentativecrosssectioninFigure
7.26.
Effective width calculated with
longitudinal stresses (S11) at top,
midplane, and bottom of the plate
Top of plate
Midplane of plate
Bottom of plate
Figure7.24Throughthethicknessvariationofeffectivewidthofaplatewithoutahole
Effective Width calculated with

longitudinal stresses (S11) at top,
midplane, and bottom of the plate
Top of plate
Midplane of plate
Bottom of plate
Figure7.25Throughthethicknessvariationofeffectivewidthofaplatewithaslottedhole
287
Longitudinal (S11) stress

variation across width of plate
1
top of plate
midplane of plate
bottom of plate
0.9
0.8
0.7
x/h
0.6
0.5
Top of plate is
fully effective
Stress
distribution used
to calculate
code-based
effective width
0.4
0.3
Tension and compression

stresses counteract each other
when calculating effective width
at the bottom of the plate
0.2
0.1
0
-1.5
Compression
-1
-0.5
Tension
0
fplate/fy
0.5
SECTION A-A
1.5
Figure7.26Throughthicknessvariationinlongitudinal(S11)stressesinaplateatfailure
7.2 Nonlinear finite element modeling of columns

withholes
A more extensive study of ABAQUS nonlinear finite element capabilities of cold
formed steel columns with holes is now presented. Simulation to collapse of the 24
column experiments described in Chapter 5 is performed, considering solution
sensitivity to specific modeling parameters including initial imperfections, residual
stresses and the coldwork of forming, nonlinear material modeling, and column
boundary conditions. A modeling protocol is developed which produces results
consistent with column experiments. This modeling tool is employed to explore the
288
validity of proposed Direct Strength Method equations for coldformed steel members
withholespresentedinChapter8.
7.2.1 Modeling protocol development

7.2.1.1
Modeldimensionsandfiniteelementmeshing
The collapse behavior of the 24 column specimens is simulated with the general
purposefiniteelementprogramABAQUS(ABAQUS2007a).Allcolumnsaremodeled
withS9R5reducedintegrationninenodethinshellelements(seeSection2.1fordetails
ontheS9R5element).Thefiniteelementmeshforeachspecimeniscreatedwithcustom
Matlab code developed by the author (see Appendix A); the mesh is consistent with
S9R5 meshing guidelines summarized in Section 2.4. The centerline Csection
dimensionsinputintoABAQUSarecalculatedusingtheouttooutdimensionsofeach
columnspecimenprovidedinTable5.3.Thecrosssectioncorneranglesareassumedas
right angles (even though they were measured to be off of 90 degrees, see Table 5.4)
since the distortional imperfection magnitudes obtained in Section 7.2.1.5 are derived
based on a nominal crosssection with 90 degree corners*. The average base metal
thicknessforeachspecimen(i.e.,theaverageoftbare,w,tbare,f1,andtbare,f2fromTable5.5)and
column length L from Table 5.6 are used to construct the ABAQUS models, as are the
locationoftheslottedwebholesrelativetothecenterlineofthewebprovidedinTable
5.8.
* The measured flangeweb and weblip angles were not considered because of initial difficulties matching the
experiment results to the simulations. To resolve these difficulties, a simplified model with nominal dimensions was
implemented.(Modelingwithplasticityattheproportionallimitwasfoundtobethetruecauseofthediscrepancies,see
Section7.2.1.4.)Considerationoftheactualcrosssectiondimensions,includingtheflaredwebflangecornersmeasured
intheexperiments,iswarrantedandisanimportantpointoffuturestudy.
289
7.2.1.2
Boundaryconditionsandapplicationofloading
The specimen boundary conditions in ABAQUS are defined to simulate the
experiment boundary conditions as shown in Figure 7.27. The nodes on the loaded
column face are coupled together in the direction of loading (1 direction) with an
ABAQUSpinnedrigidbodyconstraint.Thisconstraintensuresthatallnodesonthe
loaded face of the column translate together, while the rotational degrees of freedom
remainindependent(asinthecaseofplatenbearing).Atotalimposeddisplacementof
0.20inchesisappliedtothereferencenodeoftheABAQUSrigidbodyoveraseriesof
steps (see Section 7.2.1.3) to simulate the displacement control loading applied by the
bottom platen during the experiment. Frictioncontact boundary conditions were also
evaluated in ABAQUS as described in Appendix J although their influence on the
ultimatestrengthofthecolumnspecimenswasdeterminedtobeminimal.
constrained in 1, 2 and 3
1
2
3
ABAQUS pinned rigid

body reference node
constrained in 2 to 6
directions, ensures that all
nodes on loaded surface
move together in 1 direction
Apply uniform displacement in ABAQUS to

simulate displacement control of experiments
Figure7.27ABAQUSboundaryconditionssimulatingcolumnexperiments
290
7.2.1.3
Nonlinearsolutionmethod
ThemodifiedRiksmethod(*STATIC,RIKS)isemployedasthesolutionalgorithmin
this study. The preliminary nonlinear finite element studies on stiffened elements
demonstrated that the modified Riks method was able to capture the complete load
deformation response when imposed displacements are used to load the member (see
Figure7.12).ABAQUSautomatictimesteppingwasenabled,withthesuggestedinitial
stepsizesetto0.005,themaximumstepsizelimitedto0.01,andthemaximumnumber
ofincrementsequalto300allinputbytheuser.
7.2.1.4
Materialmodeling
Steel yielding and plasticity is simulated in ABAQUS using a classical metal
plasticityapproachwithisotropichardening.AMisesyieldsurfaceisdefinedwiththe
true stress and true plastic strain obtained from uniaxial tensile coupon tests for each
specimen.Threestressstraincurves(westflange,eastflange,andweb)wereobtained
for each specimen (see Section 5.2.6). The experimentally obtained engineering stress
straincurvesareconvertedtotruestressandstrainandthenaveragedpointbypointto
produceayieldstress,proportionallimit,andtruestressstraincurveforeachspecimen.
(The true plastic strains and associated stresses are input into ABAQUS with the
*PLASTIC command.) For Mises stresses below the yieldstress, linear elastic material
behaviorisassumedwhereE=29500ksiand=0.3.
291
Preliminary nonlinear modeling efforts for this study determined that including
plastic strains starting at the proportional limit resulted in ABAQUS simulation

predictions that were as much as 25% lower than the column tested strengths. An
example of the average true stressstrain curves with plasticity starting at the
proportional limit are provided in Figure 7.28a for specimen 362124NH and Figure
7.29aforspecimen600124NH(averagetruestressstraincurvesareprovidedforall24
columnspecimensinAppendixH).Otherresearchershaveobtainedsimulationresults
consistentwithexperimentsbyassumingthatplasticstrainsinitiateonlyaftertheyield
stress (determined with the 0.2% offset method) is reached (Schafer 1997; Yu 2005;
Schafer et al. 2006) a material modeling approach that proved to be successful at
predictingthecolumnexperimentpeakloadsforthisstudyalso.Thetruestressstrain
curves in Appendix H were therefore modified to ensure that plasticity initiates in
ABAQUS only after the yield stress is reached for the gradually yielding stressstrain
curves(362S16233specimens,seeFigure7.28b)andattheinitiationoftheyieldplateau
for the sharpyielding curves (600S16233 specimens, see Figure 7.29b). Figure 7.30
demonstratesthedisparitybetweentheexperimentandFEsimulationloaddeformation
responseforcolumnspecimen600124NHwhenplasticityinitiatesattheproportional
limit.Thereasonsforthisdisparityareunclearandwarrantfuturestudy.Additional
researchworkisplannedtostudythedetailsofmetalplasticityinABAQUSbyloading
asinglefiniteelementtofailureintension.
292
100
100
90
90
80
80
true stress, ksi
ksi
60
50
40
(a)
30
20
10
0
70
true stress
true stress, ksi
true plastic
strain
70
0.02
0
0.001
0.002
0.007
0.012
0.017
0.027
0.037
0.047
0.057
0.067
33.1
46.1
51.9
60.3
64.9
68.4
74.0
78.1
81.3
83.8
86.2
true plastic
true stress
strain
ksi
0
55.1
0.003
60.3
0.008
64.9
0.013
68.4
0.023
74.0
0.033
78.1
0.043
81.3
0.053
83.8
0.063
86.2
60
50
40
(b)
30
20
10
0.04
0.06
0.08
true plastic strain
0.1
0.12
0.02
0.04
0.06
0.08
true plastic strain
0.1
0.12
Figure7.28ABAQUSplasticstraincurveforspecimen362124NHassuming(a)plasticityinitiatesatthe
proportionallimitand(b)plasticityinitiatesat0.2%offsetyieldstress
100
100
90
90
80
80
true stress, ksi
60
50
40
30
(a)
20
10
0
0.02
true stress
0
0.001
0.002
0.007
0.012
0.017
0.027
0.037
0.047
0.057
0.067
0.077
0.087
0.097
0.107
0.04
0.06
0.08
true plastic strain
70
ksi
37.5
54.7
58.3
60.0
61.5
64.0
70.2
74.4
77.5
80.0
81.9
83.5
84.9
86.1
87.2
true stress, ksi
true plastic
strain
70
60
50
40
30
(b)
20
10
0.1
0.12
0.02
true plastic
true stress
strain
ksi
0
58.3
0.005
60.0
0.01
61.5
0.015
64.0
0.025
70.2
0.035
74.4
0.045
77.5
0.055
80.0
0.065
81.9
0.075
83.5
0.085
84.9
0.095
86.1
0.105
87.2
0.04
0.06
0.08
true plastic strain
0.1
0.12
Figure7.29ABAQUSplasticstraincurveforspecimen600124NHassuming(a)plasticityinitiatesatthe
proportionallimitand(b)plasticityinitiatesatthebeginningoftheyieldplateau(refertoAppendixHfor
thedetailsonthedevelopmentofthiscurve).
293
14
plasticity at proportional limit
plasticity at yield stress
experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
axial displacement, in.
0.16
0.18
0.2
Figure7.30InfluenceofABAQUSmaterialmodelontheloaddeformationresponseofspecimen600124
NH(workthisfigurewithFigure7.29)
7.2.1.5
Initialgeometricimperfections
The ultimate strength and failure mechanisms of coldformed steel columns are
sensitivetoinitialgeometricimperfections,asdemonstratedinthepreliminarystudies
on stiffened elements in Section 7.1.5. In this study, the sympathetic local (L) and
distortional (D) elastic buckling modes are obtained with eigenbuckling analyses for
each column specimen and imposed on the nominal geometry in each finite element
model. (An ABAQUS .fil file is created for each eigenbuckling analysis which is then
calledfromthenonlinear.inpfilewiththe*IMPERFECTIONcommand).Theboundary
conditions at both specimen ends are assumed to be warping free when obtaining the
imperfection shapes (see Figure 4.2 for definition) to ensure consistency with CUFSM
boundary conditions. Specimens with and without holes are modeled with the same
elasticbucklingimperfectionshapesbyfillingtheholeswithadditionalfiniteelements
294
asshowninFigure7.31.Thisprocedureensuresthattheloaddisplacementbehaviorof
boththeholeandnoholespecimensarecomparedonequivalentbasis(bothwillhave
the no holeL and D imperfection shapes). Filling in the holes is necessary (instead of
eliminating them completely) because it preserves the nodal numbering and geometry
of thespecimenswithholes, making it convenient to superimpose the L and Dmodes
ontotheinitialnodalgeometryinABAQUS.
Hole is filled in with

S9R5 elements to
produce no hole local
buckling shape
Type1 imperfection (L)
Type 2 imperfection (D)
Figure7.31SlottedholesarefilledwithS9R5elementstoobtainnoholeimperfectionshapes
The magnitudes of the L and D imperfections are determined with the same
probabilistictreatmentusedforthestiffenedelementstudiesinSection7.1(Schaferand
Pekz 1998). Finite element simulations with L and D imperfection magnitudes
correspondingtothe25thand75thpercentilesoftheCDFinFigure7.32areperformedfor
eachspecimentoobtainarangeofsimulatedloaddisplacementresponsestocompare
to experiment results. FE simulations are also performed using the L and D elastic
295
bucklingmodeshapesandimperfectionmagnitudesmeasureddirectlyfromthecolumn
specimens.Inthiscasethelocalimperfectionmagnitudeforeachspecimenistakenas
the maximum deviation from the average web elevation as reported in Table 5.9. The
distortional imperfection magnitude for each specimen is determined by finding the
largest measured angular deviation from 90 degrees along each specimen and
calculatingtheassociatedflangelipdisplacementasshowninFigure7.33.TheType1
imperfectionmagnitudesmeasuredintheexperimentsareoften2to3timeslargerthan
the 75th percentile CDF magnitudes as shown in Table 7.2. The Type 2 imperfection
magnitudes for the 362S16233 specimens also are 2 to 3 times larger than the 75th
percentileCDFmagnitudes,primarilybecausethesespecimenstendedtoopenupatthe
sawnends(i.e.,flangewebanglesincreasedabove90degrees)whentheyweresawcut
from full stud lengths. Other researchers have studied this observed change in cross
section after sawcutting (Wang et al. 2006). The 600S16233 specimens were less
sensitive to this sawcutting effect, resulting in measured distortional imperfection
magnitudesconsistentwiththe75thpercentileoftheimperfectionCDF.
1.00
Type 1
Type 2
Type 1
Type 2
(L)
(D)
Probability (X < x)
0.80
0.60
P(X < x)
0.25
0.50
0.75
0.95
0.99
0.40
0.20
Type 1
d/t
0.14
0.34
0.66
1.35
3.87
0.50
0.66
Type 2
d/t
0.64
0.94
1.55
3.44
4.47
1.29
1.07
0.00
0.0
1.0
2.0
3.0
4.0
5.0
d/t
CDF of Maximum Imperfection
296
6.0
7.0
Figure7.32LandDimperfectionmagnitudesdescribedwithaCDF(SchaferandPekz1998)
Deviation from reference

cross section (measured at
X=6, 18 in. for the the short
specimens and X=12, 18, 30
and 36 in. for the 48 in. long
specimens).
D = max(Bi sin (i
where i=1 or 2
Nominal cross-section
1
B1
B2
90 degrees
90 degrees
Figure7.33Methodformeasuringdistortionalimperfectionmagnitudesfromexperiments
Table7.2Localanddistortionalimperfectionmagnitudes
Specimen
362-1-24-NH
362-2-24-NH
362-3-24-NH
362-1-24-H
362-2-24-H
362-3-24-H
362-1-48-NH
362-2-48-NH
362-3-48-NH
362-1-48-H
362-2-48-H
362-3-48-H
600-1-24-NH
600-2-24-NH
600-3-24-NH
600-1-24-H
600-2-24-H
600-3-24-H
600-1-48-NH
600-2-48-NH
600-3-48-NH
600-1-48-H
600-2-48-H
600-3-48-H
Type 1 Imperfection Magnitude (L)
Type 2 Imperfection Magnitude (D)
25% CDF
0.005
0.005
0.005
0.005
0.005
0.006
0.005
0.006
0.005
0.005
0.005
0.006
0.006
0.006
0.006
0.006
0.006
0.006
0.006
0.006
0.006
0.006
0.006
0.006
25% CDF
0.025
0.025
0.025
0.025
0.025
0.025
0.025
0.025
0.025
0.025
0.025
0.026
0.028
0.028
0.028
0.027
0.026
0.028
0.028
0.028
0.028
0.027
0.027
0.028
75% CDF Measured

0.025
0.038
0.025
0.054
0.025
0.036
0.026
0.052
0.025
0.058
0.026
0.044
0.026
0.071
0.026
0.080
0.026
0.057
0.026
0.084
0.026
0.066
0.026
0.050
0.029
0.061
0.029
0.075
0.029
0.096
0.028
0.062
0.027
0.089
0.028
0.087
0.029
0.071
0.028
0.077
0.029
0.073
0.028
0.095
0.028
0.049
0.028
0.068
75% CDF Measured

0.060
0.200
0.060
0.174
0.060
0.205
0.061
0.195
0.059
0.159
0.061
0.140
0.061
0.168
0.061
0.163
0.060
0.184
0.061
0.161
0.061
0.174
0.062
0.156
0.068
0.106
0.068
0.114
0.068
0.114
0.065
0.064
0.064
0.124
0.067
0.102
0.067
0.090
0.067
0.077
0.067
0.074
0.066
0.084
0.067
0.062
0.067
0.099
The initial outofstraightness of each column specimen was measured in the MTS
machineunderasmallpreloadbeforethestartofeachtest.Thisglobalimperfectionis
also superimposed on the nodal geometry for each specimen finite element model as
297
showninFigure7.34.Themagnitudeoftheglobalimperfection,g,isprovidedinTable
7.3.
Loading Line
g (+ shown)
Specimen COG (Typ.)
Section a-a
Figure7.34DefinitionofoutofstraightnessimperfectionsimplementedinABAQUS
Table7.3Outofstraightnessimperfectionmagnitudes
Specimen
362-1-24-NH
362-2-24-NH
362-3-24-NH
362-1-24-H
362-2-24-H
362-3-24-H
362-1-48-NH
362-2-48-NH
362-3-48-NH
362-1-48-H
362-2-48-H
362-3-48-H
600-1-24-NH
600-2-24-NH
600-3-24-NH
600-1-24-H
600-2-24-H
600-3-24-H
600-1-48-NH
600-2-48-NH
600-3-48-NH
600-1-48-H
600-2-48-H
600-3-48-H
298
g
in.
-0.024
0.004
0.038
-0.012
0.034
-0.023
0.047
-0.028
0.012
0.066
0.013
-0.003
-0.063
-0.141
0.063
-0.078
0.076
0.069
-0.036
-0.087
-0.049
-0.098
0.072
0.020
7.2.1.6
Residualstressesandequivalentplasticstrains
Chapter6describesageneralmethodforpredictingthethroughthicknessresidual
stressesandstrainsincoldformedsteelmemberswhichcanthenbereadilyinputinto
ABAQUS.The prediction method assumes that residual stresses and plastic strains
occuroverthefullcrosssectionfromcoiling,uncoiling,andflatteningofthesheetcoil.
The coiling residual stresses are largest when the sheet thickness t is large (>0.068 in.)
and the yield stress is low (<40 ksi). The predicted coiling, uncoiling, and flattening
residual stresses (and plastic strains) are zero in this study because the column
specimens have a relatively low sheet thickness (~0.040 in.) and high yield stress (~ 60
ksi).
Residual stresses and plastic strains from the rollforming of the crosssection are
consideredinthisstudy.ThesestressesareappliedinABAQUSwiththeelementlocal
coordinate system shown in Figure 7.35 starting from section point 1 (SNEG). The
transverse residual stress distribution (2direction) is provided in Figure 7.36 and the
longitudinaldistribution(1direction)inFigure7.37asafunctionofyieldstressyield(yield
islistedinTable5.13foreachspecimen).PlasticstrainsareinputintoABAQUSinvon
Mises space and therefore only plastic strain magnitudes are required, not a specific
direction.Thehighestinitialstrainsoccurattheinnerandoutersurfacesofthecorners
asshowninFigure7.38.pisapproximatedusingtheprocedureoutlinedinFigure6.17.
299
Element normal
SPOS
1
SNEG
2
Figure7.35ABAQUSelementlocalcoordinatesystemforusewithresidualstressdefinitions
SPOS
-0.50yield
-yield
+yield
+0.50yield
SNEG
Figure7.36Transverseresidualstressdistributionappliedatthecornersofthecrosssection
SPOS
+0.05yield
+0.50yield
-0.50yield
-0.05yield
SNEG
Figure7.37Longitudinalresidualstressdistributionappliedatthecornersofthecrosssection
300
SPOS
SNEG
Figure7.38Equivalentplasticstraindistributionatthecornersofthecrosssection
The transverse residual stress distribution has the special property that it is self
equilibratingforbothmomentandaxialforce,i.e.thetotalforceandmomentthrough
thethicknessiszero.Thisselfequilibratingcharacteristicensuresthatnodeformation
(orredistributionofstress)willoccurinABAQUSintheinitialstate.Thelongitudinal
stress distribution is selfequilibrating for axial force but not for moment. The
deformationsassociatedwiththisoutofbalancebendingmomentareinfinitesimaland
verysmallredistributionsinstressareobserved(0.1ksi)intheinitialstate.
Thenumberofelementsectionpointsthroughthethicknessdictatestheaccuracyof
the residual stress distribution. If only a small number of section points are used, the
discontinuity in stress at the middle thickness cannot be modeled accurately and
excessive transverse deformations of the crosssection will occur. Figure 7.39
demonstratesthedecreaseinunbalancedthroughthicknesstransversemoment,MUB,as
the number of section points increase (sheet thickness is assumed as t=0.040 in. and
yield=60 ksi). As the number of section points decrease, the residual stress approaches
0.50My,whereMyistheyieldmomentofthesheetsteelperunitwidthdefinedas:
301
t2
M y = yield
6
(7.4)
55 section points are used in the specimen finite element models for this study as a
compromise between model accuracy and computational cost. ABAQUS limits the
maximumnumberofsectionpointsto250fortheS9R5element(ABAQUS2007b).
0.5
0.45
0.4
0.35
MUB/My
0.3
0.25
0.2
0.15
0.1
0.05
0
20
40
60
80
# of element through-thickness section points
100
Figure7.39Influenceofsectionpointsontheunbalancedmoment(accuracy)ofthetransverseresidual
stressdistributionasimplementedinABAQUS
7.2.2 Modeling protocol validation

7.2.2.1
Ultimatestrengthandfailuremechanisms
The nonlinear finite element protocol presented in Section 7.2.1 is demonstrated to
be a viable, conservative predictor of peak load when compared to the experiment

resultsinTable7.4.ThemeanoftheexperimentalpeakloadPtesttoABAQUSpeakload
302
PABAQUSratiosare1.03(25thpercentileimperfectionCDF),1.05(75thpercentileimperfection
CDF), and 1.11 (measured imperfections). In a few cases (and always with specimens
withholes),ABAQUSwasnotabletoobtainthepeakload,eitherbecausethemodified
Riks solution algorithm reversed the direction of the applied load (similar to that
observedinFigure7.12forstiffenedelements)orbecausetheABAQUScouldnotfind
equilibriumandterminatedthesimulation.Asimperfectionmagnitudesincreased,the
modifiedRikssolutionalgorithmwasmoresuccessfulatreachingpeakload.Thistrend
is hypothesized to occur because for small imperfection magnitudes a specific
deformationpatternisnotestablishedandmanyequilibriumpathsexistnearpeakload,
whereas for larger imperfection magnitudes a dominate deformation shape and
equilibrium path are defined early in the simulation. Nonlinear FE loaddisplacement
behavioris providedforarepresentativesampleofspecimensinFigure7.40toFigure
7.47,includingtheloaddisplacementcurvesanddeformedshapeatcollapse(compare
these simulated shapes to the pictures of experiments in Appendix F). FE simulation
loaddisplacementcurvesareprovidedforallspecimensinAppendixI.
Table7.4ComparisonofnonlinearFEsimulationpeakloadstoexperiments
303
Specimen
Ptest
kips
10.48
10.51
10.15
10.00
10.38
9.94
9.09
9.49
9.48
8.95
9.18
9.37
11.93
11.95
12.24
12.14
11.62
11.79
11.15
11.44
11.29
11.16
11.70
11.16
362-1-24-NH
362-2-24-NH
362-3-24-NH
362-1-24-H
362-2-24-H
362-3-24-H
362-1-48-NH
362-2-48-NH
362-3-48-NH
362-1-48-H
362-2-48-H
362-3-48-H
600-1-24-NH
600-2-24-NH
600-3-24-NH
600-1-24-H
600-2-24-H
600-3-24-H
600-1-48-NH
600-2-48-NH
600-3-48-NH
600-1-48-H
600-2-48-H
600-3-48-H
Average
Standard deviation
DNC
25th percentile
imperfection CDF
PABAQUS Ptest/PABAQUS
kips
10.26
1.02
10.13
1.04
10.21
0.99
9.22
1.09
8.83
1.18
9.19
1.08
9.48
0.96
9.40
1.01
9.26
1.02
8.97
1.00
8.91
1.03
8.58
1.09
12.14
0.98
12.10
0.99
12.10
1.01
DNC
--11.10
1.05
DNC
--11.27
0.99
11.27
1.02
11.37
0.99
DNC
--DNC
--DNC
--1.03
0.05
75th percentile
imperfection CDF
PABAQUS Ptest/PABAQUS
kips
9.88
1.06
9.70
1.08
9.85
1.03
9.08
1.10
8.70
1.19
9.11
1.09
9.34
0.97
9.27
1.02
8.89
1.07
8.73
1.02
8.63
1.06
DNC
--12.03
0.99
12.01
1.00
11.99
1.02
11.63
1.04
11.08
1.05
11.76
1.00
11.14
1.00
11.39
1.00
11.18
1.01
10.22
1.09
DNC
--10.35
1.078
1.05
0.05
Measured imperfections
PABAQUS
kips
8.72
8.82
8.69
8.48
8.27
8.78
7.76
8.36
7.44
8.30
8.26
ED
11.83
11.74
11.64
11.45
10.82
11.49
11.32
11.30
11.04
ED
10.17
10.30
Ptest/PABAQUS
1.20
1.19
1.17
1.18
1.26
1.13
1.17
1.14
1.28
1.08
1.11
--1.01
1.02
1.05
1.06
1.07
1.03
0.98
1.01
1.02
--1.15
1.08
1.11
0.08
Did Not Complete, Abaqus terminated before finding the peak load
ED Excessive distortion - Abaqus error, imperfection magnitude causes element distortional
Theinitialelasticslopeofthe25%CDFand75%CDFFEloaddisplacementcurves
are consistent with experimental results as shown in Figure 7.40 to Figure 7.47,
demonstratingthattheelasticmaterialmodelingassumptionsandspecimendimensions
areconsistentwiththeexperiments.Theinitialslopeoftheloaddisplacementcurveis
also sensitive to imperfection magnitudes, and therefore the similarities between
experiment and the FE results confirm the assumption that the 25th and 75th percentile
imperfectionmagnitudesintheFEsimulationsproducephysicallyrealisticresults.This
is contrary to the FE simulations with measured imperfections for the 362S16233
specimens(forexample,seeFigure7.40),wheretheinitialloaddisplacementslopeand
peak load are 15% to 30% less than the experimental results (see Table 7.4 and Figure
7.40). The FE simulations for the 600S16233 specimens are much less sensitive to
imperfectionmagnitudes(forexample,seeFigure7.44).Themaximumdifferenceintest
to predicted ratio between the three imperfection levels (25% CDF, 75% CDF, and
measured)inTable7.4forthe600S16233specimensis3%.
304
The postpeak ductility of the column specimens is often underpredicted in the
ABAQUSnonlinearfiniteelementmodels.Thecollapsemechanismofacolumndictates
itsductilityandinsomecasesitspeakload.Forexample,outwarddistortionalbuckling
has been shown to produce lower column strengths than inward distortional buckling
(SilvestreandCamotim2005).ThisobservationcouldexplainwhytheFEsimulationsof
the362S16233specimenswithholes(whichexhibitoutwarddistortionalbuckling)have
alowerpeakloadandductilitythantheexperimentresults(allthreespecimensexhibit
inward distortional buckling, see Appendix F). Another factor influencing column
ductility may be the ABAQUS material modeling effect discussed in Section 7.2.1.4.
Whenplasticityisconsideredattheproportionallimit(seeFigure7.30)thepeakofthe
loaddisplacement curve is flattened which ismore consistent with experiment results.
Thesehypothesesmotivateimportantfutureworktobetterunderstandmetalplasticity
andmaterialmodelinginABAQUSandalsotheinfluenceofimperfectionshapesonFE
columnductilityandstrengthpredictions.
305
14
Experiment
FE, 25% Imperfection CDF
FE, 75% imperfection CDF
FE, measured imperfections
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
0.16
0.18
0.2
75% Imperfection CDF
Figure7.40Loaddisplacementresponseofspecimen362124NH
14
Experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
0.16
0.18
0.2
Figure7.41Loaddisplacementresponseofspecimen362124H
306
14
Experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
0.16
0.18
0.2
14
Experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
0.16
0.18
0.2
307
14
Experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
0.16
0.18
0.2
14
Experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
0.16
0.18
0.2
308
14
Experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
0.16
0.18
0.2
14
Experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
0.16
0.18
0.2
309
7.2.2.2
Influenceofresidualstressesandinitialplasticstrains
Residualstresses(RS)andinitialplasticstrains(PS)fromthemanufacturingprocess
are approximated with the prediction method in Chapter 6 and then input into
ABAQUSasdiscussedinSection7.2.1.6.Figure7.48highlightstheireffectontheload
deformation response of specimen 600124NH. A small increase in peak load
(approximately 2%) is observed when just initial plastic strains are considered at the
corners, which simulates the increase in apparent yield stress from strain hardening.
Theincreaseinstrengthisminimalbecausetheinfluenceofthestiffenedcornersisoffset
bythelargeproportionofunformedsteel(i.e.,flats)inthecrosssection.Thetransverse
andlongitudinalresidualstressescreatedbycoldbendingofthecrosssectionalsohave
aminimalimpactontheloaddeformationresponseforthisspecimen,primarilybecause
theplasticstrainsatthecornersarehigh(pispredictedtobelargeas0.20atthecorner
outerfibers)whichincreasestheapparentyieldstressinABAQUSandpreventsalossin
stiffnessatthecorners,evenwiththepresenceofthethroughthicknessresidualstresses
inthecolumn.Similarloaddisplacementtrendsarealsoobservedforspecimen3621
24NHasshowninFigure7.49.
Residual stresses and plastic strains are expected to have a larger influence on the
ultimate strength of members with crosssections made from thicker sheet steel, since
coiling and uncoiling of the sheet steel will impart residual stresses and plastic strains
around the entire crosssections (see Figure 6.19). Future research is planned to study
the influence of throughthickness residual stresses and plastic strains on yielding
310
patterns and failure modes of coldformed steel members. The ABAQUS metal
plasticitymodelwithisotropicversuskinematichardeningalsoneedsfurtherstudyto
determine if one is better than the other when considering the influence of residual
stressesandstrains(seeSection6.7foramoredetaileddiscussion).
14
without RS or PS
with RS and PS
experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
0.16
0.18
0.2
Figure7.48Influenceofresidualstresses(RS)andplasticstrains(PS)ontheFEloaddisplacementresponse
ofspecimen600124NH
311
14
without RS or PS
with RS and PS
experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
0.16
0.18
0.2
Figure7.49Influenceofresidualstresses(RS)andplasticstrains(PS)ontheFEloaddisplacementresponse
ofspecimen362124NH.
312
Chapter 8
The Direct Strength Method for coldformed steel members with holes
ThenonlinearfiniteelementcapabilitydevelopedinChapter7isnowemployedto
evaluate proposed Direct Strength Method (DSM) formulations for coldformed steel
members with holes. Several hundred coldformed steel columns and beams with
standardSSMAstructuralstudcrosssections(SSMA2001)andwithvaryingwebhole
sizes,shapes,andspacingsaresimulatedtocollapseinABAQUS.Theelasticbuckling
properties of these members (Pcrl, Pcrd, and Pcre for columns and Mcrl, Mcrd, and Mcre for
beams),includingthepresenceoftheholes,areapproximatedwiththeCUFSMelastic
buckling prediction methods developed in Chapter 4. The corresponding ultimate
strengths(obtainedfromtheABAQUSsimulations)aremergedwiththeelasticbuckling
informationintoasimulatedexperimentsdatabasewhichisutilizedtoidentifypotential
modifications to the existing DSM local, distortional, and global failure prediction
curves. Specific DSM options are proposed from these comparisons, which are then
313
comparedtotheexperimentelasticbucklingandtestedstrengthdatabasesinChapter4
to formalize the final proposed DSM recommendations for coldformed steel members
withholes.
8.1 DSMforcolumnswithholes
8.1.1 Database of simulated column experiments
Simulated experiments were conducted on 211 Csection columns with evenly
spaced slotted or circular web holes in ABAQUS. Column lengths and crosssections
werespecificallyselectedwithcustomMatlabcodeemployingtheexistingDSMdesign
curves to identify columns predisposed to local, distortional, and global buckling type
failures. The crosssections were chosen from a catalog of 99 industry standard C
sections published by the Steel Stud Manufacturers Association (SSMA 2001). The
nominal outtoout dimensions provided in the SSMA catalog were converted to
centerline dimensions and then constructed in ABAQUS with the meshing procedure
describedinSection7.2.1.1.Evenlyspacedcircularorslottedwebholeswereplacedin
the columns with hole spacing S (defined in Figure 3.2) varying between 12 and 22
inches.Theholeswerecenteredtransverselyinthewebandtheirdepth,hhole,wasvaried
suchthattheratioofthenetcrosssectionalarea,Anet,tothegrosscrosssectionalarea,Ag,
rangedbetween0.60and1.0.
TheABAQUS boundaryconditionsandapplicationofloading,describedinFigure
8.1,areimplementedtobeconsistentwithCUFSM,i.e.pinnedpinnedandfreetowarp
with a uniform stress applied at the member ends. These boundary conditions were
314
specifically chosen to permit the use of CUFSM simplified elastic buckling methods
whenpredictingtheelasticbucklingbehaviorofcolumnswithholes.(Ifpinnedpinned
warpingfixed end conditions or fixedfixed end conditions were used the elastic
buckling predictions would have required modifications factors, see Eq. (4.8) for an
example). CUFSMboundaryconditionsrepresentalower boundon memberstrength
arethereforeconsideredconservativeindesign.Consistentnodalloadsareappliedto
simulate the uniform compressive stress at the column ends (see Section 7.2.1.2 for
information on S9R5 consistent nodal loads). The loads (areference load of 1 kip was
appliedateachendinABAQUS)aredistributedoverthefirsttwosetsofcrosssection
nodestoavoidlocalizedfailuresattheloadededges.
End cross-section nodes
restrained in 2 and 3
Node centered in flange at

longitudinal midline restrained in 1
(to prevent rigid body motion)
Consistent nodal loads applied

over two sets of cross-section
nodes to avoid edge failures
1
4
5
2
Figure8.1ABAQUSsimulatedcolumnexperimentsboundaryconditionsandapplicationofloading
The ABAQUS simulations were performed with the modified Riks nonlinear
solution algorithm. Automatic time stepping was enabled with a suggested initial arc
315
lengthstepof0.25(theRiksmethodincrementsinunitsofenergy,inthiscasekipin.),a
maximumstepsizeof0.75,andthemaximumnumberofsolutionincrementssetat300.
Metal plasticity was simulated with the material modeling procedure described in
Section7.2.1.4.Theplastictruestressstraincurveforspecimen362148HinAppendix
Hwasassumedforallcolumnmodels(butmodifiedsothatplasticitystartsattheyield
stress,seeSection7.2.1.4),wherethesteelyieldstressFy=58.6ksi.Residualstressesand
initialplasticstrains,asdiscussedinSection7.2.1.6,werenotconsideredintheABAQUS
models because their implementation requires further validation and they were not
observed to markedly influence column ultimate strength (see Figure 7.48 and Figure
7.49).
Imperfections were imposed on the initial column geometry in ABAQUS with
custom Matlab code which combines the local, distortional buckling, and global cross
section mode shapes from CUFSM along the column length. Two simulations were
performed for each column, one model with 25% CDF local and distortional
imperfectionmagnitudesandL/2000globalimperfections(whereListhelengthofthe
column) and the other model with 75% CDF local and distortional imperfection
magnitudesandaglobalimperfectionmagnitudeofL/1000(seeSection7.2.1.5forlocal
anddistortionalimperfectiondefinitions).
Theglobalimperfectionmagnitudeassumptionsarebasedonhotrolledcolumnout
ofstraightness measurements (Galambos 1998b) because no formal guidelines are

currently available for coldformed steel columns. The use of hotrolled steel column
imperfection magnitudes is consistent with the DSM approach for global buckling
316
controlledfailures.DSMemploysthesameglobaldesigncurveasthatspecifiedbythe
Structural Stability Research Council (SSRC) for hotrolled steel (Galambos 1998b),
thereby indirectly assuming that the influence of hotrolled steel global imperfection
magnitudes also apply to coldformed steel. The global imperfection shape of the
columns in the simulation database was eitherweakaxis flexural buckling or flexural
torsionalbuckling,dependingonthecrosssectiondimensionsandlengthofthecolumn.
Csectionsarenotsymmetricabouttheirweakbendingaxis,andthereforethedirection
of the global imperfection influences the predicted strength when weakaxis flexural
bucklingdefinestheglobalimperfectionshape(e.g.,webincompressionfrombowing
or flange lips in compression from bowing). Simulations with both L/1000 and
L/2000 imperfection magnitudes were performed to capture this strength effect for
weakaxisflexuralbucklingmodeshapes.Globalimperfectionswerenotconsideredfor
columns with L/D18 (i.e., stockier columns with a low sensitivity to global
imperfections),whereDistheouttooutflangewidthofthecolumn.
Thelocal(Pcrl),distortional(Pcrd),andglobal(Pcre)criticalelasticbucklingloadswere
predicted for each column with custom Matlab code based on the CUFSM prediction
methods described in Section 4.2.7. The database of simulated column experiments,
including crosssection type, column and hole geometry, simulated ultimate strength
(Ptest25 and Ptest75) and critical elastic buckling loads for each column (including the
presenceofholes)isprovidedinAppendixK.
317
8.1.2 Distortional buckling study
Agroupof20columnsfromtheSSMAcolumnsimulationdatabasewaschosento
evaluatetheinfluenceoftheratioAnet/Ag onthetestedstrengthofcolumnspredictedto
collapsewithadistortionalfailuremode.Agisthegrosscrosssectionalareaofacolumn
and Anet is the crosssectional area at the location of a hole. In this study the column
length,L,isheldconstantat24in.andthecolumnwidthsrangefrom6in.to12in.The
SSMA crosssections chosen have relatively thick sheet steel (t up to 0.1017 in.) which
preventsalocalbucklingtypefailure.Thewebofeachcolumnhastwocircular holes
wheretheholespacingS=12in(seeFigure3.2forthedefinitionofS).Theholedepth
(diameter),hhole,isvariedforeachcolumntoproduceAnet/Agof1.0(noholes),0.9,0.8,0.7,
and 0.6. Refer to Appendix K, Study Type D, for specific crosssection and hole
geometry information for each column. Figure 8.2 provides an example of an SSMA
600S25097structuralstudcolumnconsideredinthestudy.
SSMA 800S250-97 structural stud column
Anet/Ag
1.0
0.90
0.80
0.70
0.60
Figure8.2SSMA800S25097structuralstudwithwebholesconsideredintheDSMdistortionalbuckling
study
ThesimulationresultsforAnet/Ag=1.0,0.9,0.8,0.7,and0.6arecomparedtotheDSM
distortionalbucklingpredictioncurveinFigure8.4toFigure8.8.Thecolumnstrengths,
Ptest25 and Ptest75, without holes (Anet/Ag=1.0)are consistent with theDSM design curve as
318
showninFigure8.4a,confirmingtheviabilityofthenonlinearsimulationprotocol.The
mean and standard deviation of the simulated test to predicted ratio is 1.10 and 0.10
respectivelyfor25%CDFlocalanddistortionalimperfections,and1.06and0.13for75%
CDF imperfections (global imperfections are not considered in these stocky columns).
For the columns with holes, the simulated test strengths diverge from the DSM
predictioncurveasdistortionalslenderness, d=(Pyg/Pcrd)0.5,decreasesasshowninFigure
8.5atoFigure8.8a(Pygisthesquashloadofthecolumncalculatedwiththegrosscross
sectional area Ag). This divergent trend in Ptest with decreasing d can be explained as
follows.When dishigh(i.e.PcrdislowrelativetoPyg),thecolumnstrengthislowerthan
Pygbecausethecollapsemechanismiscontrolledbydistortionalbucklingdeformations.
The presence of a hole may decrease Pcrd (as predicted with the method in Section
4.2.7.2), but the distortional failure mechanismstill dominates in this case. When d is
low, Pcrd is much higher than Pyg and the column is not as sensitive to distortional
deformation.Instead,thecolumnfailsbyyieldingofthecrosssection.Whenaholeis
added,theyieldingofthecrosssectionoccursatthelocationofthehole(i.e.,atthenet
section) resulting in the collapse of the unstiffened strips adjacent to the hole. This
collapse is accompanied by distortional and global deformations caused by the
reduction in stiffness at the net section. These two column failure mechanisms, a
distortional buckling failure (when d is high) and yielding and collapse of the net
section(whendislow),arecomparedinFigure8.3.
319
Yielding and collapse of the unstiffened strips adjacent

to the holes accompanied by distortional and global
(weak axis flexural) deformation
Distortional buckling failure
Anet/Ag
1.0
0.90
0.80
0.70
0.60
Ptest25/Pyg
0.60
0.59
0.57
0.54
0.45
1.33
1.36
1.39
1.42
1.45
Figure8.3SSMA800S25097structuralstudfailuremodetransitionfromdistortionalbucklingtoyieldingat
thenetsection
TheobservationsfromthisstudyareusedtoformulateamodifiedDSMdistortional
curve for columns with holes which captures the failure mechanism transition from
yielding at the net crosssection to a distortional type failure mode and limits the
strengthofthecolumntoitssquashloadatthenetsection:
The nominal axial strength, Pnd, for distortional buckling shall be calculated in accordance with the
following:
(a) For d d1
Pnd = Pynet
(caponcolumnstrength)
(b) For d1<dd2

Pnd = P Pynet Pd 2
ynet

d1
d2
(yieldcontroltransition)
(c) For d > d2
Pnd = 1 0.25 crd

Py
where
0 .6
Pcrd
P
y
0 .6
(existingDSMdistortionalcurve)
Py
d1
= 0.561(Pynet Py )
Py Pcrd
d2
= 0.561 14 (Pynet Py )0.4 13
Pd2
= 1 0.25 (1 d 2 )
1.2
)(1
1 .2
d2
Py
Pynet = FyAnet0.6Py
= Column cross-sectional area at the location of hole(s)
Anet
Pcrd
= Critical elastic distortional column buckling load including hole(s)
320
ThemodifiedDSMdistortionalcurveisaddedinFigure8.5btoFigure8.8basAnet/Ag
decreases,simulatingthetransitionfromtheexistingDSMcurvetothecappedcolumn
strength exhibited by the simulated test data. The linear portion of the modified
prediction curve represents the unstiffened strip distortional collapse mechanism and
thenonlinearportionrepresentsacollapsemechanismdrivenbydistortionalbuckling.
ThisproposedmodificationtotheDSMdistortionalpredictioncurvewillbecompared
againstthecolumnexperimentsdatabasedevelopedinSection4.2.6.1asapartofseveral
proposedDSMoptionsconsideredlaterinthischapter.
1.4
1.4
DSM (no hole)
FE 25% CDF imperfections
0.8
0.8
0.6
0.6
0.4
0.4
0.2
0.2
0.5
1.5
2.5
3.5
DSM (no hole)

DSM (proposed, with holes)
1.2
Ptest/Py
Ptest/Py
1.2
0.5
distortional slenderness, d=(Py /P crd)0.5
1.5
2.5
3.5
Figure8.4Comparisonofsimulatedcolumnstrengths(Anet/Ag=1.0)to(a)theexistingDSMdistortional
bucklingdesigncurveandto(b)theproposedDSMdistortionalbucklingcurveforcolumnswithholes
1.4
1.4
DSM (no hole)
0.8
0.8
0.6
0.6
0.4
0.4
0.2
0.2
0.5
1.5
2.5
3.5
DSM (no hole)

1.2
Ptest/Py
Ptest/Py
1.2
0.5
1.5
2.5
3.5
321
1.4
1.4
DSM (no hole)
0.8
0.8
0.6
0.6
0.4
0.4
0.2
0.2
0.5
1.5
2.5
3.5
DSM (no hole)

1.2
Ptest/Py
Ptest/Py
1.2
0.5
1.5
2.5
3.5
1.4
1.4
DSM (no hole)
0.8
0.8
0.6
0.6
0.4
0.4
0.2
0.2
0.5
1.5
2.5
3.5
DSM (no hole)

1.2
Ptest/Py
Ptest/Py
1.2
0.5
1.5
2.5
3.5
1.4
1.4
DSM (no hole)
0.8
0.8
0.6
0.6
0.4
0.4
0.2
0.2
0.5
1.5
2.5
3.5
DSM (no hole)

1.2
Ptest/Py
Ptest/Py
1.2
0.5
1.5
2.5
3.5
322
8.1.3 Global buckling study
This study compares simulated strengths to DSM predictions of coldformed steel
columnswithholespredictedtoexperienceaglobalfailure.Aglobalfailureistriggered
by yielding for a stocky column and flexural or flexuraltorsional buckling for slender
columns.NomodificationsareproposedtotheDSMglobalbucklingdesigncurvefor
columns with holes, as the influence of holes on short columns will be accounted for
withtheDSMlocalbucklingdesigncurve(seeSection8.1.4).Forexample,whenPne=Pyg,
Pnl will always be made less than or equal to Pynet, and therefore the nominal column
strength,Pn,willalwaysbelessthanorequaltoPynet.
Agroupof18columnspredisposedtoaglobalfailurewereselectedfromtheSSMA
columnsimulationdatabase.Inthisstudythecolumnlength,L,variedfrom8in.to96
in. to consider a wide range of global column slenderness, c=(Pyg/Pcre)0.5. The SSMA
crosssections are purposely selected with low local buckling slenderness (i.e., sections
withthickersheetsteeluptot=0.1017in.andrelativelynarrowflangesandwebs).DSM
predictsthatlocalbucklingdoesnotinfluenceglobalbucklingbehaviorwhen l0.776.
ThewebofeachcolumncontainsevenlyspacedslottedholeswheretheholespacingS
variesfrom8in.to22in.Theholelength,Lhole,isheldconstantat4in.,whilethehole
depth, hhole, is varied for each column to produce Anet/Ag of 1.0 (no holes), 0.9, and 0.8.
(Refer to Appendix K, Study Type G, for specific column crosssection and hole
geometry information.) The four columns with the lowest global slenderness (for
example, Specimen ID # 137 to 140 in Appendix K) were modeled with circular holes
insteadofslottedholesbecausetheslottedholesresultedinimpracticalcolumnlayouts,
323
with the hole extending over more than 50% of the column length. The global
imperfectionshapeforfiveofthelongercolumnswasweakaxisflexuralbuckling,and
therefore four simulated strengths are determined for these columns (instead of the
typical two): 25% CDF local and distortional imperfections with L/2000 global
imperfections and 75% CDF local and distortional imperfections with L/1000 global
imperfections.
Figure8.9toFigure8.11comparethesimulatedcolumnstrengthstotheDSMglobal
predictioncurveasAnet/Agdecreases.Thesimulatedstrengthsforcolumnswithoutholes
areconsistentwiththeDSMglobalpredictioncurveasshowninFigure8.9a.Themean
and standard deviation of the simulated test to predicted ratio for columns without
holes is 1.06 and 0.05 respectively for 25% CDF local and distortional imperfections
L/2000 global imperfection and 0.95 and 0.07 for 75% local and distortional
imperfectionsL/1000globalimperfection.
Figure8.10aandFigure8.11ademonstratethatforcolumnswithholes,thepredicted
strengthsareconsistentwiththeDSMglobaldesigncurvewhenglobalslenderness cis
greater than 2. Most of the columns in this region fail by weakaxis flexural buckling.
Whencisbetween1and2,allofthecolumnsfailbyflexuraltorsionalbucklingandthe
simulatedcolumnstrengths(with25%CDFimperfections)are20%higherthantheDSM
predictions. This conservative trend is caused by the simplified prediction method
developed in Section 4.3.2.3, which is know tobe a conservative predictor of Pcrewhen
torsionalbucklinginfluencestheglobalbucklingmode.WhenPcreisunderpredicted,the
global slenderness increases, which shifts the tested data off of the DSM design curve;
324
theshiftisespeciallyclearinFigure8.11a.Thisobservationfurthermotivatesthefuture
worktostudytheinfluenceofholesonthewarpingtorsionconstant,Cw.
Simulated column strengths diverge below the DSM prediction curve when c
decreases and Anet/Ag increases as shown in Figure 8.10a and Figure 8.11a. These
columns are short, ranging in length from 8 in.to 26 in., and exhibit a yielding failure
mode at the net section, similar to that observed in the distortional failure study in
Figure 8.3. This observation supports the proposed modification to the DSM
distortionalbucklingcurve,whichaccuratelypredictsthestrengthsofthesecolumnsas
shown in Figure 8.9b and Figure 8.11b, where the diverging data points are plotted
againstthemodifiedDSMdistortionalpredictioncurve.Thisobservationreiteratesthe
conclusion drawn in the distortional buckling study, that yielding and collapse of the
unstiffenedstripsadjacenttoaholeinfluencebothdistortionalandglobalfailuremodes
asslendernessdecreases.
1.4
1.4
DSM (no hole)
no holes, FE 25% CDF imperfections
0.8
0.8
0.6
0.6
0.4
0.4
0.2
0.2
0.5
1.5
2.5
3.5
DSM (no hole)

1.2
Ptest/Py
Ptest/Py
1.2
global slenderness, c =(Py /Pcre)0.5
0.5
1.5
2.5
3.5
Figure8.9Comparisonofsimulatedcolumnstrengths(Anet/Ag=1.00)to(a)theexistingDSMglobalbuckling
designcurveandto(b)theproposedDSMdistortionalbucklingcurveforcolumnswithholes
325
1.4
1.4
DSM (no hole)
slotted holes, FE 25% CDF imperfections
circular holes, FE 25% CDF imperfections
1.2
0.8
Ptest/Py
Ptest/Py
0.6
0.8
0.6
0.4
0.4
0.2
0.2
0.5
1.5
2.5
3.5
DSM (no hole)

1.2
0.5
1.5
2.5
3.5
Figure8.10Comparisonofsimulatedcolumnstrengths(Anet/Ag=0.90)to(a)theexistingDSMglobalbuckling
designcurveandto(b)theproposedDSMdistortionalbucklingcurveforcolumnswithholes
1.4
1.4
DSM (no hole)
1.2
0.8
Ptest/Py
Ptest/Py
0.6
0.8
0.6
0.4
0.4
0.2
0.2
0.5
1.5
2.5
3.5
DSM (no hole)

1.2
0.5
1.5
2.5
3.5
Figure8.11Comparisonofsimulatedcolumnstrengths(Anet/Ag=0.80)to(a)theexistingDSMglobal
8.1.4 Local buckling study
The distortional buckling failure study in Section 8.1.2 and the global buckling
failurestudyinSection8.1.3demonstratedthatthepresenceofholesdecreasesultimate
strengthwhencoldformedsteelcolumnsfailbyyieldingandcollapseoftheunstiffened
stripsadjacenttoaholeatthenetcrosssection.Holeswereobservedtohaveaminimal
influence on ultimate strength when the column failure mode was dictated by elastic
buckling. Thegoalofthisstudyistodetermineifthistrendisconsistentforcolumns
withholesexperiencinglocalglobalbucklinginteractionatfailure.
326
ElevencolumnsfromthesimulationdatabaseinAppendixKwerechosenforthis
study. The columns have SSMA crosssections and lengths which result in a local
bucklingslenderness, l,rangingfrom0.8to3.0.Thecolumnlength,L,variesfrom24
in. to 88 in. and column widths range from 3.5 in. to 12 in. The web of each column
containsevenlyspacedcircularholeswheretheholespacingSvariesfrom12in.to17
in.Theholedepth(diameter),hhole,isvariedforeachcolumntoproduceAnet/Agof1.0(no
holes),0.80,and0.65.RefertoAppendixK,StudyTypeL,forspecificcolumncross
sectionandholegeometryinformation.
Thesimulatedultimatestrengthsofthe11columnswithoutholes,Ptest,arecompared
totheDSMlocalbucklingstrengthprediction,Pnl,inFigure8.12.Thesimulatedtestto
predicted ratios are more variable than those observed in the distortional and global
failure studies but on average are close to unity, with a trend towards increasingly
conservative predictions with increasing l as shown in Figure 8.12a. The mean and
standarddeviationofthesimulatedtesttopredictionratiois1.05and0.14respectively
for25%CDFlocalanddistortionalimperfectionsL/2000globalimperfectionsand1.03
and0.15for75%localanddistortionalimperfectionsL/1000globalimperfections.
327
1.4
1.2
1.2

0.6
0.8
0.4
0.2
0.2
0.5
1.5
2.5
3.5
local slenderness, =(Pne/Pcr )0.5

l
l

0.6
0.4
0.8
Ptest/Pn
Ptest/Pn
1.4
0.5
1.5
2.5
3.5
Figure8.12Comparisonofcolumntesttopredictionratiosforcolumns(Anet/Ag=1.0)failingbylocalglobal
bucklinginteractionasafunctionof(a)localslenderness(b)globalslenderness
Figure8.13andFigure8.14comparethesimulatedstrengthsofthe11columnstothe
predictedstrength,Pnl,asAnet/Agdecreasesfrom1.0(nohole),to0.80,to0.65.(InFigure
8.13andFigure8.14theteststrengthsarethoseassociatedwithPtest25+inAppendixK,i.e.
the 25% CDF local and distortional imperfection magnitudes and +L/2000 global
imperfection magnitudes.) Figure 8.13a compares the simulated strengths Ptest25+ to Pnl
without the influence of holes. (The localglobal buckling interaction complicates the
comparisonbecausePnlandlarebothafunctionofPne.ByinitiallyassumingthatPneis
notinfluencedbythehole,theeffectofholesizeonsimulatedstrengthismoreclearly
observed.)Aslocalslenderness(l)decreasesinFigure8.13a(i.e.,theinfluenceoflocal
bucklingonmemberstrengthdecreases)thetestedstrengthbecomesmoresensitiveto
increasing hole size (i.e., decreasing Anet/Ag), diverging below the prediction Pnl by as
muchas40%whenl=0.75.
Figure8.14ademonstratesthatthesensitivityofcolumnstrengthtoadecreaseinAnet/Ag
isrelatedtotheratioofPynettoPne.WhenPynet/Pneishigh,thestrengthsensitivitytoAnet/Ag
328
islowbecauseglobalbucklinginitiatesthecolumnfailure.AsPne/Pynetapproachesunity,
column failure is initiated by unstiffened strip buckling and yielding at the net cross
section and therefore the sensitivity of column strength to Anet/Ag increases. A column
with the largest drop in strength with increasing hole size is the SSMA 350S16268
column with L=34 in. and S=17 in. shown in Figure 8.15. In this case a large hole
(Anet/Ag=0.65) causes the collapse of the net section resulting in an unfavorable and
suddenweakaxisflexuralfailureanda42%strengthreductionwhencomparedtothe
same column without holes. The SSMA 350S16268 column with smaller holes
(Anet/Ag=0.80)failsinacombinationofdistortionalandflexuraltorsionalbucklingwitha
12%strengthreduction.
Figure8.13bplotsthesameinformationasFigure8.14a,exceptnowPneiscalculated
usingPcre includingtheinfluenceofholes.For8outofthe11columns,theprediction
Pnl shifts from unconservative to slightly conservative, even for large holes. One
exception is the SSMA 800S25043 column with L=74 in. and S=12 in. shown in Figure
8.16,wherethestrengthpredictionbecomesoverlyconservativeasAnet/Agincreases.Pcre
ispredictedtodecreaseby45%whenAnet/Ag=0.65,althoughthetestedstrengthdecreases
by only 10%. Figure 8.16 demonstrates that the Csection web is susceptible to local
buckling, and that the presence of holes does not adversely affect the failure mode in
this case. The strengthpredictions for the SSMA 350S16268 column (Figure 8.15)and
theSSMA350S16254columnwithL=24in.andS=12in.areviablewhenAnet/Ag=0.80,but
are underestimated by 20% with Option 4 (5) when Anet/Ag=0.65 because of the
introductionofanunstableweakaxisflexuralfailuremodetriggeredbythecollapseof
329
the net section. A hinge forms at the location of the net section, and the global
slendernessishighenoughthatthecolumnbecomessusceptibletoaflexural buckling
mode. These hinge failures are not observed in the distortional buckling study (see
Section8.1.2)becausetheglobalslendernessofthecolumnsislower(i.e.theweakaxis
flexural stiffness is higher), avoiding a global buckling failure. Option 6 accurately
predicts the strength of the SSMA 350S16268 column and the SSMA 350S16254
columnsbecausethemethodassumesthattheglobalstrength,Pne,isreducedbyPynet/Py.
1.5
SSMA 800S250-43 column
1.5
Anet/Ag
1.0
1
1
Ptest/Pn
Ptest/Pn
0.80
0.65
0.5
0.5
SSMA 350S162-68 column
0.5
1.5
2.5
3.5
local slenderness, =(Pne/Pcr )0.5

l
l
0.5
1.5
2.5
3.5
local slenderness, =(P ne/Pcr )0.5

l
l
Figure8.13Comparisonofcolumntesttopredictionratiosforcolumnsfailingbylocalglobalbuckling
interactionwithPnecalculated(a)withouttheinfluenceofholes(b)andwiththeinfluenceofholes
330
1.5
1
Ptest/Pn
Ptest/P n
1.5
Increase in point size,

Anet/Ag=1.0,0.80,0.65
0.5
0.5
1.5
2.5
P ynet/Pne
3.5
0.5
4.5
0.5
1.5
2.5
P ynet/Pne
3.5
4.5
Figure8.14Comparisonofcolumntesttopredictionratiosforcolumnsfailingbylocalglobalbuckling
interactionasafunctionofPynet/PnewherePneiscalculated(a)withouttheinfluenceofholes(b)andwiththe
influenceofholes
Distortional,
flexural-torsional failure
Distortional,
flexural-torsional failure
Weak-axis flexural failure
Anet/Ag
1.0
0.80
0.65
Ptest/Ptest,no hole
1.0
0.88
0.58
Figure8.15SSMA350S16268columnfailuremodechangesfromdistortionalflexuraltorsionalbuckling
failuretoweakaxisflexureasholesizeincreases(L=34in.)
331
Local global interaction

at failure
Local global interaction

at failure
Unstiffened strip buckling

global interaction at failure
Anet/Ag
1.0
0.80
0.65
Ptest/Ptest,no hole
1.0
0.95
0.90
Figure8.16SSMA800S25043(L=74in.)columnweblocalbucklingchangestounstiffenedstripbucklingat
peakloadasholesizeincreases
Theobservationsfromthisstudyarenowemployedtoproposetwooptionsforthe
DSM local buckling design curve for columns with holes. The presence of holes
influenced the tested strength of the coldformed steel columns over the full range of
localslendernessconsidered.Thisresultwasdifferentfromthedistortionalandglobal
failurestudies,whereholeswereobservedtoreducestrengthonlyfromthecollapseand
yielding at the net section as slenderness decreased. The strength reduction from the
holeswaspredictedinDSMfor8outofthe11columns,whenPnlwascalculatedwith
Pneincludedtheinfluenceofholes(compareFigure8.13atoFigure8.13b).Atransition
similartothatproposedfortheDSMdistortionaldesigncurveisstill justifiedthough,
especiallywhenPynet/Pne1(seeFigure8.14b),tocapturetheyieldingandcollapseatthe
netsectionobservedincolumnswithlowlocalandglobalslenderness.Thestrengthof
twocolumnswithlargeholeswereunderpredictedbecauseofunstableglobalcollapse
initiatedbyyieldingatthenetsection,motivatingtheimplementationofalimitonhole
332
size(i.e.,Anet/Ag)toensuretheviabilityoftheDSMapproach.Twomodificationoptions
areproposedfortheDSMlocaldesigncurvebasedontheseconclusions:
Local Buckling (Option A)
The nominal axial strength, Pnl, for local buckling shall be calculated in accordance with the following:
(a) For l l1
Pnl = Pne Pynet
(b) For l1<ll2
P
( l l1 )
ynet
l2
Pnl = P
ynet
l1
l2
P
(c) For l > l2
Pnl = 1 0.15 Pcrl
0.4
Pne
where
l
P
crl
Pne
(yieldtransitionwhenPynet/Pne1)
0. 4
Pne
(DSMlocalbucklingcurve,unchanged)
= Pne Pcrl
l1
= 0.776(Pynet Pne ) 0.776
l2 = 0.776 1.7 (Pynet Pne )1.6 0.7 , Pynet/Pne1
= 0.776, Pynet/Pne>1
= 1 0.15(1 l 2 )0.8 (1 l 2 )0.8 Pne
Pynet = FyAnet0.6Py
Pl2
(notransitionwhenPynet/Pne>1)
(limitreductionofthenetsectionto0.6Py)
Anet = Column cross-sectional area at the location of hole(s)

Pcrl = Critical elastic local column buckling load including hole(s)
Local Buckling (Option B)

(a) For l l1
Pnl = Pynet (Pne /Py )
(b) For l1<ll2
Pynet (Pne Py ) Pl 2
(l l1 )
P
l 2 l1
Pnl = P Pne
ynet
(c) For l > l2
Pnl = 1 0.15 Pcrl
Pne
where
l
l1
l2
0.4
P
crl
Pne
0. 4
Pne
= Pne Pcrl
= 0.776(Pynet Py )
= 0.776 1.7 (Pynet Py )1.6 0.7
= 1 0.15(1 l 2 )
Pynet = FyAnet0.6Py
Pl2
0.8
)(1
l2
0.8
Pne
333

Option A imposes a transition from the DSM local buckling curve to column
strengthatthenetsection,Pynet,whenPynet<PneasshowninFigure8.17aforthecasewhen
Pne=Py (i.e., stub columns) and Pynet=0.8Pyg. When Pynet>Pne, Option A assumes that holes
influence only the critical elastic buckling loads (Pcrl, Pcre) and otherwise do not change
thefailuremodeofthecolumn;thiscaseisdemonstratedinFigure8.17cwhenPcre=Pyg.
OptionBalsoimposesatransitiontothenetcolumnstrengthfromtheDSMlocalfailure
curve,althoughthetransitionisassumedtooccurforall valuesofPynet/Pne.Inessence,
theOptionBcurveforstubcolumnsshowninFigure8.17aisscaleddownbasedonthe
ratio Pynet/Py. The result is an additional reduction in predicted strength for global
column failures without local buckling interaction that is not captured by Option A.
This difference between Option A and Option B is highlighted in Figure 8.17b, where
Pynet=0.8Pyg and Pcre = 5Pyg. The validity of both options are evaluated in the following
section against the simulation database and the experiment database assembled in
Chapter4.
1.4
1.4
DSM local curve (no hole)
DSM local curve (Option A)
DSM local curve (Option B)
1.2
1.4
1.2
Pcre=100Pyg
Pcre=5Pyg
0.8
0.6
0.6
0.4
0.4
0.4
0.2
0.2
0.2
0.5
1.5
2.5
local slenderness, =(P ne/P cr )0.5

l
l
3.5
0.5
1.5
2.5

l
l
Pynet=0.8Pyg
Pn /Py
0.6
Pcre=Pyg
0.8
Pynet=0.8Pyg
Pn /Py
Pynet=0.8Pyg
Pn /Py
0.8

1.2
3.5
0.5
1.5
2.5
3.5

l
l
Figure8.17ComparisonofDSMlocalbucklingdesigncurveoptionswhenPynet=0.80Pygand(a)Pcre=100Pyg,
(b)Pcre=5Pyg,and(c)Pcre=Pyg
334
8.1.5 Presentation and evaluation of DSM options
SixoptionsforextendingDSMtocolumnswithholesareevaluatedinthissection.
The options range from simple substitutions in the existing code to more involved
modifications, including the incorporation of the design curve transitions discussed in
Section8.1.2andSection8.1.4fordistortionalandlocalbuckling.
8.1.5.1
DescriptionofDSMoptions
Option 1: Include hole(s) in Pcr determinations, ignore hole otherwise

This method, in presentation, appears identical to currently available DSM expressions
The nominal axial strength, Pne, for flexural, or torsional- flexural buckling is
2
for c 1.5 Pne = 0.658 c Py
0.877
for c > 1.5 Pne = 2 Py = 0.877Pcre

c
where
Py Pcre
Py
= AgFy
Pcre= Critical elastic global column buckling load (including hole(s))
Ag
Local Buckling
335

P
for l > 0.776 Pnl = 1 0.15 crl
Pne
where
0 .4
P
crl
Pne
0.4
Pne
Pne Pcrl
Pcrl = Critical elastic local column buckling load (including hole(s))

Pne is defined above.
for d 0.561 Pnd = Py
for d > 0.561 Pnd = 1 0.25 crd

Py
where
0 .6
Pcrd
P
y
0 .6
Py
= Py Pcrd
Pcrd = Critical elastic distortional column buckling load (including hole(s))
336
Option 2: Include hole(s) in Pcr determinations, Use Pynet everywhere

The only change in this method is to replace Py with Pynet
for cnet 1.5 Pne = (0.658
for cnet > 1.5
where
cnet
2
cnet
)P
ynet
0.877
Pne = 2 Pynet = 0.877 Pcre
cnet
Pynet Pcre
Pynet = AnetFy
= net area of the column
Anet
Local Buckling
P
for l > 0.776 Pnl = 1 0.15 crl
Pne
where
0 .4
P
crl
Pne
0.4
Pne
Pne Pcrl

Pne is defined above
for dnet 0.561 Pnd = Pynet
for dnet > 0.561 Pnd = 1 0.25 crd

Pynet
where
dnet
0.6
Pcrd
Pynet
0.6
Pynet
= Pynet Pcrd
337
Option 3: Cap Pnl and Pnd, otherwise no strength change, include hole(s) in Pcr
This method puts bounds in place and assumes local-global interaction happens at full Pne
2
for c 1.5 Pne = 0.658 c Py
0.877
for c > 1.5 Pne = 2 Py = 0.877Pcre

c
where
Py Pcre

Py
Ag
= AgFy
Local Buckling
for l 0.776 Pnl = Pne Pynet
P
for l > 0.776 Pnl = 1 0.15 crl
Pne
where
0 .4
Pcrl
Pne
0 .4
Pne Pynet
Pne Pcrl

Pne is defined in Section above.
Pynet = AnetFy
Anet
= net area of the column
for d 0.561 Pnd = Py Pynet
0.6
0.6
Pcrd Pcrd
Py P.ynet
for d > 0.561 Pnd = 1 0.25
Py Py
where
= Py Pcrd
338
Option 4: Cap Pnl, transition Pnd, include hole(s) in Pcr determinations

This method puts bounds and transition in place, assumes local-global interaction at full Pne
2
for c 1.5 Pne = 0.658 c Py
0.877
for c > 1.5 Pne = 2 Py = 0.877Pcre

c
where
Py Pcre

Py
Ag
= AgFy
Local Buckling
for l 0.776 Pnl = Pne Pynet
P
for l > 0.776 Pnl = 1 0.15 crl
Pne
where
0 .4
Pcrl
Pne
0 .4
Pne Pynet
Pne Pcrl

Pne is defined in Section above.
Pynet = AnetFy
Anet = net area of the column
The nominal axial strength, Pnd, for distortional buckling shall be calculated in accordance with the
following:
(a) For d d1
Pnd = Pynet
(b) For d1<dd2

Pnd = P Pynet Pd 2
ynet

d1
d2
(c) For d > d2
Pnd = 1 0.25 crd

Py
where
0 .6
Pcrd
P
y
0 .6
d1
= 0.561(Pynet Py )
Py
Py Pcrd
d2
= 0.561 14 (Pynet Py )0.4 13
Pd2
= 1 0.25 (1 d 2 )1.2 (1 d 2 )1.2 Py
Pynet = FyAnet0.6Py
Pcrd
= Critical elastic distortional column buckling load including hole(s)
339
Option 5: Transition Pnl (Option A), transition Pnd, include hole(s) in Pcr determinations
2
for c 1.5 Pne = 0.658 c Py
0.877
for c > 1.5 Pne = 2 Py = 0.877Pcre

c
where
Py Pcre

Py
= AgFy
Ag
Local Buckling
(a) For l l1
Pnl = Pne Pynet
(b) For l1<ll2
P
( l l1 )
ynet
l2
Pnl = P
ynet
l 2 l1
P
(c) For l > l2
Pnl = 1 0.15 Pcrl
0.4
Pne
where
l
l1
P
crl
Pne
0. 4
Pne
= Pne Pcrl
= 0.776(Pynet Pne ) 0.776
l2 = 0.776 1.7 (Pynet Pne )1.6 0.7 , Pynet/Pne1
= 0.776, Pynet/Pne>1
= 1 0.15(1 l 2 )0.8 (1 l 2 )0.8 Pne
Pynet = FyAnet0.6Py
Pl2
(notransitionwhenPynet/Pne>1)

SameasOption4
340
Option 6: Transition Pnl (Option B), transition Pnd, include hole(s) in Pcr determinations
2
for c 1.5 Pne = 0.658 c Py
0.877
for c > 1.5 Pne = 2 Py = 0.877Pcre

c
where
Py Pcre

Py
= AgFy
Ag
Local Buckling
(a) For l l1
Pnl = Pynet (Pne /Py )
(b) For l1<ll2
Pynet (Pne Py ) Pl 2
(l l1 )
P
l 2 l1
Pnl = P Pne
ynet
(c) For l > l2
Pnl = 1 0.15 Pcrl
Pne
where
l
l1
l2
0.4
Pl2
P
crl
Pne
0. 4
Pne
= Pne Pcrl
= 0.776(Pynet Py )
= 0.776 1.7 (Pynet Py )1.6 0.7
= 1 0.15(1 l 2 )
Pynet = FyAnet0.6Py
0.8
)(1
l2
0.8
Pne

SameasOption4
341
8.1.6 DSM comparison to column test simulation database
The six DSM prediction options for coldformed steel columns with holes are
evaluated with the simulated column experiment database developed in Section 8.1.1
and summarized in Appendix K. (Tested strengths with and without global
imperfectionsareprovidedinAppendixK.Thesimulatedstrengthsconsideredinthis
study contain global imperfections, except for stocky columns with L/D<18whereD is
the column flange width). The simulated data is compared against DSM predictions
whileevaluatingdatatrendsagainstmemberslenderness,holesize(Anet/Ag),andcolumn
dimensionsL/h,wherehistheflatwebwidthofacolumn.
Figure 8.18 to Figure 8.21 compare the simulated test data to predictions for local,
distortional,andglobalbucklingcontrolledcolumnfailures.Option1isidenticaltothe
existing DSM approach for columns without holes, except the critical elastic buckling
loads(Pcrl,Pcrd,andPcre)aredeterminedwiththeinfluenceofholes.Option1isobserved
to be an accurate predictor of strength when l, d, and c are high, but results in
unconservative predictions (by as much as 30 % for distortional buckling controlled
specimens, see Figure 8.20) as l, d, and c decreases below 1.5. The unconservative
predictionsoccurbecauseOption1doesnotaccountforthecolumnstrengthlimitPynet,
nordoesitaccountforatransitionfromanelasticbucklingcontrolledfailuretoayield
controlledfailureatthenetsectiondiscussedinSection8.1.2andSection8.1.3.
Option2isobservedtobeaconservativepredictorinFigure8.18toFigure8.21for
highl,d,andcanddemonstratesimprovedaccuracyoverOption1whenslenderness
decreases and hole size increases (see Figure 8.20). Option 2 replaces Pynet everywhere
342
withintheexistingDSMformulation,whichhastheeffectofincreasingl,d,andcand
decreasingpredictedstrength.Option3testtopredictedtrendsaresimilartoOption1
with increasingly unconservative predictions as slenderness decreases, demonstrating
thatthePynetlimitonPnlandPndinOption3arenotfullyeffectiveatcapturingtheyield
transitiontothenetsection.Option4isidenticaltoOption3excepttheyieldtransition
ontheDSMdistortionalcurvedevelopedinSection8.1.2isemployedtoprovideamore
accurate prediction of the netsection yielding influence. Option 4 demonstrates an
improvementinaccuracyoverOption3,althoughitoverpredictsthestrengthofthetwo
columnsdiscussedinSection8.1.4(SSMA350S16268andSSMA350S16254columns),
where large holes caused a sudden weakaxis flexural buckling failure. Option 5
includes both local and distortional yield transitions, although the predictions are
identicaltoOption4becausethedistortionaltransitionalwayspredictslowerstrengths
thanthelocaltransitionforthecolumnsconsidered.Option6deviatesfromtheother
approaches and accounts for the presence of holes by reducing Pnl by the ratio Pynet/Py
when lislessthan0.776;thisoptionalsoalwaysincludingalocalbucklingtransition
(Option 5 imposes a transition on the DSM local buckling design curve only when
Pynet<Pne, see Figure 8.17). The reduction in Pnl shifts the global bucklingcontrolled
specimensinOptions1through5totheDSMlocalbucklingcurveinOption6,resulting
inconservativepredictionswithdecreasingl.
Table 8.1 summarizes the testtopredicted ratio statistics for the six DSM options.
Thestandarddeviation(SD)isusefulwhencomparingthemethods,becauseitprovides
a metric for how well the trends in strength are following the prediction curves. (The
343
mean is also an important statistic but can hide unconservative prediction trends in
some columns with overconservative predictions in other columns). A low standard
deviation is appealing because it enables higher strength reduction factors in a design
code.Thestrengthreductionfactor isalsoprovidedforeachoption. iscalculated
withthefollowingequationfromChapterFoftheSpecification(AISIS1002007):
= C (M m Fm Pm )e
o VM2 +VF2 + C PVP2 +VQ2
(8.1)
wherethecalibrationcoefficientC =1.52forLRFD,themeanvalueofthematerialfactor
Mm=1.10 for concentrically loaded compression members, the mean value of the
fabricationfactorFm=1.0,themeanvalueoftheprofessionalfactorPm=1.0,thecoefficient
of variation (COV) of the material factor Vm=0.10, the COV of the fabrication factor
Vf=0.05,theCOVoftheloadeffectVq=0.21forLRFD,andthecorrectionfactorCp=1.The
COV of the test results, Vp, is calculated as the ratio of the standard deviation to the
meanofthetesttopredictedstatisticsinTable8.1.
Nooneoptionstandsoutabovetherestwhenstudyingthetable,althoughOption2,
3, and 4 (5) have the most evenly distributed statistics between local and distortional
buckingcolumngroups.Theobservationsfromthiscomparisonwillbecombinedwith
theDSMcomparisontotheexperimentaldatabaseinthenextsection.
Table8.1DSMtesttopredictedstatisticsforcolumnsimulations
Option
1
Description
Py everywhere
Mean
1.06
Local buckling
SD
0.15
0.83
# of tests
93
Mean
1.07
SD
# of tests
0.17
0.82
178
Mean
1.11
Global buckling
SD
0.21
0.78
# of tests
114
Pynet everywhere
1.14
0.13
0.86
93
1.24
0.18
0.83
176
1.15
0.18
0.82
116
Cap Pnl, Pnd
1.06
0.15
0.83
93
1.09
0.17
0.82
186
1.13
0.21
0.79
106
Transition Pnd, Cap Pnl
1.08
0.14
0.85
89
1.04
0.19
0.79
200
1.16
0.19
0.82
96
Transition Pnd and Pnl (Option A)
1.08
0.14
0.85
89
1.04
0.19
0.79
200
1.16
0.19
0.82
96
Transition Pnd and Pnl (Option B)
1.07
0.20
0.78
221
1.10
0.15
0.85
164
---
---
---
344
1.5
1.5
Option 2 - Pynet everywhere
Pn/Pne
Pn/Pne
Option 1 - Py everywhere
0.5
0.5
1.5
2.5
0.5
0.5
1.5
l=(Pne/Pcrl)0.5
1.5
Pn/Pne
Pn/Pne
Option 4 - cap Pn , transition Pnd

l
0.5
0.5
1.5
2.5
DSM Local Prediction

Local Controlled Test
0.5
0.5
1.5
2.5
2.5
l=(Pne/Pcrl)0.5
l=(Pne/Pcrl)
1.5
1.5
Option 5 - transition Pn (Option A), transition Pnd
Option 6 - transition Pn (Option B), transition Pnd
Pn/Pne
Pn/Pne
0.5
0.5
2.5
1.5
Option 3 - cap Pn , Pnd
l
l=(Pne/Pcrl)0.5
0.5
1.5
2.5
0.5
0.5
0.5
1.5
l=(Pne/Pcrl)0.5
l=(Pne/Pcrl)
Figure8.18Comparisonofsimulatedteststrengthstopredictionsforcolumnswithlocalbucklingcontrolledfailuresasafunctionoflocalslenderness(tested
strengthisnormalizedbyPne)
345
1.5
1.5
Pn/Py
Pn/Py
0.5
0.5
0.5
1.5
2.5
0.5
1.5
l=(Pne/Pcrl)0.5
2.5
l=(Pne/Pcrl)0.5
1.5
1.5
l

l
Pn/Py
Pn/Py

0.5
0.5
0.5
1.5
2.5
0.5
1.5
0.5
2.5
2.5
l=(Pne/Pcrl)0.5
l=(Pne/Pcrl)
1.5
1.5
Pn/Py
Pn/Py
0.5
0.5
0.5
1.5
2.5
0.5
0.5
1.5
l=(Pne/Pcrl)0.5
l=(Pne/Pcrl)
Figure8.19Comparisonofsimulatedteststrengthstopredictionsforcolumnswithlocalbucklingcontrolledfailuresasafunctionoflocalslenderness(tested
strengthisnormalizedbyPyg)
346
1.5
1.5
Pn/Py
Pn/Py
DSM Dist. Prediction

Dist. Controlled Test
0.5
0.5
0.5
1.5
2.5
0.5
d=(Py /Pcrd)0.5
1.5
2.5
2.5
2.5
dnet=(Pynet/Pcrd)0.5
1.5
1.5
l

l
Pn/Py
Pn/Py
0.5
0.5
0.5
1.5
2.5
0.5
0.5
1.5
d=(Py /Pcrd)0.5
d=(Py /Pcrd)
1.5
1.5
Pn/Py
Pn/Py
0.5
0.5
0.5
1.5
2.5
0.5
0.5
1.5
d=(Py /Pcrd)0.5
d=(Py /Pcrd)
Figure8.20Comparisonofsimulatedteststrengthstopredictionsforcolumnswithdistortionalbucklingcontrolledfailuresasafunctionofdistortional
slenderness
347
1.5
1.5
Pn/Py
Pn/Py
DSM Global Prediction

Global Controlled Test
0.5
0.5
0.5
1.5
2.5
0.5
c =(Py /Pcre)0.5
1.5
2.5
2.5
2.5
cnet=(Pynet/Pcre)0.5
1.5
1.5
l

l
Pn/Py
Pn/Py
0.5
0.5
0.5
1.5
2.5
0.5
0.5
1.5
c =(Py /Pcre)0.5
c =(Py /Pcre)
1.5
1.5
Pn/Py
Pn/Py
0.5
0.5
0.5
1.5
2.5
0.5
0.5
1.5
c =(Py /Pcre)0.5
c =(Py /Pcre)
Figure8.21Comparisonofsimulatedteststrengthstopredictionsforcolumnswithglobalbucklingcontrolledfailures(i.e.,nolocalinteraction)asafunctionof
globalslenderness
348
1.5
Ptest/Pn
Ptest/Pn
1.5
0.5
Local Controlled
0.5
0
0.5

1.5
2.5
0.5
1.5
1.5
0.5
0.5
1.5
2.5
0.5
1.5
Ptest/Pn
Ptest/Pn
1.5
0.5
1.5
1.5
2.5
2.5
0.5

l
0.5
l=(Pne/Pcrl)0.5
l=(Pne/Pcrl)
2.5

l
0.5
0.5

l
0
1.5
l=(Pne/Pcrl)0.5
Ptest/Pn
Ptest/Pn
l=(Pne/Pcrl)0.5

l
2.5
0.5
0.5
1.5
l=(Pne/Pcrl)0.5
l=(Pne/Pcrl)
Figure8.22Testtopredictedratiosforlocalbucklingcontrolledsimulatedcolumnfailuresasafunctionoflocalslenderness
349
1.5
Ptest/Pn
Ptest/Pn
1.5
0.5
Distortional Controlled
0.5
0
0.5

1.5
2.5
0.5
1.5
1.5
0.5
0.5
1.5
2.5
0.5
1.5
Ptest/Pn
Ptest/Pn
1.5
0.5
1.5
1.5
2.5
2.5
0.5

l
0.5
d=(Py /Pcrd)0.5
d=(Py /Pcrd)
2.5

l
0.5
0.5

l
0
1.5
Ptest/Pn
Ptest/Pn
d=(Py /Pcrd)0.5

l
2.5
0.5
0.5
1.5
d=(Py /Pcrd)0.5
d=(Py /Pcrd)
Figure8.23Testtopredictedratiosfordistortionalbucklingcontrolledsimulatedcolumnfailuresasafunctionofdistortionalslenderness
350
1.5
Ptest/Pne
Ptest/Pne
1.5
0.5
0.5
0
0.5
1.5
2.5
0.5
1.5
1.5
0.5
0.5
1.5
2.5
0.5
1.5
Ptest/Pne
Ptest/Pne
1.5
0.5
1.5
1.5
2.5
2.5
0.5
l

l
0.5
c =(Py /Pcre)0.5
c =(Py /Pcre)
2.5

l
0.5
0.5

l
0
1.5
Ptest/Pne
Ptest/Pne
c =(Py /Pcre)0.5
Global Controlled
2.5
0.5
0.5
1.5
c =(Py /Pcre)0.5
c =(Py /Pcre)
Figure8.24Testtopredictedratiosforsimulatedglobalbucklingcontrolledcolumnfailures(i.e.,nolocalbucklinginteraction)asafunctionofglobalslenderness
351
1.5
Ptest/Pn
Ptest/Pn
1.5
0.5
0.5
0
0.5
0.6
0.7
0.8
0.9
0
0.5
0.6
0.7
1.5
1.5
0.5
0.7
0.8
0.9
0
0.5
0.6
0.7
1.5
Ptest/Pn
Ptest/Pn
0.8
0.9
0.9
Anet/Ag
1.5
0.5
0.5

l
0.6

l
Anet/Ag
0
0.5
0.9
0.5

l
0.6
0.8
Anet/Ag
Ptest/Pn
Ptest/Pn
Anet/Ag
0
0.5
Local Controlled
0.7
0.8

l
0.9
0
0.5
Anet/Ag
0.6
0.7
0.8
Anet/Ag
Figure8.25Testtopredictedratiosforsimulatedlocalbucklingcontrolledcolumnfailuresasafunctionofnetcrosssectionalareatogrosscrosssectionalarea
352
1.5
Ptest/Pn
Ptest/Pn
1.5
0.5
0.5
0
0.5
0.6
0.7
0.8
0.9
0
0.5
0.6
0.7
1.5
1.5
0.5
0.7
0.8
0.9
0
0.5
0.6
0.7
1.5
Ptest/Pn
Ptest/Pn
0.8
0.9
0.9
Anet/Ag
1.5
0.5
0.5

l
0.6

l
Anet/Ag
0
0.5
0.9
0.5

l
0.6
0.8
Anet/Ag
Ptest/Pn
Ptest/Pn
Anet/Ag
0
0.5
0.7
0.8

l
0.9
0
0.5
Anet/Ag
0.6
0.7
0.8
Anet/Ag
Figure8.26Testtopredictedratiosforsimulateddistortionalbucklingcontrolledcolumnfailuresasafunctionofnetcrosssectionalareatogrosscrosssectional
area
353
1.5
Ptest/Pn
Ptest/Pn
1.5
0.5
0.5
0
0.5
0.6
0.7
0.8
0.9
0
0.5
0.6
0.7
1.5
1.5
0.5
0.7
0.8
0.9
0
0.5
0.6
0.7
1.5
Ptest/Pn
Ptest/Pn
0.8
0.9
0.9
Anet/Ag
1.5
0.5
0.5

l
0.6

l
Anet/Ag
0
0.5
0.9
0.5

l
0.6
0.8
Anet/Ag
Ptest/Pn
Ptest/Pn
Anet/Ag
0
0.5
Global Controlled
0.7
0.8

l
0.9
0
0.5
Anet/Ag
0.6
0.7
0.8
Anet/Ag
Figure8.27Testtopredictedratiosforsimulatedglobalbucklingcontrolledcolumnfailuresasafunctionofnetcrosssectionalareatogrosscrosssectionalarea
354
1.5
Ptest/Pn
Ptest/Pn
1.5
0.5
Local Controlled
0.5
0
10
15
20
L/h
25
30
35
40
1.5
1.5
Ptest/Pn
Ptest/Pn
0.5
10
15
20
L/h
25
30
35
40
1.5
1.5
0.5
10
20
L/h
25
30
35
40
15
20
L/h
25
10
15
20
L/h
25
30
35
40
30
35
40
0.5

l
0
15

l
Ptest/Pn
Ptest/Pn
10
0.5

l
0

l
30
35
40
10
15
20
L/h
25
Figure8.28Testtopredictedratiosforsimulatedlocalbucklingcontrolledcolumnfailuresasafunctionofcolumnlength,L,toflatwebwidth,h
355
1.5
Ptest/Pn
Ptest/Pn
1.5
0.5
0.5
0
10
15
20
L/h
25
30
35
40
1.5
1.5
Ptest/Pn
Ptest/Pn
0.5
10
15
20
L/h
25
30
35
40
1.5
1.5
0.5
10
20
L/h
25
30
35
40
15
20
L/h
25
10
15
20
L/h
25
30
35
40
30
35
40
0.5

l
0
15

l
Ptest/Pn
Ptest/Pn
10
0.5

l
0

l
30
35
40
10
15
20
L/h
25
Figure8.29Testtopredictedratiosforsimulateddistortionalbucklingcontrolledcolumnfailuresasafunctionofcolumnlength,L,toflatwebwidth,h
356
1.5
Ptest/Pn
Ptest/Pn
1.5
0.5
Global Controlled
0.5
0
10
15
20
L/h
25
30
35
40
1.5
1.5
Ptest/Pn
Ptest/Pn
0.5
10
15
20
L/h
25
30
35
40
1.5
1.5
0.5
10
20
L/h
25
30
35
40
15
20
L/h
25
10
15
20
L/h
25
30
35
40
30
35
40
0.5

l
0
15

l
Ptest/Pn
Ptest/Pn
10
0.5

l
0

l
30
35
40
10
15
20
L/h
25
Figure8.30Testtopredictedratiosforsimulatedglobalbucklingcontrolledcolumnfailuresasafunctionofcolumnlength,L,towebwidth,h
357
8.1.7 DSM comparison to experimental column database
The six DSM options are now compared to the column experiment database first
assembled in Section 4.2.6.2. The database contains the elastic buckling properties of
eachcolumn,includingthepresenceofholesandtheinfluenceofboundaryconditions,
aswellasthetestedstrengths.Figure8.31throughFigure8.34comparetheexperiment
strengths to DSM predictions for local, distortional, and global buckling controlled
column failures. Option 1 is observed to be an accurate predictor of column strength
when local, distortional, and global slenderness are high, but overpredicts the tested
strength as slenderness decreases. This trend is consistent with the simulated
experimentcomparisoninSection8.1.6andemphasizestheneedforalimitoncolumn
strength when yielding at the net section controls the failure of a column with holes.
Option 2 is even more conservative in this study when compared to the simulated
column study because the tested specimens considered only have one hole, and
therefore employing Pynet produces unrealistically high column slenderness. Option 3
shifts column specimens from the global buckling failure group to the local buckling
failuregroupwiththePynetlimitonPnl,resultinginimprovedaccuracywhencompared
toOption2.FourcolumnsintheOption3distortionalbucklingfailuregrouparestill
overpredicted by more than 10% though as observed in Figure 8.33. Option 4 and
Option 5 improve the accuracy of the underpredicted specimen strengths in Option 3
withtheadditionofthedistortionalandlocalyieldcontroltransitionstothenetsection.
Option6isanoverlyconservativepredictorofcolumnsfailingbyglobalbuckling.
358
Table 8.2 summarizes the testtopredicted ratio statistics for all columns in the
database. Options 3, 4, and 5 are identified as the methods with the mean closest to
unityandwiththeloweststandarddeviations.Thestatisticsforjustthestubcolumns
(c<0.20) in Table 8.3 confirm the viability of DSM Options 3, 4, and 5, and provides
more direct evidence that holes limit the column strength to the net section Pynet; the
mean testtopredicted ratio is 0.84 for global (yielding) failures of stub columns
employingOption1.
Table8.2DSMtesttopredictedratiostatisticsforcolumnexperiments
Option
1
Description
Py everywhere
Mean
1.03
Local buckling
SD
0.11
0.87
# of tests
52
Mean
1.09
SD
# of tests
0.16
0.83
15
Mean
1.06
Global buckling
SD
0.17
0.82
# of tests
11
Pynet everywhere
1.17
0.09
0.89
47
1.22
0.13
0.87
15
1.17
0.15
0.85
16
Cap Pnl, Pnd
1.07
0.08
0.90
42
1.06
0.13
0.85
29
1.16
0.09
0.90
1.07
0.08
0.90
40
1.10
0.11
0.87
33
1.19
0.08
0.90
1.06
0.08
0.89
47
1.13
0.10
0.89
26
1.19
0.08
0.90
1.12
0.15
0.84
56
1.14
0.10
0.89
22
---
---
---
Table8.3DSMtesttopredictedratiostatisticsforcolumnexperiments(stubcolumnsonly)
Option
1
Description
Py everywhere
Mean
0.98
Local buckling
SD
0.10
0.88
# of tests
33
Mean
0.83
SD
# of tests
0.01
0.92
3
Mean
0.84
Global buckling
SD
0.08
0.88
# of tests
3
Pynet everywhere
1.12
0.07
0.90
28
1.03
0.06
0.91
1.07
0.12
0.86
Cap Pnl, Pnd
1.03
0.06
0.91
23
1.00
0.12
0.86
16
---
---
---
1.04
0.06
0.91
21
1.06
0.11
0.87
18
---
---
---
1.03
0.07
0.90
28
1.11
0.10
0.88
11
---
---
---
1.03
0.07
0.90
29
1.11
0.10
0.88
10
---
---
---
359
1.5
1.5
Pn/Pne
Pn/Pne
0.5
0.5
1.5
2.5
0.5
0.5
1.5
0.5
1.5

l
0.5

0.5
1.5
2.5
0.5
0.5
1.5
l=(Pne/Pcrl)0.5
2.5
2.5
l=(Pne/Pcrl)0.5
1.5
1.5
1
Pn/Pne
1
Pn/Pne
1
Pn/Pne
Pn/Pne
0.5
2.5
1.5
l
l=(Pne/Pcrl)0.5
l=(Pne/Pcrl)
0.5
1.5
2.5
0.5
0.5
0.5
1.5
l=(Pne/Pcrl)0.5
l=(Pne/Pcrl)
Figure8.31Comparisonofexperimentalteststrengthstopredictionsforcolumnswithlocalbucklingcontrolledfailuresasafunctionoflocalslenderness(tested
strengthisnormalizedbyPne)
360
1.5
1.5
Pn/Py
Pn/Py
0.5
0.5
0.5
1.5
2.5
0.5
1.5
l=(Pne/Pcrl)0.5
2.5
l=(Pne/Pcrl)0.5
1.5
1.5
l

l
Pn/Py
Pn/Py

0.5
0.5
0.5
1.5
2.5
0.5
1.5
0.5
2.5
2.5
l=(Pne/Pcrl)0.5
l=(Pne/Pcrl)
1.5
1.5
Pn/Py
Pn/Py
0.5
0.5
0.5
1.5
2.5
0.5
0.5
1.5
l=(Pne/Pcrl)0.5
l=(Pne/Pcrl)
Figure8.32Comparisonofexperimentalteststrengthstopredictionsforcolumnswithlocalbucklingcontrolledfailuresasafunctionoflocalslenderness(tested
strengthisnormalizedbyPy)
361
1.5
1.5
Pn/Py
Pn/Py

0.5
0.5
0.5
1.5
2.5
0.5
d=(Py /Pcrd)0.5
1.5
2.5
2.5
2.5
1.5
1.5
l

l
Pn/Py
Pn/Py
0.5
0.5
0.5
1.5
2.5
0.5
0.5
1.5
d=(Py /Pcrd)0.5
d=(Py /Pcrd)
1.5
1.5
Pn/Py
Pn/Py
0.5
0.5
0.5
1.5
2.5
0.5
0.5
1.5
d=(Py /Pcrd)0.5
d=(Py /Pcrd)
Figure8.33Comparisonofexperimentalteststrengthstopredictionsforcolumnswithdistortionalbucklingcontrolledfailuresasafunctionofdistortional
slenderness
362
1.5
1.5
Pn/Py
Pn/Py
DSM Global Prediction

Global Controlled Test
0.5
0.5
0.5
1.5
2.5
0.5
c =(Py /Pcre)0.5
1.5
2.5
2.5
2.5
1.5
1.5
l

l
Pn/Py
Pn/Py
0.5
0.5
0.5
1.5
2.5
0.5
0.5
1.5
c =(Py/Pcre)0.5
c =(Py /Pcre)
1.5
1.5
Pn/Py
Pn/Py
0.5
0.5
0.5
1.5
2.5
0.5
0.5
1.5
c =(Py/Pcre)0.5
c =(Py /Pcre)
Figure8.34Comparisonofexperimentalteststrengthstopredictionsforcolumnswithglobalbucklingcontrolledfailuresasafunctionofglobalslenderness
363
1.5
Ptest/Pn
Ptest/Pn
1.5
0.5
Local Controlled
0.5
0
0.5

1.5
2.5
0.5
1.5
1.5
0.5
0.5
1.5
2.5
0.5
1.5
Ptest/Pn
Ptest/Pn
1.5
0.5
1.5
1.5
2.5
2.5
0.5

l
0.5
l=(Pne/Pcrl)0.5
l=(Pne/Pcrl)
2.5

l
0.5
0.5

l
0
1.5
l=(Pne/Pcrl)0.5
Ptest/Pn
Ptest/Pn
l=(Pne/Pcrl)0.5

l
2.5
0.5
0.5
1.5
l=(Pne/Pcrl)0.5
l=(Pne/Pcrl)
Figure8.35Testtopredictedratiosforexperimentlocalbucklingcontrolledcolumnfailuresasafunctionoflocalslenderness
364
1.5
Ptest/Pn
Ptest/Pn
1.5
0.5
0.5
0
0.5

1.5
2.5
0.5
1.5
1.5
0.5
0.5
1.5
2.5
0.5
1.5
Ptest/Pn
Ptest/Pn
1.5
0.5
1.5
1.5
2.5
2.5
0.5

l
0.5
d=(Py /Pcrd)0.5
d=(Py /Pcrd)
2.5

l
0.5
0.5

l
0
1.5
Ptest/Pn
Ptest/Pn
d=(Py /Pcrd)0.5

l
2.5
0.5
0.5
1.5
d=(Py /Pcrd)0.5
d=(Py /Pcrd)
Figure8.36Testtopredictedratiosforexperimentdistortionalbucklingcontrolledcolumnfailuresasafunctionofdistortionalslenderness
365
1.5
Ptest/Pne
Ptest/Pne
1.5
0.5
Global Controlled
0.5
0
0.5

1.5
2.5
0.5
1.5
1.5
0.5
0.5
1.5
2.5
0.5
1.5
Ptest/Pne
Ptest/Pne
1.5
0.5
1.5
1.5
2.5
2.5
0.5

l
0.5
c =(Py/Pcre)0.5
c =(Py /Pcre)
2.5

l
0.5
0.5

l
0
1.5
Ptest/Pne
Ptest/Pne
c =(Py /Pcre)0.5

l
2.5
0.5
0.5
1.5
c =(Py/Pcre)0.5
c =(Py /Pcre)
Figure8.37Testtopredictedratiosforexperimentglobalbucklingcontrolledcolumnfailuresasafunctionofglobalslenderness
366
1.5
Ptest/Pn
Ptest/Pn
1.5
0.5
Local Controlled
0.5
0
0.5
0.6
0.7
0.8
0.9
0
0.5
0.6
0.7
1.5
1.5
0.5
0.6
0.7
0.8
0.9
0
0.5
0.6
0.7
1.5
Ptest/Pn
Ptest/Pn
0.8
0.9
0.9
Anet/Ag
1.5
0.5
0.5

l
0.6

l
Anet/Ag
0
0.5
0.9
0.5

l
0
0.5
0.8
Anet/Ag
Ptest/Pn
Ptest/Pn
Anet/Ag
0.7
0.8

l
0.9
0
0.5
Anet/Ag
0.6
0.7
0.8
Anet/Ag
Figure8.38Testtopredictedratiosforexperimentlocalbucklingcontrolledcolumnfailuresasafunctionofnetcrosssectionalareatogrosscrosssectionalarea
367
1.5
Ptest/Pn
Ptest/Pn
1.5
0.5
0.5
0
0.5
0.6
0.7
0.8
0.9
0
0.5
0.6
0.7
1.5
1.5
0.5
0.6
0.7
0.8
0.9
0
0.5
0.6
0.7
1.5
Ptest/Pn
Ptest/Pn
0.8
0.9
0.9
Anet/Ag
1.5
0.5
0.5

l
0.6

l
Anet/Ag
0
0.5
0.9
0.5

l
0
0.5
0.8
Anet/Ag
Ptest/Pn
Ptest/Pn
Anet/Ag
0.7
0.8

l
0.9
0
0.5
Anet/Ag
0.6
0.7
0.8
Anet/Ag
Figure8.39Testtopredictedratiosforexperimentdistortionalbucklingcontrolledcolumnfailuresasafunctionofnetcrosssectionalareatogrosscrosssectional
area
368
1.5
Ptest/Pn
Ptest/Pn
1.5
0.5
0.5
0
0.5
0.6
0.7
0.8
0.9
0
0.5
0.6
0.7
1.5
1.5
0.5
0.7
0.8
0.9
0
0.5
0.6
0.7
1.5
Ptest/Pn
Ptest/Pn
0.8
0.9
0.9
Anet/Ag
1.5
0.5
0.5

l
0.6

l
Anet/Ag
0
0.5
0.9
0.5

l
0.6
0.8
Anet/Ag
Ptest/Pn
Ptest/Pn
Anet/Ag
0
0.5
Global Controlled
0.7
0.8

l
0.9
0
0.5
Anet/Ag
0.6
0.7
0.8
Anet/Ag
Figure8.40Testtopredictedratiosforexperimentglobalbucklingcontrolledcolumnfailuresasafunctionofnetcrosssectionalareatogrosscrosssectionalarea
369
1.5
Ptest/Pn
Ptest/Pn
1.5
0.5
Local Controlled
0.5
0
10
15
20
L/h
25
30
35
40
1.5
1.5
Ptest/Pn
Ptest/Pn
0.5
10
15
20
L/h
25
30
35
40
1.5
1.5
0.5
10
20
L/h
25
30
35
40
15
20
L/h
25
10
15
20
L/h
25
30
35
40
30
35
40
0.5

l
0
15

l
Ptest/Pn
Ptest/Pn
10
0.5

l
0

l
30
35
40
10
15
20
L/h
25
Figure8.41Testtopredictedratiosforexperimentlocalbucklingcontrolledcolumnfailuresasafunctionofcolumnlength,L,toflatwebwidth,h
370
1.5
Ptest/Pn
Ptest/Pn
1.5
0.5
0.5
0
10
15
20
L/h
25
30
35
40
1.5
1.5
Ptest/Pn
Ptest/Pn
0.5
10
15
20
L/h
25
30
35
40
1.5
1.5
0.5
10
20
L/h
25
30
35
40
15
20
L/h
25
10
15
20
L/h
25
30
35
40
30
35
40
0.5

l
0
15

l
Ptest/Pn
Ptest/Pn
10
0.5

l
0

l
30
35
40
10
15
20
L/h
25
Figure8.42Testtopredictedratiosforexperimentdistortionalbucklingcontrolledcolumnfailuresasafunctionofcolumnlength,L,toflatwebwidth,h
371
1.5
Ptest/Pn
Ptest/Pn
1.5
0.5
Global Controlled
0.5
0
10
15
20
L/h
25
30
35
40
1.5
1.5
Ptest/Pn
Ptest/Pn
0.5
10
15
20
L/h
25
30
35
40
1.5
1.5
0.5
10
20
L/h
25
30
35
40
15
20
L/h
25
10
15
20
L/h
25
30
35
40
30
35
40
0.5

l
0
15

l
Ptest/Pn
Ptest/Pn
10
0.5

l
0

l
30
35
40
10
15
20
L/h
25
Figure8.43Testtopredictedratiosforexperimentglobalbucklingcontrolledcolumnfailuresasafunctionofcolumnlength,L,towebwidth,h
372
8.1.8 Recommendations DSM for columns with holes
Options3,4,and5arepresentedasviableproposalsforextendingDSMtocolumns
with holes. This recommendation is based on the testtopredicted statistics and data
trends presented in Section 8.1.6 and Section 8.1.7, and also considers the effort to
implement the modifications and their ease of use by design engineers. Option 3
accountsforthereductionincolumnstrengthfromthepresenceofholesbycappingPnl
and Pnd at Pynet. This is a simple modification to implement in the Specification and
avoids additional calculation work for a design engineer (except for that required to
calculatethecriticalelasticbucklingloadsincludingtheinfluenceoftheholes).Options
4and5arerefinementsofOption3,wherethecaponPnlandPndbecomesatransition
from an elastic buckling controlled failure mode to a yield controlled failure at Pynet.
These two methods require additional effort from the designer when compared to
Option3,buttheyhaveanimportantadvantage.Options4and5aremorecloselytied
to the failure mechanisms influencing column strength because they capture the yield
transitiontothenetsectionintheirpredictions.Thetransitionsincreasetheprobability
that strength will be accurately predicted for general column and hole geometries.
Option5hastheadditionaladvantageofcapturingtheinfluenceofayieldtransitionfor
closed crosssections that do not experience distortional buckling. This generality is
whatmotivatestheuseoftheDirectStrengthMethod(AISIS1002007,Appendix1).
373
8.2 DSMforlaterallybracedbeamswithholes
8.2.1 Database of simulated column experiments
Simulatedexperimentswereconductedon125Csectionlaterallybracedbeamswith
evenlyspacedcircularwebholesinABAQUS.Crosssectionswerespecificallyselected
withcustomMatlabcodeemployingtheexistingDSMdesigncurvestoidentifybeams
predisposed to local and distortional bucklingcontrolled failures. The crosssections
were chosen from a catalog of 99 industry standard Csections published by the Steel
Stud Manufacturers Association (SSMA 2001). The nominal outtoout dimensions
provided in the SSMA catalog were converted to centerline dimensions and then
constructedinABAQUSwiththemeshingproceduredescribedinSection7.2.1.1.The
beams in the database have a constant length L=48 in. to accommodate multiple local
anddistortionalbucklinghalfwavesalongthebeam.Evenlyspacedcircularwebholes
wereplacedinthecolumnswithholespacingS(definedinFigure3.2)of16inches(i.e.,
threeevenlyspacedholes).Theholeswerecenteredtransverselyinthewebandtheir
depth(diameter),hhole,wasvariedsuchthattheratioofthenetmomentofinertia,Inet,to
thegrosscrosssectionalarea,Ig,rangedbetween0.85and1.0.
TheABAQUS boundaryconditionsandapplicationofloading,describedinFigure
8.44, are implemented to be consistent with CUFSM, i.e. pinnedpinned and freeto
warpwithauniformstressappliedatthememberends.Eachbeamislaterallybraced
by restraining the compression flange at the midlength of the beam. (Initial modeling
trials, where all nodes centered in the compression flange were laterally restrained,
374
resulted in simulated strengths 25% higher than DSM predictions for beams without
holes.)Consistentnodalloadswereappliedtosimulatethelinearstressgradientatthe
beam ends (see Section 7.2.1.2 for information on S9R5 consistent nodal loads). The
loads (a reference moment of 1 kipin. was applied at each end in ABAQUS) were
distributedoverthefirsttwosetsofcrosssectionnodestoavoidlocalizedfailuresatthe
loadededges.
Node centered in compression flange at
longitudinal midline restrained in 1 (to prevent
rigid body motion) and 3 (for laterally bracing)

2
5
4
1
Moment applied as consistent nodal loads

over two sets of cross-section nodes to
avoid edge failures (Typ.)
Figure8.44ABAQUSsimulatedbeamexperimentsboundaryconditionsandapplicationofloading
The ABAQUS simulations were performed with the modified Riks nonlinear
solution algorithm. Automatic time stepping was enabled with a suggested initial arc
length step of 1 (the Riks method increments in units of energy, in this case kipin.), a
maximum step size of 3, and the maximum number of solution increments set at 300.
Metal plasticity was simulated with the material modeling procedure described in
Section7.2.1.4.Theplastictruestressstraincurveforspecimen362148HinAppendix
H was assumed for all column models (but modified such that plasticity starts at the
yieldstress,seeSection7.2.1.4),wherethesteelyieldstressFy=58.6ksi.Residualstresses
375
and initial plastic strains, as discussed in Section 7.2.1.6, were not considered in the
ABAQUS models because their implementation requires further validation and they
werenotobservedtomarkedlyinfluencecolumnultimatestrength(seeFigure7.48and
Figure7.49).
ImperfectionswereimposedontheinitialbeamgeometryinABAQUSwithcustom
Matlab code which combines the local and distortional buckling crosssection mode
shapes from CUFSM along the column length. Two simulations were performed for
each beam, one model with 25% CDF local and distortional imperfection magnitudes
andtheothermodelwith75%CDFlocalanddistortionalimperfectionmagnitudes(see
Section7.2.1.5forlocalanddistortionalimperfectiondefinitions).
The local (Mcrl) and distortional (Mcrd) critical elastic buckling loads were predicted
for each beam with custom Matlab code based on the CUFSM prediction methods
describedinSection4.3.Thedatabaseofsimulatedbeamexperiments,includingcross
section type, column and hole geometry, simulated ultimate strength (Mtest25 and Mtest75)
and critical elastic buckling loads for each beam (including the presence of holes) is
providedinAppendixL.
8.2.2 Local buckling study
TwelvebeamsfromthesimulationdatabaseinAppendix Lwerechosentostudy
the influence of web holes on the ultimate strength of laterally braced beams
predisposedtoalocalbucklingcontrolledfailure.ThebeamshaveSSMAcrosssections
whichresultinalocalbucklingslenderness,l,rangingfrom1.3to2.0.(Theslenderness
range considered here is relatively narrow because only 12 of the 99 SSMA cross
376
sections, when employed as laterally braced beams, are controlled by a local buckling
failure. The majority of beam crosssections are predicted to exhibit a distortional
bucklingcontrolled failure.) The web of each beam contains three evenly spaced
circularholeswheretheholespacingS=16in.Theholedepth(diameter),hhole,isvaried
foreachbeamtoproduceInet/Igof1.0(noholes),0.95,0.90,and0.85.RefertoAppendix
L,StudyTypeL,forspecificbeamcrosssectionandholegeometryinformation.
The simulation results for Inet/Ig =1.0, 0.95, 0.90, and 0.85, are compared to the DSM
distortionalbucklingpredictioncurveinFigure8.46toFigure8.49.Thebeamstrengths,
Mtest25 and Mtest75, without holes (Inet/Ig =1.0) are consistent with the DSM design curve as
showninFigure8.46a,confirmingthatthenonlinearsimulationprotocoldevelopedfor
columns in Section 7.2 is also viable when conducting coldformed steel beam
simulations.Themeanandstandarddeviationofthesimulatedtesttopredictedratiois
1.05 and 0.05 respectively for 25% CDF local and distortional imperfections, and 1.03
and 0.05 for 75% CDF local and distortional imperfections. For the beams with holes,
thesimulatedteststrengthsdivergefromtheDSMpredictioncurveaslocalslenderness,
l=(Myg/Mcrl)0.5, decreases as shown in Figure 8.47a to Figure 8.49a (Myg is the yield
momentofthecolumncalculatedwiththegrosscrosssectionalareaIg).Thisdivergent
trend in Mtest with decreasing l is consistent with the column results with holes
discussedinSection8.1,whereelasticbucklingcontrolledthefailurewhenslenderness
was high and transitioned to yielding and collapse of the net section as slenderness
decreased.Figure8.45showstheloaddeformationresponseatultimatelimitstatefor
anSSMA800S16243beamconsideredinthisstudy,andhighlightsthetransitionfrom
377
an elastic buckling controlledfailure to a yield controlledfailure at the net section as

holesizeincreases.
Elastic buckling controlled failure
Yielding and collapse of net section
Inet/Ig
1.0
0.95
0.90
Mtest25/Myg
l
0.72
0.68
0.64
0.85
0.61
1.45
1.63
1.45
1.45
Figure8.45SSMA800S16243beamwithwebholesconsideredintheDSMlocalbucklingstudy
Two modification options are proposed for the DSM local buckling beam design
curve:
Local Buckling (Option A)
The nominal flexural strength, Mnl, for local buckling shall be calculated in accordance with the following:
(a) For l l1
Mnl = Mne Mynet
(caponbeamstrength)
(b) For l1<ll2
l 2 l l1
Mnl = M
ynet M ynet M l 2
l l 2 l1
(c) For l > l2
Mnl = 1 0.15 M crl
where
l
l1
M ne
0.4
M
crl
M ne
(nonlinearyieldtransitionwhenMynet/Mne1)
0.4
M ne
= M ne M crl
= 0.776( M ynet M ne ) 0.776
l2 = 0.776 2.4(M ynet M ne )3.5 1.4 , Mynet/Mne1

= 0.776, Mynet/Mne>1
l2 = 1 0.15(1 l 2 )0.8 (1 l 2 )0.8 M ne

Mynet = SfnetFy0.80My
(notransitionwhenMynet/Mne>1)
(limitreductionofthenetsectionto0.8My)
Sfnet = Section modulus at the hole(s) referenced to the extreme fiber at first yield
Mcrl = Critical elastic local beam buckling load including hole(s)
Local Buckling (Option B)

The nominal axial strength, Mnl, for local buckling shall be calculated in accordance with the following:
378
(a) For l l1
Mnl = Mynet (Mne /My)
(b) For l1<ll2
Mnl = M M ne M M ne M l 2 l l1
ynet
ynet
l2
M
y
M
y
(c) For l > l2
Mnl = 1 0.15 M crl
where
l
l1
M ne
0.4
M
crl
M ne
l l 2 l1
0.4
M ne
= M ne M crl
= 0.776( M ynet M y ) 0.776
l2 = 0.776 2.4(M ynet M y )3.5 1.4
(nonlinearyieldtransition)
l2 = 1 0.15(1 l 2 )
0.8
)(1
l2
0.8
M ne
TheframeworkforOptionAandOptionBisbasedontheproposedmodifications
to the DSM local buckling column design curve presented in Section 8.1.4. Option A
imposes a transition from the DSM local buckling curve to the net section limit, Mynet,
whenMynet<Mne.WhenMynet>Mne,OptionAassumesthatholesinfluenceonlythecritical
elastic buckling loads (Mcrl, Mcre) but otherwise do not change the failure mode of the
beam.OptionBalsoimposesatransitiontothenetbeamstrengthfromtheDSMlocal
failurecurve,althoughinthiscasetheyieldtransitionoccursforallvaluesofMynet/Mne.
Theproposedtransitionfromtheelasticbucklingfailureregimetotheyieldplateauis
nonlinearforbothOptionsAandBasdemonstratedinFigure8.47atoFigure8.49a,in
contrast to the linear transition for coldformed steel columns with holes (see Section
8.1.4).
379
Allbeamsconsideredinthisstudyarelaterallybraced,i.e.global(lateraltorsional)
bucklingdoesnotinfluencebeamstrength,andthereforeOptionAandBwillproduce
thesamestrengthpredictions.Thevalidityofbothoptionsforlaterallybracedbeamsis
evaluatedinthefollowingsectionwiththesimulationdatabaseinAppendixLandthe
experiment database assembled in Chapter 4. Future work is planned to evaluate
Option A and B for unbraced coldformed steel beams with holes, where lateral
torsionalbucklinginfluencesbeamstrength.
1.4
1.4
DSM (no holes)
0.8
0.8
0.6
0.6
0.4
0.4
0.2
0.2
0.5
1.5
2.5
local slenderness, =(My /Mcr )0.5

l
l
DSM (no holes)

1.2
Mtest/Mn
Mtest/Mn
1.2
0.5
1.5
2.5

l
l
Figure8.46Comparisonofsimulatedbeamstrengths(Inet/Ig=1.0,noholes)to(a)theexistingDSMlocal
bucklingdesigncurveandto(b)theproposedDSMlocalbucklingcurveforbeamswithholes
1.4
1.4
DSM (no holes)
0.8
0.8
0.6
0.6
0.4
0.4
0.2
0.2
0.5
1.5

l
l
2.5
DSM (no holes)

1.2
Mtest/Mn
Mtest/Mn
1.2
0.5
1.5
2.5

l
l
Figure8.47Comparisonofsimulatedbeamstrengths(Inet/Ig=0.95)to(a)theexistingDSMlocalbuckling
designcurveandto(b)theproposedDSMlocalbucklingcurveforbeamswithholes
380
1.4
1.4
DSM (no holes)
0.8
0.8
0.6
0.6
0.4
0.4
0.2
0.2
0.5
1.5
2.5
DSM (no holes)

1.2
Mtest/Mn
Mtest/Mn
1.2
0.5

l
l
1.5
2.5

l
l
1.4
1.4
DSM (no holes)
0.8
0.8
0.6
0.6
0.4
0.4
0.2
0.2
0.5
1.5
2.5
DSM (no holes)

1.2
Mtest/Mn
Mtest/Mn
1.2

l
l
0.5
1.5

l
l
2.5
8.2.3 Distortional buckling study
A group of 11 beams from the SSMA beam simulation database was chosen to
evaluate the influence of the ratio Inet/Ig on the tested strength of beams predicted to
collapse with a distortional failure mode. (Ig is the gross moment of inertia of a beam
andInetisthemomentofinertiaatthelocationofahole.)ThebeamshaveSSMAcross
sectionswhichresultinadistortionalbucklingslenderness, d,rangingfrom0.6to1.6.
(AllSSMAcrosssections,employedasbeamsandcontrolledbyadistortionalbuckling
failure,liewithinthisslendernessrange.)Inthisstudythebeamdepthsrangefrom4in.
381
to12in.ThewebofeachbeamhasthreecircularholeswheretheholespacingS=16in
(seeFigure3.2forthedefinitionofS).Theholedepth(diameter),hhole,isvariedforeach
beamtoproduceInet/Igof1.0(noholes),0.95,and0.90.RefertoAppendixL,StudyType
D,forspecificcrosssectionandholegeometryinformation.
The simulation results for Inet/Ig =1.0, 0.95, and 0.90 are compared to the DSM
distortionalbucklingpredictioncurveinFigure8.51toFigure8.53.Thebeamstrengths,
Mtest25 and Mtest75, without holes (Inet/Ig =1.0) are consistent with the DSM distortional
buckling design curve as shown in Figure 8.51a, with a trend of increasingly
conservativepredictionsasdistortionalslendernessincreases.Themeanandstandard
deviation of the simulated test to predicted ratio is 1.08 and 0.08 respectively for 25%
CDFlocalanddistortionalimperfections,and1.02and0.12for75%CDFimperfections.
Forthebeamswithholes,thesimulatedteststrengthsdemonstrateaslightdivergence
fromtheDSMpredictioncurveasdistortionalslenderness, d=(Myg/Mcrd)0.5,decreasesas
showninFigure8.52aandFigure8.53a(Mygistheyieldmomentofthebeamcalculated
with the gross crosssectional area Ig). (Figure 8.52a and Figure 8.53a also demonstrate
that Mcrd, predicted with the simplified method in Section 4.3.2.2, increases distortional
slendernessandshiftsthesimulateddataoffofthepredictioncurve.Futureresearchis
plannedtoimprovetheaccuracyofthissimplifiedmethod.)Thisdivergenttrendin
MtestwasalsoobservedinthelocalbucklingcontrolledbeamstudyinSection8.2.2and
the column studies presented in Section 8.1. As d decreases, the beam failure mode
transitions from a distortional buckling failure to yielding and collapse of the net
section.Figure8.50highlightsthistransitionfortheSSMA550S16254beamconsidered
382
in this study by comparing the deformed shape at ultimate limit state as hole size
increases.
Elastic buckling controlled failure
Inet/Ig
Mtest25/Myg
d
Yielding and collapse of net section
1.0
0.95
0.90
0.89
0.90
0.83
0.87
0.92
0.93
Figure8.50SSMA550S16254structuralstudfailuremodetransitionfromdistortionalbucklingtoyielding
atthenetsection
TheobservationsfromthisstudyareusedtoformulateamodifiedDSMdistortional
curve for beams with holes which captures the failure mechanism transition from
yielding at the net crosssection to a distortional type failure mode and limits the
strengthofthebeamtotheyieldmomentatthenetsection:
The nominal flexural strength, Mnd, for distortional buckling shall be calculated in accordance with the
following:
(a) For d d1
Mnd = M ynet
(b) For d1<dd2
ynet
d2
Mnd = M
ynet
d 2 d1
M
(c) For d > d2
M crd
Mnd =
1 0.22 M
y
where
d
d1
d2
d2
0. 5
M
crd
M
y
0 .5
My
M y M crd
= 0.673( M ynet M y )
= 0.673 1.7 (M ynet M y )1.7 0.7

= (1 0.22(1 d 2 ))(1 d 2 )M y
= Section modulus at the hole(s) referenced to the extreme fiber at first yield
Sfnet
Mcrd
= Critical elastic distortional beam buckling load including hole(s)
383
ThemodifiedDSMdistortionalcurveisaddedinFigure8.51btoFigure8.53basInet/Ig
decreases, simulating the transition from the existing DSM curve to the net section
strength limit exhibited by the simulated test data. The linear portion of themodified
prediction curve represents the unstiffened strip distortional collapse mechanism and
thenonlinearportionrepresentsacollapsemechanismdrivenbydistortionalbuckling.
ThisproposedmodificationtotheDSMdistortionalpredictioncurvewillbecompared
against the beam experiments database developed in Section 4.3.1 as a part of several
proposedDSMoptionsconsideredlaterinthischapter.
1.4
1.4
DSM (no hole)
0.8
0.8
0.6
0.6
0.4
0.4
0.2
0.2
0.5
1.5
2.5
DSM (no hole)

1.2
Mtest/My
Mtest/My
1.2
distortional slenderness, d=(My /Mcrd)0.5
0.5
1.5
2.5
Figure8.51Comparisonofsimulatedbeamstrengths(Inet/Ig=1.0)to(a)theexistingDSMdistortional
bucklingdesigncurveandto(b)theproposedDSMdistortionalbucklingcurveforbeamswithholes
1.4
1.4
DSM (no hole)
0.8
0.8
0.6
0.6
0.4
0.4
0.2
0.2
0.5
1.5
2.5
DSM (no hole)

1.2
Mtest/My
Mtest/My
1.2
0.5
1.5
2.5
384
1.4
1.4
DSM (no hole)
0.8
0.8
0.6
0.6
0.4
0.4
0.2
0.2
0.5
1.5
2.5
DSM (no hole)

1.2
Mtest/My
Mtest/My
1.2
0.5
1.5
2.5
8.2.4 Presentation and evaluation of DSM options
SixoptionsforextendingDSMtolaterallybracedbeamswithholesareevaluatedin
this section. The options range from simple substitutions in the existing code to more
involved modifications, including the incorporation of the design curve transitions
discussedinSection8.2.2andSection8.2.3forlocalanddistortionalbuckling.
385
8.2.4.1
DescriptionofDSMoptions
Option 1: Include hole(s) in Mcr determinations, ignore hole otherwise

This method, in presentation, appears identical to currently available DSM expressions
Lateral-Torsional Buckling
The nominal flexural strength, Mne, for lateral-torsional buckling shall be calculated in accordance with the
following:
(a) for Mcre < 0.56 My
Mne = Mcre
(b) for 2.78My Mcre 0.56My M = 10 M 1 10M y

ne
y
(c) for Mcre >2.78My
36M cre
Mne= My
where
Mcre= Critical elastic global beam buckling load (including hole(s))
Local Buckling
(a) For l 0.776
Mnl = Mne
(b) For l > 0.776
Mnl = 1 0.15 M crl
where
l
0.4
M ne
M
crl
M ne
0.4
M ne
= M ne M crl

Mne=
defined in section above
following:
(a) For d 0.673
Mnd = M y
(b) For d >0.673
M crd
Mnd =
1 0.22 M
y
where
d
Mcrd
0. 5
M
crd
M
y
0 .5
My
M y M crd
386
Option 2: Include hole(s) in Mcr determinations, Use Mynet everywhere

The only change in this method is to replace My with Mynet
following:
(a) for Mcre < 0.56 Mynet
Mne = Mcre
(b) for 2.78Mynet Mcre 0.56Mynet M = 10 M 1 10M ynet

ne
ynet
(c) for Mcre >2.78Mynet
36M cre
Mne= Mynet
where
Local Buckling
(a) For l 0.776
Mnl = Mne
(b) For l > 0.776
Mnl = 1 0.15 M crl
where
l
M ne
0.4
M
crl
M ne
0.4
M ne
= M ne M crl

Mne = defined in section above
following:
(a) For dnet 0.673
Mnd = M ynet
(b) For dnet >0.673

0 .5
M crd M crd
Mnd =

1
0
.
22
M M
ynet ynet
where
dnet
= M ynet M crd
Mcrd
0. 5
M ynet
387
Option 3: Cap Mnl and Mnd, otherwise no strength change, include hole(s) in Mcr
This method puts bounds in place and assumes local-global interaction happens at full Mne
following:
Mne = Mcre

ne
y
36M cre
Mne= My
where
Local Buckling
(a) For l 0.776
Mnl = Mne Mynet
(b) For l > 0.776
Mnl = 1 0.15 M crl
where
l
M ne
0.4
M
crl
M ne
0.4
M ne
= M ne M crl

Mne=
following:
(a) For d 0.673
Mnd = My Mynet
(b) For d >0.673

0. 5
M crd M crd
Mnd =
1 0.22 M M
y
where
d
= M y M crd
Mcrd
0 .5
My
388
Option 4: Cap Mnl, transition Mnd, include hole(s) in Mcr determinations

This method puts bounds and transition in place, assumes local-global interaction at full Mne
following:
Mne = Mcre

ne
y
36M cre
Mne= My
where
Local Buckling
(a) For l 0.776
Mnl = Mne Mynet
(b) For l > 0.776
Mnl = 1 0.15 M crl

M
ne
where
l
0.4
M
crl
M ne
0.4
M ne
= M ne M crl

Mne=
The nominal flexural strength, Mnd, for distortional buckling shall be calculated in with the following:
(a) For d d1
Mnd = M ynet
(b) For d1<dd2
ynet
d2
Mnd = M
ynet
d1
d2
M
(c) For d > d2

0. 5
Mnd = 1 0.22 M crd M crd
M M
y y
where
d
= M y M crd
d1
d2
d2
0 .5
My
= 0.673( M ynet M y )
= 0.673 1.7 (M ynet M y )1.7 0.7

= (1 0.22(1 d 2 ))(1 d 2 )M y
Sfnet
= Section modulus at the hole(s) referenced to the extreme fiber at first yield
Mcrd
389
Option 5: Transition Mnl (Option A), transition Mnd, include hole(s) in Mcr determinations
following:
Mne = Mcre

ne
y
36M cre
Mne= My
where
Local Buckling
(a) For l l1
Mnl = Mne Mynet
(caponbeamstrength)
(b) For l1<ll2
l 2 l l1
Mnl = M
ynet M ynet M l 2
l l 2 l1
(c) For l > l2
Mnl = 1 0.15 M crl
where
l
l1
M ne
0.4
M
crl
M ne
(nonlinearyieldtransitionwhenMynet/Mne1)
0.4
M ne
= M ne M crl
= 0.776( M ynet M ne ) 0.776
l2 = 0.776 2.4(M ynet M ne )3.5 1.4 , Mynet/Mne1

= 0.776, Mynet/Mne>1
l2 = 1 0.15(1 l 2 )
0.8
)(1
l2
0.8
(notransitionwhenMynet/Mne>1)
M ne
SameasOption4
390
Option 6: Transition Mnl (Option B), transition Mnd, include hole(s) in Mcr determinations
2
for c 1.5 Pne = 0.658 c Py
0.877
for c > 1.5 Pne = 2 Py = 0.877Pcre

c
where
Py Pcre

Py
= AgFy
Ag
Local Buckling
The nominal axial strength, Mnl, for local buckling shall be calculated in accordance with the following:
(a) For l l1
Mnl = Mynet (Mne /My)
(b) For l1<ll2
Mnl = M M ne M M ne M l 2 l l1
ynet
ynet
l2
M
y
M
y
(c) For l > l2
Mnl = 1 0.15 M crl
where
l
l1
M
ne
0.4
M
crl
M ne
l l 2 l1
0.4
M ne
= M ne M crl
= 0.776( M ynet M y ) 0.776
l2 = 0.776 2.4(M ynet M y )3.5 1.4
(nonlinearyieldtransition)
l2 = 1 0.15(1 l 2 ) (1 l 2 ) M ne
0.8
0.8
SameasOption4
391
8.2.5 DSM comparison to beam test simulation database
The six DSM prediction options for coldformed steel beams with holes are
evaluated with the simulated laterally braced beam experiment database developed in
Section8.2.1andsummarizedinAppendixL.Thesimulateddataiscomparedagainst
DSMpredictionswhileevaluatingdatatrendsagainstmemberslendernessandholesize
(Inet/Ig),andspantodepthratio(L/H).
Figure8.51andFigure8.52comparethesimulatedtestdatatopredictionsforlocal
anddistortionalbucklingcontrolledbeamfailures.Option1isidenticaltotheexisting
DSM approach for beams without holes, except the critical elastic buckling loads (Mcrl,
Mcrd,andMcre)aredeterminedwiththeinfluenceofholes.Option1isobservedtobea
accurate predictor of local buckling controlled failure strengths, although distortional
predictionsareconservativewhendishighandunconservativebyasmuchas20%asd
decreases below 1.5 (see Figure 8.52). The unconservative predictions occur because
Option1doesnotaccountforthecolumnstrengthlimitMynet,nordoesitaccountfora
transitionfromanelasticbucklingcontrolledfailuretoayieldcontrolledfailureat the
netsection.
Option2isobservedtobeaconservativepredictorinFigure8.51andFigure8.52for
high l, and d and demonstrates improved accuracy over Option 1 when slenderness
decreasesandholesizeincreases(seeFigure8.52).Option2replacesMyneteverywhere
within the existing DSM formulation, which has the effect of increasing l and d and
decreasingpredictedstrength.Option3testtopredictedtrendsaresimilartoOption1
with increasingly unconservative predictions as slenderness decreases, demonstrating
392
that the Mynet limits on Mnl and Mnd in Option 3 are not fully effective at capturing the
yield transition to the net section. Option 4 is identical to Option 3 except the yield
transition on the DSM distortional curve is employed to provide a more accurate
predictionofthenetsectionyieldinginfluence.Option4demonstratesanimprovement
in distortional bucklingcontrolled prediction accuracy when d < 1, although the
strengthof11beamsareunderpredictedbyupto15%whend=1.3.Option5accurately
predictsthestrengthofthese11beamswiththeaddedtransitiononthelocalbuckling
design curve. (Option 6 is the same as Option 5 because the beams considered are
laterallybraced).
Table 8.1 summarizes the testtopredicted ratio statistics and strength reduction
factorforthesixDSMoptions(seeEq.(8.1)foradefinitionof).Nooneoptionstands
outabovetherestwhenstudyingthetable,althoughtheobservationsfromFigure8.51
and Figure 8.52 support Options 3,4, and 5(6) as the methods most closely tied to
underlying collapse mechanisms at ultimate limit state. The observations from this
comparisonwillbeemployedalongwiththeDSMcomparisontothebeamexperimental
databaseinthenextsectiontosupporttherecommendedDSMmodifications.
Table8.4DSMtesttopredictedstatisticsforlaterallybracedbeamsimulations
Option
Description
Local buckling
SD
0.09
0.89
# of tests
44
Mean
1.06
SD
0.13
0.86
# of tests
160
My everywhere
Mean
1.07
Mynet everywhere
1.05
0.10
0.88
50
1.07
0.12
0.86
154
Cap Mnl, Mnd
1.07
0.09
0.89
44
1.06
0.13
0.86
160
Transition Mnd, Cap Mnl
1.07
0.09
0.89
44
1.06
0.13
0.86
160
Transition Mnd and Mnl (Option A, B)
1.01
0.11
0.87
72
1.09
0.12
0.87
132
5,6
393
1.5
1.5
Option 1 - My everywhere
Option 2 - Mynet everywhere

1
Mn/My
Mn/My
0.5
0.5
1.5
2.5
0.5
0.5
1.5
l=(My /Mcrl)0.5
1.5
Option 4 - cap Mn , transition Mnd

l
0.5

0.5
1.5
2.5
0.5
0.5
1.5
0.5
2.5
2.5
l=(My /Mcrl)0.5
l=(My /Mcrl)
1.5
1.5
Option 5 - transition Mn (Option A), transition Mnd
Option 6 - transition Mn (Option B), transition Mnd
1
Mn/My
1
Mn/My
1
Mn/My
Mn/My
0.5
2.5
1.5
Option 3 - cap Mn , Mnd
l
l=(My /Mcrl)0.5
0.5
1.5
2.5
0.5
0.5
0.5
1.5
l=(My /Mcrl)0.5
l=(My /Mcrl)
Figure8.54Comparisonofsimulatedteststrengthstopredictionsforlaterallybracedbeamswithlocalbucklingcontrolledfailuresasafunctionoflocal
slenderness
394
1.5
1.5

1
Mn/My
Mn/My
0.5
0.5
1.5
2.5
0.5
0.5
d=(My /Mcrd)0.5
2.5
2.5
1
Mn/My
Mn/My
2.5
0.5
0.5
1.5
2.5
0.5
0.5
0.5
1.5
d=(My /Mcrd)0.5
d=(My /Mcrd)
1.5
1.5
1
Mn/My
1
Mn/My
1.5
l
0.5
1.5
dnet=(Mynet/Mcrd)0.5
1.5

0.5
1.5
2.5
0.5
0.5
0.5
1.5
d=(My /Mcrd)0.5
d=(My /Mcrd)
Figure8.55Comparisonofsimulatedteststrengthstopredictionsforlaterallybracedbeamswithdistortionalbucklingcontrolledfailuresasafunctionof
distortionalslenderness
395
1.5
Mtest/Mn
Mtest/Mn
1.5
0.5
0.5
0
0.5
1.5
2.5
0.5
1.5
1.5
1.5
0.5
0.5
1.5
2.5
0.5
Mtest/Mn
Mtest/Mn
1.5
0.5
2.5
2.5
0.5
1.5
1.5
l=(My /Mcrl)0.5
1.5
0.5
l=(My /Mcrl)
0.5
2.5
0.5

l
0
l=(My /Mcrl)0.5
Mtest/Mn
Mtest/Mn
l=(My /Mcrl)0.5
Local Controlled
2.5
0.5
0.5
1.5
l=(My /Mcrl)0.5
l=(My /Mcrl)
Figure8.56Testtopredictedratiosforlocalbucklingcontrolledsimulatedlaterallybracedbeamfailuresasafunctionoflocalslenderness
396
1.5
Mtest/Mn
Mtest/Mn
1.5
0.5
0.5
0
0.5

1.5
2.5
0.5
1.5
1.5
0.5
0.5
1.5
2.5
0.5
Mtest/Mn
Mtest/Mn
1.5
0.5
2.5
2.5
0.5
1.5
1.5
d=(My /Mcrd)0.5
1.5
0.5
d=(My /Mcrd)
2.5
0.5
0.5

l
0
1.5
Mtest/Mn
Mtest/Mn
d=(My /Mcrd)0.5
2.5
0.5
0.5
1.5
d=(My /Mcrd)0.5
d=(My /Mcrd)
Figure8.57Testtopredictedratiosfordistortionalbucklingcontrolledsimulatedlaterallybracedbeamfailuresasafunctionofdistortionalslenderness
397
1.5
Mtest/Mn
Mtest/Mn
1.5
0.5
0.5
0
0.5
0.6
0.7
0.8
0.9
0
0.5
0.6
0.7
1.5
1.5
0.5
0.7
0.8
0.9
0
0.5
0.6
0.7
0.9
0.9
Inet/Ig
1.5
Mtest/Mn
Mtest/Mn
0.8
1.5
0.5
0.5
0.6
Inet/Ig
0
0.5
0.9
0.5

l
0.6
0.8
Inet/Ig
Mtest/Mn
Mtest/Mn
Inet/Ig
0
0.5
Local Controlled
0.7
0.8
0.9
0
0.5
Inet/Ig
0.6
0.7
0.8
Inet/Ig
Figure8.58Testtopredictedratiosforsimulatedlocalbucklingcontrolledlaterallybracedbeamfailuresasafunctionofnetcrosssectionalmomentofinertiato
grosscrosssectionalmomentofinertia
398
1.5
Mtest/Mn
Mtest/Mn
1.5
0.5
0.5
0
0.5
0.6
0.7
0.8
0.9
0
0.5
0.6
0.7
1.5
1.5
0.5
0.7
0.8
0.9
0
0.5
0.6
0.7
0.9
0.9
Inet/Ig
1.5
Mtest/Mn
Mtest/Mn
0.8
1.5
0.5
0.5
0.6
Inet/Ig
0
0.5
0.9
0.5

l
0.6
0.8
Inet/Ig
Mtest/Mn
Mtest/Mn
Inet/Ig
0
0.5
0.7
0.8
0.9
0
0.5
Inet/Ig
0.6
0.7
0.8
Inet/Ig
Figure8.59Testtopredictedratiosforsimulateddistortionalbucklingcontrolledlaterallybracedbeamfailuresasafunctionofnetcrosssectionalmomentof
inertiatogrosscrosssectionalmomentofinertia
399
1.5
Mtest/Mn
Mtest/Mn
1.5
0.5
0.5
0
10
1.5
1.5
0.5
10
Mtest/Mn
Mtest/Mn
10
10
L/H
1.5
0.5
0.5
1.5
10
L/H
0.5

l
0
6
L/H
Mtest/Mn
Mtest/Mn
L/H
Local Controlled
10
L/H
6
L/H
Figure8.60Testtopredictedratiosforsimulatedlocalbucklingcontrolledlaterallybracedbeamfailuresasafunctionofcolumnlength,L,tobeamdepth,H
400
1.5
Mtest/Mn
Mtest/Mn
1.5
0.5
0.5
0
10
1.5
1.5
0.5
10
Mtest/Mn
Mtest/Mn
10
10
L/H
1.5
0.5
0.5
1.5
10
L/H
0.5

l
0
6
L/H
Mtest/Mn
Mtest/Mn
L/H
10
L/H
6
L/H
Figure8.61Testtopredictedratiosforsimulateddistortionalbucklingcontrolledlaterallybracedbeamfailuresasafunctionofcolumnlength,L,toH
401
8.2.6 DSM comparison to experimental beam database
The six DSM options are now compared to the laterally braced beam experiment
database first assembled in Section 4.3.1. The database contains the elastic buckling
propertiesofeachbeam,includingthepresenceofholesandtheinfluenceofboundary
conditions,aswellasthetestedstrengths.Figure8.62throughFigure8.65comparethe
experiment strengthstoDSMpredictionsforlocalanddistortionalbucklingcontrolled
beamfailures.(Thelocalanddistortionalslendernessisobtainedwiththepurelocal
anddistortionalelasticbucklingloadsLandDinthisstudy,nottheLHandDHmodes
describedinSection4.3).Thetestedstrengthsarelowerthanthepredictionsoverawide
range of local and distortional slenderness. These trends were first observed in a
preliminary DSM comparison (Moen and Schafer 2007a), and possible reasons for the
difference between test and predictions were hypothesized, including experimental
error,errorinthedeterminationofelasticbucklingloads,andtheinfluenceoftheangle
strapsonthecalculationofthedistortionalcriticalelasticbucklingload.Thebeamsin
thedatabasehaverelativelysmallholes,withInet/Igrangingfrom0.96to0.99asshownin
Figure8.64andFigure8.65,whichsuggeststhatthepresenceofholesshouldnothavea
significantimpactontestedstrength.Thetesttopredictedstatisticsarethesameforthe
sixDSMoptionsasshowninTable8.5.Itisconcludedthattheexperimentaldatabase,
initscurrentform,cannotbeusedtoevaluatetheproposedDSMmodifications.Future
workisplannedtoinvestigatethedifferencesbetweentheDSMpredictionsandtested
strengthsforthisdata.Inaddition,morerecenttestsoncoldformedsteelbeamswith
402
holes will be added to the database. Experiments on beams with larger holes are also
needed.
Table8.5DSMtesttopredictedratiostatisticsforbeamexperiments
Option
Description
Local buckling
SD
0.12
0.85
# of tests
55
Mean
0.87
SD
0.14
0.81
# of tests
89
My everywhere
Mean
0.88
Mynet everywhere
0.88
0.12
0.85
55
0.87
0.14
0.81
89
Cap Mnl, Mnd
0.88
0.12
0.85
55
0.87
0.14
0.81
89
Transition Mnd, Cap Mnl
0.88
0.12
0.85
55
0.87
0.14
0.81
89
Transition Mnd and Mnl (Option A, B)
0.88
0.12
0.85
55
0.87
0.14
0.81
89
5,6
403
1.5
1.5

1
Mn/My
Mn/My
0.5
0.5
1.5
2.5
0.5
0.5
1.5
l=(My /Mcrl)0.5
1.5

l
0.5

0.5
1.5
2.5
0.5
0.5
1.5
0.5
2.5
2.5
l=(My /Mcrl)0.5
l=(My /Mcrl)
1.5
1.5
1
Mn/My
1
Mn/My
1
Mn/My
Mn/My
0.5
2.5
1.5
l
l=(My /Mcrl)0.5
0.5
1.5
2.5
0.5
0.5
0.5
1.5
l=(My /Mcrl)0.5
l=(My /Mcrl)
Figure8.62Comparisonofexperimentalteststrengthstopredictionsforlaterallybracedbeamswithlocalbucklingcontrolledfailuresasafunctionoflocal
slenderness
404
1.5
1.5

1
Mn/My
Mn/My
0.5
0.5
1.5
2.5
0.5
0.5
d=(My /Mcrd)0.5
2.5
2.5
1
Mn/My
Mn/My
2.5
0.5
0.5
1.5
2.5
0.5
0.5
0.5
1.5
d=(My /Mcrd)0.5
d=(My /Mcrd)
1.5
1.5
1
Mn/My
1
Mn/My
1.5
l
0.5
1.5
1.5

0.5
1.5
2.5
0.5
0.5
0.5
1.5
d=(My /Mcrd)0.5
d=(My /Mcrd)
Figure8.63Comparisonofexperimentalteststrengthstopredictionsforlaterallybracedbeamswithdistortionalbucklingcontrolledfailuresasafunctionof
distortionalslenderness
405
1.5
Mtest/Mn
Mtest/Mn
1.5
0.5
0.5
0
0.5
0.6
0.7
0.8
0.9
0
0.5
0.6
0.7
1.5
1.5
0.5
0.7
0.8
0.9
0
0.5
0.6
0.7
0.9
0.9
Inet/Ig
1.5
Mtest/Mn
Mtest/Mn
0.8
1.5
0.5
0.5
0.6
Inet/Ig
0
0.5
0.9
0.5

l
0.6
0.8
Inet/Ig
Mtest/Mn
Mtest/Mn
Inet/Ig
0
0.5
Local Controlled
0.7
0.8
0.9
0
0.5
Inet/Ig
0.6
0.7
0.8
Inet/Ig
Figure8.64Testtopredictedratiosforexperimentallocalbucklingcontrolledlaterallybracedbeamfailuresasafunctionofnetcrosssectionalmomentofinertia
togrosscrosssectionalmomentofinertia
406
1.5
Mtest/Mn
Mtest/Mn
1.5
0.5
0.5
0
0.5
0.6
0.7
0.8
0.9
0
0.5
0.6
0.7
1.5
1.5
0.5
0.7
0.8
0.9
0
0.5
0.6
0.7
0.9
0.9
Inet/Ig
1.5
Mtest/Mn
Mtest/Mn
0.8
1.5
0.5
0.5
0.6
Inet/Ig
0
0.5
0.9
0.5

l
0.6
0.8
Inet/Ig
Mtest/Mn
Mtest/Mn
Inet/Ig
0
0.5
0.7
0.8
0.9
0
0.5
Inet/Ig
0.6
0.7
0.8
Inet/Ig
Figure8.65Testtopredictedratiosforexperimentaldistortionalbucklingcontrolledlaterallybracedbeamfailuresasafunctionofnetcrosssectionalmomentof
inertiatogrosscrosssectionalmomentofinertia
407
8.2.7 Recommendations DSM for beams with holes
Options3,4,5,and6arepresentedasviableproposalsforextendingDSMtobeams
with holes. This recommendation is based on the testtopredicted statistics and data
trends from the simulation studies presented in Section 8.2.5, and also considers the
efforttoimplementthemodificationsandtheireaseofusebydesignengineers.Option
3accountsforthereductioninbeamstrengthfromthepresenceofholesbylimitingMnl
and Mnd to Mynet. This is a simple modification to implement in the Specification and
avoids additional calculation work for a design engineer (except for that required to
calculatethecriticalelasticbucklingloadsincludingtheinfluenceoftheholes).Options
4and5arerefinementsofOption3,wherethecaponMnlandMndbecomesatransition
from an elastic buckling controlled failure mode to a yield controlled failure at Mynet.
These two methods require additional effort from the designer when compared to
Option3,buttheyhaveanimportantadvantage.Options4,5,and6aremoreclosely
tied to the failure mechanisms influencing column strength because they capture the
yield transition to the net section in their predictions. The transitions increase the
probability that strength will be accurately predicted for general beam and hole
geometries.Option5hastheadditionaladvantageofcapturingtheinfluenceofayield
transition for closed crosssections that do not experience distortional buckling.
Additionalnonlinearfiniteelementsimulationsandexperimentsareneededtovalidate
the proposed modifications to theDirect Strength Method for beamssubject to lateral
torsionalbucklingatultimatelimitstate.
408
Chapter 9
Conclusions and proposed future work
9.1 Conclusions
ProposedDirectStrengthMethoddesignequationsarenowinplaceforcoldformed
steelmemberswithholes.Thedevelopmentofthemethodwasinitiatedwiththinshell
finite element eigenbuckling studies in ABAQUS on thin plates and full coldformed
steelmemberswithholes.Thebucklingoftheunstiffenedstripsadjacenttoaholeina
thin plate influenced, and sometimes controlled, the critical elastic buckling stress of
individualcrosssectionelements.Unstiffenedstripbucklingwasalsocloselyassociated
withdistortionalbucklingmodesatthelocationoftheholesinCsectioncolumnsand
beams. Large holes and closelyspaced holes locally stiffened thin rectangular plates
and the webs of Csection columns, resulting in buckling away from the holes. The
elasticbucklingstudiesledtousefuldesignguidelinesandtools,includingholespacing
limits (which prevent cumulative reductions in elastic stiffness along the length of a
member)andsimplifiedelasticbucklingpredictionmethodsforlocal,distortional,and
globalbucklingdevelopedasanalternativetofiniteelementeigenbucklinganalysis.
TheviabilityoftheDSMframeworkforcoldformedsteelmemberswithholeswas
established early in this research using existing test results and the elastic buckling
properties of coldformed steel column and beam specimens with holes. Additional
409
experimental work evaluated the influence of holes on the loaddeformation response

and failure mechanisms for short and intermediate length Csection columns. During
the experiments, holes were observed to locally stiffen the web of the intermediate
lengthCsectioncolumnsandpreventeddynamicmodeswitching(fromlocalbuckling
to distortional buckling) near peak load. Holes were also observed to decrease post
peakductilityforcolumnswhentheholesizewaslargerelativetothewebwidth(e.g.,
the362S16233specimens).
Results from the experimental program were used to validate a nonlinear finite
elementmodelingprotocol.Aconcertedeffortwasmadetosimulatetheinitialstateofa
coldformedsteelmemberintheprotocol,includingimperfectionmagnitudesbasedon
measurement statistics and residual stresses and initial plastic strains from the cold
forming process predicted with a mechanicsbased approach. The nonlinear finite
elementmodelingcapabilitywasusedtoconstructalargedatabaseofsimulatedcolumn
and beam experiments with a wide range of hole sizes, spacings, and Csection
dimensions.Simulationresultsdemonstratedthatascrosssectiondistortionalor local
slendernessdecreased,thefailureofacoldformedsteelmemberwithholesoccurredby
yielding and collapse of the unstiffened stripsat thenet crosssection. Collapse of the
unstiffened strips sometimes triggered unstable global failure modes in columns with
large holes, i.e., as hole size approached Anet=0.60Ay. (Global instabilities caused by
yieldingatpeakloadwerenotstudiedforbeamswithholesinthisthesis,onlylaterally
bracedbeamswereconsidered.)ModificationstothelocalanddistortionalDSMcurves
were made to account for this unique netsection failure mechanism with a deliberate
410
transitionandcaponmemberstrength.ThefinalproposedDSMmethodformembers
withholeswasvalidatedwithexistingexperimentaldataandthesimulatedexperiments
database.
9.2 Futurework
Severalinterestingfutureresearchtopicsresultedfromtheelasticbucklingstudies,
experiments,andnonlinearfiniteelementsimulationsinthisthesis.Futureresearchis
plannedtofollowuponmanyoftheseideasandquestions.Themajorpointsoffuture
study,organizedbyresearchtopic,arelistedbelow.
ThinshellfiniteelementmodelinginABAQUS(Chapter2)
The S9R5 meshing guidelines developed in this thesis were developed primary for
eigenbucklinganalyses.Meshingguidelineswhichensureaccurateresultsinnonlinear
finite element simulations are also needed. Studies are ongoing to develop rules for
determining the minimum number of throughthickness finite element integration
points, themesh density required for linear and quadratic finite element formulations,
andlimitsoninitialelementdistortionandcurvature.
Elasticbucklingofcoldformedsteelcrosssectionalelementswithholes(Chapter3)
1.
2.
The simplified elastic buckling prediction method presented in this thesis for
unstiffened elements loaded with uniaxial compression is empirically derived.
A mechanicsbased unstiffened element prediction method is warranted as a
topicoffutureresearchtoimprovethegeneralityofthemethod.
Anelementbasedelasticbucklingpredictionmethodwhichaccountsforstress
gradientsonunstiffenedelementswithholesisneededtoaddressadesigncase
engineersmayencounterinpractice.
411
3.
4.
Elastic buckling studies are planned to develop elementbased simplified

methodsforholepatternsfoundinstorageracks.
The elementbased elastic buckling prediction methods provide a convenient
method to calculate Fcr (the critical elastic buckling stress) for general hole
shapes,sizes,andspacingsforuseintheAISIS10007effectivewidthmethod.
Work is planned to evaluate introduce these simplified approaches into the
effectivewidthmethod.
Elasticbucklingofcoldformedsteelmemberswithholes(Chapter4)
1. Yu and Davis, OrtizColberg, Rhodes and MacDonald, Rhodes and Schnieder,
andPuetal.performedtestsoncolumnspecimenswithmultiplediscreteholes
or hole patterns. The elastic buckling properties and tested strengths of these
specimenswillbeaddedtotheexperimentdatabase,inadditiontotestsonrack
sections.
2. Automated elastic buckling modal identification tools are needed to identify
local, distortional, and global buckling modes in thinshell finite element
eigenbuckling analysis. Research is ongoing to develop this capability with an
implementationsimilartothatoftheconstrainedfinitestripmethod.
3. Work continues on the development and validation of the CUFSM elastic
bucklingapproximatemethodsdevelopedandtheextensionofthesemethodsto
members with hole patterns (e.g., storage racks). A general procedure for
implementing CUFSM constraints in the local buckling prediction method is
needed.Also,thecurrentassumptionthatthewarpingtorsionconstantCw=0ata
hole produces conservative global elastic buckling predictions for columns and
beams. Additional research is needed to derive a mechanicsbased
approximationforCwatahole.
Experimentsoncoldformedsteelcolumnswithholes(Chapter5)
1.
2.
Amoredefinitivemethodofmeasuringthebasemetalthicknessofcoldformed
steelmemberswithazincgalvaniccoatingisneeded.Currentstandardpractice
istoremovethezinccoatingwithhydrochloricacidoraferricchloridesolution.
It is difficult to know when all of the zinc has been removed though since the
zinc chemically interacts with the base metal during the initial application.
Experiments are planned to determine the influence of the zinc coating on
ultimatestrength.
Research work is planned to evaluate the influence of sheet coiling on the
measuredyieldstressintensilecoupons.IthasbeenhypothesizedbyProfessor
Rasmussen at the University of Sydney that the same coiling curvature which
causes residual stresses in coldformed steel members also affects yield stress
measurementsintensiletests.
412
Residualstressesandplasticstrainsincoldformedsteelmembers(Chapter6)
1. Experimental work is planned to validate the prediction model presented in
Chapter6relatingcoilingresidualstressestothecoilingradius,sheetthickness,
andyieldstress.
2. Research is ongoing to evaluate how ABAQUS metal plasticity laws use the
residualstressandinitialplasticstraininformationandtodetermineifkinematic
hardeningoradifferentmixedhardeningruleisrequiredtoaccuratelysimulate
thecoldworkofformingeffectonloaddeformationresponse.
3. Nonlinearfiniteelementstudiesareplannedtoidentifytheinfluenceofthrough
thicknessresidualstressesandplasticstrainsontheloaddeformationresponse,
ultimate strength, and failure mechanisms of coldformed steel beams and
columns.
4. Hancocketal.providesamethodwhichaccountsforthecoldworkofforming
in the corners of coldformed steel crosssections when calculating ultimate
strength(Hancocketal.2001).TheresearchinChapter6providesnewinsight
intotherelationshipbetweenresidualstressesandinitialplasticstrainsfromthe
manufacturing process. Research work is planned to revisit Hancocks cold
work of forming method to determine if it can be supplemented with this new
research.
5. Thecurrentresidualstresspredictionmethodassumesanelasticperfectlyplastic
materialmodel.Researchworkisongoingtointroducetheeffectofsteelstrain
hardeningintothepredictionmethod.
Nonlinearfiniteelementmodelingofcoldformedsteelmembers(Chapter7)
1. Theuseofmeasuredimperfectionmagnitudesinsteadofstatisticaldistributions
is warranted as a topic of future study, especially the use of a flared cross
section,includingflangewebanglesoffof90degrees.
2. InitiatingplasticityinABAQUSatthematerialsproportionallimitreducedthe
predictedstrengthbyupto20%whencomparedtoexperimentsinChapter7.A
study is planned to simulate a single finite element in tension to evaluate the
ABAQUS implementation of metal plasticity and determine the source of the
discrepancy.
TheDirectStrengthMethodformemberswithholes(Chapter8)
1.
2.
AdditionalvalidationstudiesareplannedtocomparetheproposedDSMHoles
methodologytotheAISIS10007effectivewidthdesignmethod.
Nonlinear finite element studies of other DSM prequalified crosssections (e.g.,
Zsectionsandhatsections)aswellasracksectionswithholepatternsarealso
plannedtoexpandthesimulationdatabase.
413
414
References
ABAQUS.(2007a).ABAQUS/StandardVersion6.71.,DassaultSystmes,
http://www.simulia.com/,Providence,RI.
ABAQUS.(2007b).Personalcommunication,S9R5elementsectionnumbers.Dassault
Systmes,http://www.simulia.com/,Providence,RI.
AISIS100.(2007).NorthAmericanSpecificationfortheDesignofColdFormedSteelStructural
Members,AmericanIronandSteelInstitute,Washington,D.C.
AISITS202.(2001).StubColumnTestMethodforEffectiveAreaofColdformedSteel
Columns.2001NorthAmericanColdFormedSteelSpecification,AmericanIronand
SteelInstitute,Washington,D.C.
AISI.(2006).DirectStrengthMethodDesignGuide,AmericanIronandSteelInstitute,
Washington,D.C.
ASTM.(2004).E8M04,StandardTestMethodsforTensionTestingofMetallicMaterials
(Metric).ASTMInternational,WestConshohocken,PA.
Batson,K.D.(1992).FlexuralBehaviorofWebswithOpenings,M.S.Thesis,Universityof
MissouriRolla,Rolla.
Bernard,E.S.(1993).FlexuralBehaviourofColdFormedProfiledSteelDecking,Ph.D.Thesis,
UniversityofSydney,Australia,Sydney.
Chajes,A.(1974).PrinciplesofStructuralStability,PrenticeHallCollegeDiv,EnglewoodCliffs,
NJ.
Chen,W.F.,andHan,D.J.(1988).PlasticityforStructuralEngineers,SpringerVerlag,NewYork,
NY.
Cook,R.D.(1989).ConceptsandApplicationsofFiniteElementAnalysis,J.Wiley&Sons,New
York,NY.
CRC.(2003).StandardMathematicalTablesandFormulae,CRCPress,NewYork,NY.
Crisfield,M.A.(1981).Afastincremental\iterationsolutionprocedurethathandlessnap
through.ComputersandStructures,13,5562.
Dat,D.T.(1980).TheStrengthofColdFormedSteelColumns.CornellUniversityDepartment
ofStructuralEngineeringReportNo.804,Ithaca,NY.
415
DeBatista,E.M.,andRodrigues,F.C.(1992).Residualstressmeasurementsoncoldformed
profiles.ExperimentalTechniques,16(5),2529.
Galambos,T.(1998a).AppendixB.3,TechnicalMemorandumNo.3:StubColumnTest
Procedure.GuidetoStabilityDesignCriteriaforMetalStructures,5thEdition,John
Wiley&Sons,NewYork,NY.
Galambos,T.(1998b).GuidetoStabilityDesignCriteriaforMetalStructures,5thEdition,John
Wiley&Sons,NewYork,NY.
Hancock,G.J.,Murray,T.,andEllifritt,D.(2001).ColdformedsteelstructurestotheAISI
specification,MarcelDekker,Inc.,NewYork,NY.
Hill,R.(1950).TheMathematicalTheoryofPlasticity,OxfordUniversityPress,London,England.
Ingvarsson,L.(1975).Coldformingresidualstresses,effectofbuckling.Proceedings,Third
AnnualSpecialtyConferenceonColdFormedSteelStructures,UniversityofMissouriRolla,
85119.
Key,P.,andHancock,G.J.(1993).ATheoreticalInvestigationoftheColumnBehaviourof
ColdFormedSquareHollowSections.ThinWalledStructures,16,3164.
Kwon,Y.B.(1992).PostBucklingBehaviourofThinWalledSections,Ph.D.Thesis,University
ofSydney,Australia.
Labview.(2005).Labview,Version8.NationalInstruments,www.labview.com,Austin,TX.
Mathworks.(2007).Matlab7.5.0(R2007b).Mathworks,Inc.,www.mathworks.com.
Moen,C.D.,andSchafer,B.W.(2007a).Directstrengthdesignforcoldformedsteelmembers
withperforations,ProgressReport#3.AmericanIronandSteelInstitute,Washington,
D.C.
Moen,C.D.,andSchafer,B.W.(2007b).Directstrengthdesignforcoldformedsteelmembers
withperforations,ProgressReport#4.AmericanIronandSteelInstitute,Washington,
D.C.
Oden,J.T.,andMartins,J.A.C.(1985).Modelsandcomputationalmethodsfordynamicfriction
phenomena.ComputerMethodsinAppliedMechanicsandEngineering,52,527634.
OrtizColberg,R.A.(1981).TheLoadCarryingCapacityofPerforatedColdFormedSteel
Columns,M.S.Thesis,CornellUniversity,Ithaca,NY.
Porter,F.C.(1991).ZincHandbook,MarcelDekker,NewYork,NY.
Powell,G.,andSimons,J.(1981).Improvediterativestrategyfornonlinearstructures.
InternationalJournalforNumericalMethodsinEngineering,17,14551467.
416
Quach,W.M.,Teng,J.G.,andChung,K.F.(2004).Residualstressesinsteelsheetsdueto
coilinganduncoiling:aclosedformanalyticalsolution.EngineeringStructures,26,1249
1259.
Quach,W.M.,Teng,J.G.,andChung,K.F.(2006).Finiteelementpredictionsofresidual
stressesinpressbrakedthinwalledsteelsections.EngineeringStructures,28,16091619.
Ramm,E.(1981).Strategiesfortracingnonlinearresponsenearlimitpoints.NonlinearFinite
ElementAnalysisinStructuralMechanics:ProceedingsoftheEuropeUSWorkshop,6389.
Schafer,B.W.(1997).Coldformedsteelbehavioranddesign:analyticalandnumerical
modelingofelementsandmemberswithlongitudinalstiffeners,Ph.D.Thesis,Cornell
University,Ithaca,NY.
Schafer,B.W.,anddny,S.(2006).Bucklinganalysisofcoldformedsteelmembersusing
CUFSM:conventionalandconstrainedfinitestripmethods.EighteenthInternational
SpecialtyConferenceonColdFormedSteelStructures,Orlando,FL.
Schafer,B.W.,andPekz,T.(1998).Computationalmodelingofcoldformedsteel:
characterizinggeometricimperfectionsandresidualstresses.JournalofConstructional
Research,47,193210.
Schafer,B.W.,Sarawit,A.,andPekz,T.(2006).Complexedgestiffenersforthinwalled
members.JournalofStructuralEngineering,132(2),212226.
Schuster,R.M.(1992).TestingofPerforatedCStudSectionsinBending.Universityof
Waterloo,Ontario,Canada.
Shan,M.Y.,andLaBoube,R.A.(1994).BehaviorofWebElementsWithOpeningsSubjectedto
Bending,Shear,andtheCombinationofBendingandShear.CivilEngineeringStudy942,
ColdformedSteelSeries,UniversityofMissouriRolla,Rolla.
Shanley,F.R.(1957).StrengthofMaterials,McGrawHillBookCompany,NewYork,NY.
Silvestre,N.,andCamotim,D.(2005).Localplateanddistortionalpostbucklingbehaviorof
coldformedsteellippedchannelcolumnswithintermediatestiffeners.Seventeenth
InternationalSpecialtyConferenceonColdFormedSteelStructures:RecentResearchand
DevelopmentsinColdFormedSteelDesignandConstruction,Orlando,FL,UnitedStates,1
18.
SSMA.(2001).ProductTechnicalInformation,ICBOER4943P.SteelStudManufacturers
Association,www.ssma.com.
Timoshenko,S.P.,Gere,JamesM.(1961).TheoryofElasticStability,McGrawHill,NewYork.
USSteel.(1985).TheMaking,Shaping,andTreatingofSteel,10thEdition,Herbick&Held,
Pittsburgh,PA.
417
vonKarman,T.,Sechler,E.E.,andDonnell,L.H.(1932).Thestrengthofthinplatesin
compression.TransactionsASME,54(APM545).
Wang,X.P.,Lam,S.S.E.,andChung,K.F.(2006).Crosssectiondistortionduetocuttingofcold
formedsteellippedCsection.ThinWalledStructures,44,271280.
Weng,C.C.,andPekz,T.(1990).ResidualStressesinColdFormedSteelMembers.ASCE
JournalofStructuralEngineering,116(6),16111625.
WesternStatesClayProductsAssociation.(2004).DesignGuideforAnchoredBrickVeneerover
SteelStudSystems,www.brick_wscpa.org,Seattle,WA.
Winter,G.(1947).Strengthofthinsteelcompressionflanges.CornellUniversityEngineering
ExperimentStationReprintNo.32.
Yu,C.(2005).Distortionalbucklingofcoldformedsteelmembersinbending,Ph.D.Thesis,
JohnsHopkinsUniversity,Baltimore.
Yu,C.,andSchafer,B.W.(2006).Distortionalbucklingtestsoncoldformedsteelbeams.ASCE
JournalofStructuralEngineering,132(4),515528.
Yu,C.,andSchafer,B.W.(2007).EffectofLongitudinalStressGradientsonElasticBucklingof
ThinPlates.ASCEJournalofStructuralEngineering,133(4),452463.
Yu,W.W.(2000).ColdFormedSteelDesign,JohnW.Wiley&Sons,NewYork,NY.
Ziemian,R.,andMcGuire,W.(2005).MASTANversion3.0.4,www.mastan.com.Bucknell
University,Lewisburg,PA.
418
Appendix A
ABAQUS input file generator in Matlab
The finite element models in this thesis were generated with a custom Matlab
program which assembles a column or beam with any general crosssection (input in
CUFSMstyleformat)usingninenodeS9R5thinshellfiniteelements.Theuserhasthe
abilitytoaddholesatspecificlocationsinthemember,dictatetheboundaryconditions
and application of load, specify the material properties, and impose imperfection,
residual stresses and plastic strains to define a members initial state. Input files for
eigenbucklinganalysisandnonlinearfiniteelementsimulationscanbegenerated.The
programwasusedthroughoutthisresearchtogenerategroupsofABAQUSinputfiles
forparameterstudies.Theprogramsetupusedtogeneratethenonlinearfiniteelement
modelsofthecolumnexperimentsisprovidedhere.
419
clear all
close all
sourceloc='C:\Documents and Settings\Cris\Desktop\cmoen\Cold Formed Steel - Holes
Research\Fall 2007\runbuck development\Rev_6NL\jhab'
%This example generates ABAQUS input files for the nonlinear

%simulation of 12 different columns with SSMA C-sections and evenly spaced
%slotted holes. The column boundary conditions are pinned-pinned warping free and the
column is
%loaded from both ends with a uniform compressive stress simulated as
%consistent nodal loads on the first two sets of cross-section nodes.
%25%CDF and 75% CDF imperfections are imposed on the member geometry with
%CUFSM local and distortional buckling mode shapes.
addpath([sourceloc
addpath([sourceloc
addpath([sourceloc
addpath([sourceloc
addpath([sourceloc
load
load
load
load
SSMAxsections
SSMAnames
SSMA_wvlengths
Ag
'\functions\filewriting\'])
'\functions\holes\'])
'\functions\'])
'\templates\'])
'\'])
%SSMA
%SSMA
%SSMA
%SSMA
cross section info

name list
cFSM wavelengths for Pcrl, Pcrd
gross cross-sectional area
%define the SSMA sections to create models for

sections=[12
86
11
73
39
95
72
56
47
75
66
87]
%define the imperfection magnitudes
imptypes=[25 75]
%define hole length (slotted holes considered here)
Lhole=4
%define rough hole spacing, will be adjusted in holes section of file
S=12
%define member lengths
Lc=[34
88
24
74
420
42
78
66
56
32
74
40
80]
%define hole depth such that Anet=0.70Ag
Anetfactor=0.7
count=1
for i=1:length(sections)
section_num=sections(i)
for j=1:length(imptypes)
%MEMBER LENGTH
L=Lc(i)
%MESH ALONG LENGTH
nele=L*2;
%NUMBER OF SECTION POINTS THROUGH THE THICKNESS
sectionpoints=5
RB2
%CROSS-SECTION DIMENSIONS
%
Z
%
%
A
%
X
% D2 /
I
\ D1
% RT2/_S2___
S
___S1_\RT1
%
\
|
/
% B2 \
|
/ B1
%
\
|
/
% ___F2_\__________________/_F1___
ABAQUS Y AXIS
%
RB2
H
RB1
%Dimensions are out-to-out, angles are in degrees, t is base metal +
%coating thickness, tbare is base metal thickness
%
[H
B1
B2
D1
D2
F1
F2
S1
S2
RT1
RT2
t
tbare]
dims=SSMAxsections(section_num,2:16)
RB1
%calculate hole depth

hhole=Ag(section_num)*(1-Anetfactor)./dims(15)
%CROSS-SECTION MESHING
%number of elements around the cross section
%[D1
RT1 B1 RB1 H RB2 B2 RT2 D2]
n=[2
2
2
2
16
2
2
2
2];
%CorZ=1 C-section, CorZ=2 Z-section
CorZ=1
[node,elem]=cztemplate(CorZ,dims,n)
nnodes=length(node(:,1)); %Number of FSM cross-section nodes
%Determine FE number of nodes and increment
nL=2*nele+1; %Number of FE nodes along the length
%Determine the node numbering increment along the length
if nnodes<100
FEsection_increment=100; %so along the length the numbering goes up by 100's
else
FEsection_increment=nnodes+1;
end
%ADD ADDITIONAL NODES
nodeadd=[]
%MATERIAL PROPERTIES
421
%steel
matprops(1).name='MAT100';
matprops(1).elastic=[29500 0.3];
matprops(1).plastic=[58.6, 0
64.1517, 0.00342827
68.2188, 0.00842827
72.0304, 0.0134283
77.9752, 0.0234283
82.2224, 0.0334283
85.7249, 0.0434283
88.4053, 0.0534283
90.7405, 0.0634283
92.652, 0.0734283
94.3657, 0.0834283
95.8299, 0.0934283
97.2001, 0.103428 ];
%IMPERFECTIONS
%*****IMPERFECTIONS*****
%type=0
no imperfections
%type=1
use mode shapes from ABAQUS results file
%type=2
input from file
%type 3
impose CUFSM shapes as imperfections
imperfections.type=3;
imperfections.filename=[];
imperfections.step=[];
imperfections.mode=[]
t=dims(15)
if imptypes(j)==25
if L>24
imperfections.magnitude=[0.14*t 0.64*t L/2000]
imperfections.wavelength=[SSMA_wvlengths(section_num,1)
SSMA_wvlengths(section_num,2) L]
else
imperfections.magnitude=[0.14*t 0.64*t]
SSMA_wvlengths(section_num,2)]
end
elseif imptypes(j)==75
if L>24
imperfections.magnitude=[0.66*t 1.55*t L/1000]
SSMA_wvlengths(section_num,2) L]
else
imperfections.magnitude=[0.66*t 1.55*t]
SSMA_wvlengths(section_num,2)]
end
end
imperfections.plumb=[]
imperfections.member=[1] %1 for column, 2 for beam
%DEFINE HOLES
%Add holes to your member.
%hole.type=1 circular
%hole.type=2 rectangular
%hole.type=3 slotted w\radial ends
%hole.dimension=['width or length (ABAQUS x direction)' 'height or diameter']
%hole.location=['CUFSM cross section node (must be odd!)' 'longitudinal location'
'shift hole in direction of height']
%hole.thickness = thickness of finite elements making up hole, usually the same
as the rest of the member
%I've defined two slotted holes here in the web of the cross-section.
%number of holes
nhole=floor(L/S)
if nhole<1
422
nhole=1
end
%final hole spacing
Sfinal=floor(L/nhole)
spacing=Sfinal/2:Sfinal:L-Sfinal/2
hole.type=[3*ones(nhole,1)];
%define dimensions for slotted hole
hole.dimension=[Lhole*ones(nhole,1) hhole*ones(nhole,1)];
%define location of hole in cross-section
hole.location = [(length(node)+1)/2*ones(nhole,1) spacing' zeros(nhole,1)]
hole.thickness = [dims(1,15)*ones(length(hole.type),1)]
hole.material=[100*ones(length(hole.type),1)];
hole.groups=[100000+[1:length(hole.type)]];
hole.fill=[zeros(length(hole.type),1)];
%If you don't want holes, replace above with
%hole=[ ]
%MEMBER END LOADINGS
%Loading notation is similar to CUFSM. Apply P for compression, M for
%moment, or a combination of both. Compression at both ends of
%column are
%shown here. Loads are applied as consistent nodal loads in ABAQUS.
end1load.P=1;
end1load.Mxx=0;
end1load.Mzz=0;
end1load.M11=0;
end1load.M22=0;
end2load.P=-1;
end2load.Mxx=0;
end2load.Mzz=0;
end2load.M11=0;
end2load.M22=0;
%CALCULATE CONSISTENT NODAL LOADS ON MEMBER ENDS*****
unsymm=0
[end1cload, end2cload, A,
Ixx]=consist_endloads(node,elem,end1load,end2load,unsymm, nL, FEsection_increment);
%ABAQUS NODE SETS

%Define these node sets to apply boundary conditions in ABAQUS
%
nodesetinfo={'nodeset name' [xlim1 xlim2 xint] [ylim1 ylim2 yint] [zlim1
zlim2 zint] exclude}
%where nodes are grouped based on xlim1<=x<=xlim2 and ylim1<=y<=ylim2 and
zlim1<=z<=zlim2.
%Instead of ranges, assign xint,yint,zint to something other than zero to group
nodes at specific x,y,and z
%distance intervals
%
xlim1:xint:xlim2, ylim1:yint:ylim2, ylim1:yint:ylim2.
%The exclude command can be used to exclude previously defined node sets from the
current node set.
%exclude = 0
all nodes in range are included in nodeset
%exclude = m
excludes nodeset m from current nodeset
nodesetinfo={'ENDXZERO' [0 0 0] [-1000 1000 0] [-1000 1000 0] 0;
'ENDXL' [L L 0] [-1000 1000 0] [-1000 1000 0] 0;
'DISPDOF' [L L 0] [0 0 0] [0 0 0] 0;
'MID' [L/2 L/2 0] [max(node(:,3))-0.05 max(node(:,3))+0.05 0]
[max(node(:,2))/2-0.05 max(node(:,2))/2+0.05 0] 0};
%DEFINE SPRINGS
springs=[]
%DEFINE CONTACT SURFACES, NODE SURFACES, KINEMATIC CONSTRAINTS,....
surface.type={}
surface.type=[]
surface.local=[]
423
surface.coord=[]
surface.coupling={};
surface.interaction=[]
surface.contact=[]
surface.areadist=[]
%ABAQUS INP FILE NAME
jobname{count}=[SSMAnames{section_num} '_' num2str(i) '_' num2str(imptypes(j))];
%DEFINE ANALYSIS STEP

step(1).stepinfo={'STEP 1,' 'nlgeom, INC=' [260]};
step(1).solutiontype='Static, Riks';
step(1).solutionsteps={'0.25, ,1e-10, 1'};
step(1).solutioncontrols={ };
step(1).boundarycon={'ENDXZERO' 2 3;
'ENDXL' 2 3;
'MID' 1 1}
step(1).coupling=[]
step(1).loads={'*Cload' end1cload(:,1) 1 end1cload(:,2)./2;
'*Cload' end1cload(:,1)+200 1 end1cload(:,2)./2;
'*Cload' end2cload(:,1) 1 end2cload(:,2)./2;
'*Cload' end2cload(:,1)-200 1 end2cload(:,2)./2}
step(1).outrequest={'*Output, field, frequency=10';
'*Element Output';
'1,3,5';
'S,MISES';
'*Node Output';
'U';
'*Node Print, NSET=DISPDOF, SUMMARY=NO';
'U1,CF1'};
%WRITE ABAQUS INP FILE
%this is the important function, you can use this in for loops to generate
parameter studies
jhabnl(L, node, elem, nele, end1load, end2load, hole, nodesetinfo, surface,
nodeadd, step, jobname{count},matprops,imperfections,springs,sectionpoints)
count=count+1
end
end
%CREATE BEAST BATCH FILE
%Generates a linux batch file that will submit all of the parameter study
%.inp files to the queue manager on the beast.
ABQbeastscript(jobname,ones(length(jobname),1)*4,'cdmscript')
%Run the script at the beast command line with:
% bash cdmscript
424
Appendix B
ABAQUS element-based elastic buckling results
ThisappendixcontainsthefiniteelementplatemodeldimensionsandABAQUScritical
elastic buckling stress results (fcr ) used in the Chapter 3 elastic buckling studies on
l
stiffenedandunstiffenedelements.
425
426
9.2.1 Stiffenedelementinuniaxialcompression,
transverselycenteredholes
Model
number
hole type
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
circular
circular
circular
circular
circular
circular
circular
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
hhole
in.
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
h
in.
15.00
7.50
5.00
3.75
3.00
2.50
2.14
15.00
7.50
5.00
3.75
3.00
2.50
2.14
15.00
7.50
5.00
3.75
3.00
2.50
2.14
15.00
7.50
5.00
3.75
3.00
2.50
2.14
15.00
7.50
5.00
3.75
3.00
2.50
2.14
3.41
3.41
3.41
3.41
3.41
3.41
3.41
3.41
3.41
3.41
3.41
3.41
5.77
5.77
5.77
5.77
5.77
5.77
5.77
5.77
5.77
5.77
5.77
7.89
7.89
7.89
7.89
7.89
7.89
7.89
7.89
7.89
7.89
2.27
2.27
2.27
2.27
2.27
2.27
2.27
Lhole
in.
1.50
1.50
1.50
1.50
1.50
1.50
1.50
4.00
4.00
4.00
4.00
4.00
4.00
4.00
6.00
6.00
6.00
6.00
6.00
6.00
6.00
8.00
8.00
8.00
8.00
8.00
8.00
8.00
12.00
12.00
12.00
12.00
12.00
12.00
12.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
427
S
in.
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
96.00
48.00
32.00
24.00
19.20
16.00
13.71
12.00
10.67
9.60
8.73
8.00
96.00
48.00
32.00
24.00
19.20
16.00
13.71
12.00
10.67
9.60
8.73
96.00
48.00
32.00
24.00
19.20
16.00
13.71
12.00
10.67
9.60
12.00
16.00
20.00
24.00
28.00
32.00
36.00
L
in.
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
12.00
16.00
20.00
24.00
28.00
32.00
36.00
t
in.
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
hole
in.
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
fcrl
ksi
0.57
2.21
4.91
8.88
14.21
20.61
28.03
0.53
1.99
4.38
8.01
13.39
20.68
28.09
0.49
1.80
3.96
7.27
12.44
20.73
28.13
0.44
1.65
3.67
6.86
12.06
20.78
28.21
0.37
1.49
3.36
6.53
11.83
20.80
22.82
9.95
9.95
9.95
9.94
9.94
9.92
9.91
9.86
9.94
9.77
9.72
9.80
3.29
3.29
3.29
3.28
3.25
3.28
3.18
3.17
3.26
3.23
3.02
1.84
1.84
1.83
1.81
1.80
1.77
1.81
1.81
1.68
1.71
26.47
25.43
25.02
24.86
24.80
24.77
24.75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
square
square
square
square
square
square
square
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
2.27
2.27
2.27
2.27
2.27
2.27
2.27
3.41
3.41
3.41
3.41
3.41
3.41
3.41
3.41
3.41
3.41
3.41
3.41
3.41
3.41
5.77
5.77
5.77
5.77
5.77
5.77
5.77
5.77
5.77
5.77
5.77
5.77
5.77
5.77
7.89
7.89
7.89
7.89
7.89
7.89
7.89
7.89
7.89
7.89
7.89
7.89
7.89
7.89
15.00
7.50
5.00
3.75
3.00
2.50
2.14
15.00
7.50
5.00
3.75
3.00
2.50
2.14
15.00
7.50
5.00
3.75
3.00
2.50
2.14
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
1.50
1.50
1.50
1.50
1.50
1.50
1.50
40.00
44.00
48.00
60.00
72.00
84.00
96.00
12.00
16.00
20.00
24.00
28.00
32.00
36.00
40.00
44.00
48.00
60.00
72.00
84.00
96.00
12.00
16.00
20.00
24.00
28.00
32.00
36.00
40.00
44.00
48.00
60.00
72.00
84.00
96.00
12.00
16.00
20.00
24.00
28.00
32.00
36.00
40.00
44.00
48.00
60.00
72.00
84.00
96.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
428
40.00
44.00
48.00
60.00
72.00
84.00
96.00
12.00
16.00
20.00
24.00
28.00
32.00
36.00
40.00
44.00
48.00
60.00
72.00
84.00
96.00
12.00
16.00
20.00
24.00
28.00
32.00
36.00
40.00
44.00
48.00
60.00
72.00
84.00
96.00
12.00
16.00
20.00
24.00
28.00
32.00
36.00
40.00
44.00
48.00
60.00
72.00
84.00
96.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0692
0.0692
0.0692
0.0692
0.0692
0.0692
0.0692
0.1038
0.1038
0.1038
0.1038
0.1038
0.1038
0.1038
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
24.74
24.73
24.72
24.69
24.68
24.67
24.67
9.86
9.98
9.94
9.95
9.95
9.95
9.95
9.95
9.95
9.95
9.95
9.95
9.95
9.95
3.57
3.31
3.26
3.30
3.30
3.29
3.29
3.29
3.29
3.29
3.29
3.29
3.29
3.29
1.81
2.02
1.87
1.81
1.83
1.85
1.84
1.84
1.84
1.84
1.84
1.84
1.84
1.84
2.14
7.96
17.45
31.73
52.49
82.03
111.50
4.81
17.87
39.05
70.42
114.77
178.44
242.93
0.56
2.21
4.95
9.02
14.32
20.62
27.99
9.2.2 Stiffenedelementinuniaxialcompression,offsetholes
Model
number
hole type
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
hhole
Lhole
hole
fcrl
in.
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
in.
15.00
15.00
15.00
15.00
15.00
15.00
15.00
7.50
7.50
7.50
7.50
7.50
7.50
5.00
5.00
5.00
5.00
5.00
3.75
3.75
3.75
3.75
3.75
3.75
3.75
3.00
3.00
3.00
3.00
3.00
3.00
3.00
2.50
2.50
2.50
2.50
2.50
2.50
2.50
2.14
2.14
2.14
2.14
2.14
2.14
2.14
in.
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
in.
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
in.
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
in.
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
in.
0.00
0.94
1.88
2.81
3.75
4.69
5.63
0.00
0.47
0.94
1.41
1.88
2.34
0.00
0.31
0.63
0.94
1.25
0.07
0.14
0.20
0.27
0.34
0.41
0.47
0.05
0.09
0.14
0.18
0.23
0.27
0.32
0.03
0.06
0.09
0.12
0.15
0.18
0.21
0.02
0.04
0.06
0.08
0.10
0.12
0.14
ksi
0.53
0.54
0.54
0.54
0.54
0.54
0.54
1.99
1.98
1.96
1.96
1.91
1.85
4.38
4.17
3.79
3.71
3.38
7.81
7.43
7.03
6.65
6.31
5.99
5.70
12.65
11.76
10.95
10.23
9.59
9.01
8.49
20.12
18.55
17.15
15.91
14.80
13.81
12.91
28.54
28.54
27.99
25.88
23.98
22.26
20.72
429
9.2.3 Stiffenedelementinbending(Y=0.50h),transversely
centeredholes
Model
number
hole type
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
hhole
in.
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
h
in.
15.00
15.00
15.00
15.00
7.50
7.50
7.50
7.50
5.00
5.00
5.00
5.00
3.75
3.75
3.75
3.75
3.00
3.00
3.00
3.00
2.50
2.50
2.50
2.50
2.14
2.14
2.14
2.14
Lhole
in.
4.00
6.00
8.00
12.00
4.00
6.00
8.00
12.00
4.00
6.00
8.00
12.00
4.00
6.00
8.00
12.00
4.00
6.00
8.00
12.00
4.00
6.00
8.00
12.00
4.00
6.00
8.00
12.00
S
in.
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
430
L
in.
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
t
in.
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
hole
in.
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
Y
in.
7.50
7.50
7.50
7.50
3.75
3.75
3.75
3.75
2.50
2.50
2.50
2.50
1.88
1.88
1.88
1.88
1.50
1.50
1.50
1.50
1.25
1.25
1.25
1.25
1.07
1.07
1.07
1.07
fcrl
ksi
3.26
3.06
2.78
2.23
10.63
8.27
6.77
5.28
17.97
13.89
11.97
10.40
27.16
22.42
20.51
19.11
41.14
36.47
34.85
33.81
64.42
60.59
59.52
59.07
106.30
104.39
104.25
58.03
9.2.4 Stiffenedelementinbending(Y=0.50h),offsetholes
Model
number
hole type
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
hhole
in.
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
h
in.
15.00
15.00
15.00
15.00
15.00
15.00
15.00
7.50
7.50
7.50
7.50
7.50
7.50
5.00
5.00
5.00
5.00
5.00
3.75
3.75
3.75
3.75
3.75
3.75
3.75
3.00
3.00
3.00
3.00
3.00
3.00
3.00
2.50
2.50
2.50
2.50
2.50
2.50
2.50
2.14
2.14
2.14
2.14
2.14
2.14
2.14
15.00
15.00
15.00
15.00
15.00
15.00
15.00
7.50
7.50
7.50
7.50
7.50
7.50
5.00
5.00
5.00
5.00
5.00
3.75
3.75
3.75
3.75
3.75
3.75
3.75
3.00
3.00
3.00
3.00
3.00
3.00
3.00
2.50
2.50
2.50
2.50
2.50
2.50
2.50
2.14
2.14
2.14
2.14
2.14
2.14
2.14
Lhole
in.
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
S
in.
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
431
L
in.
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
t
in.
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
hole
in.
0.00
-0.94
-1.88
-2.81
-3.75
-4.69
-5.63
0.00
-0.47
-0.94
-1.41
-1.88
-2.34
0.00
-0.31
-0.63
-0.94
-1.25
-0.07
-0.14
-0.20
-0.27
-0.34
-0.41
-0.47
-0.05
-0.09
-0.14
-0.18
-0.23
-0.27
-0.32
-0.03
-0.06
-0.09
-0.12
-0.15
-0.18
-0.21
-0.02
-0.04
-0.06
-0.08
-0.10
-0.12
-0.14
0.00
0.94
1.88
2.81
3.75
4.69
5.63
0.00
0.47
0.94
1.41
1.88
2.34
0.00
0.31
0.63
0.94
1.25
0.07
0.14
0.20
0.27
0.34
0.41
0.47
0.05
0.09
0.14
0.18
0.23
0.27
0.32
0.03
0.06
0.09
0.12
0.15
0.18
0.21
0.02
0.04
0.06
0.08
0.10
0.12
0.14
Y
in.
7.50
7.50
7.50
7.50
7.50
7.50
7.50
3.75
3.75
3.75
3.75
3.75
3.75
2.50
2.50
2.50
2.50
2.50
1.88
1.88
1.88
1.88
1.88
1.88
1.88
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.25
1.25
1.25
1.25
1.25
1.25
1.25
1.07
1.07
1.07
1.07
1.07
1.07
1.07
7.50
7.50
7.50
7.50
7.50
7.50
7.50
3.75
3.75
3.75
3.75
3.75
3.75
2.50
2.50
2.50
2.50
2.50
1.88
1.88
1.88
1.88
1.88
1.88
1.88
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.25
1.25
1.25
1.25
1.25
1.25
1.25
1.07
1.07
1.07
1.07
1.07
1.07
1.07
fcrl
ksi
3.26
3.34
3.38
3.41
3.44
3.45
3.44
10.63
12.10
13.35
13.80
13.97
13.92
17.96
21.25
27.86
31.03
31.18
27.74
28.67
30.05
32.01
34.74
38.57
43.91
41.09
41.45
42.28
43.66
45.76
48.87
53.49
63.17
62.48
62.30
62.68
63.73
65.62
68.69
102.33
99.94
98.12
96.92
96.40
96.69
97.99
3.26
3.14
3.04
2.96
2.91
2.81
2.57
10.63
9.72
9.53
10.17
10.87
7.36
17.96
16.95
17.88
21.78
19.29
26.90
26.93
27.25
27.88
28.87
30.28
32.22
41.54
42.34
43.55
45.23
47.44
50.31
54.02
66.02
68.21
70.99
74.42
78.63
83.69
89.64
108.66
112.46
116.44
116.35
112.54
107.24
100.88
9.2.5 Stiffenedelementinbending(Y=0.75h),offsetholes
Model
number
hole type
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
hhole
in.
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
h
in.
15.00
15.00
15.00
15.00
15.00
15.00
15.00
7.50
7.50
7.50
7.50
7.50
7.50
5.00
5.00
5.00
5.00
5.00
3.75
3.75
3.75
3.75
3.75
3.75
3.75
3.00
3.00
3.00
3.00
3.00
3.00
3.00
2.50
2.50
2.50
2.50
2.50
2.50
2.50
2.14
2.14
2.14
2.14
2.14
2.14
2.14
15.00
15.00
15.00
15.00
15.00
15.00
15.00
7.50
7.50
7.50
7.50
7.50
7.50
5.00
5.00
5.00
5.00
5.00
3.75
3.75
3.75
3.75
3.75
3.75
3.75
3.00
3.00
3.00
3.00
3.00
3.00
3.00
2.50
2.50
2.50
2.50
2.50
2.50
2.50
2.14
2.14
2.14
2.14
2.14
2.14
2.14
Lhole
in.
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
S
in.
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
432
L
in.
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
t
in.
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
hole
in.
0.00
-0.94
-1.88
-2.81
-3.75
-4.69
-5.63
0.00
-0.47
-0.94
-1.41
-1.88
-2.34
0.00
-0.31
-0.63
-0.94
-1.25
-0.07
-0.14
-0.20
-0.27
-0.34
-0.41
-0.47
-0.05
-0.09
-0.14
-0.18
-0.23
-0.27
-0.32
-0.03
-0.06
-0.09
-0.12
-0.15
-0.18
-0.21
-0.02
-0.04
-0.06
-0.08
-0.10
-0.12
-0.14
0.00
0.94
1.88
2.81
3.75
4.69
5.63
0.00
0.47
0.94
1.41
1.88
2.34
0.00
0.31
0.63
0.94
1.25
0.07
0.14
0.20
0.27
0.34
0.41
0.47
0.05
0.09
0.14
0.18
0.23
0.27
0.32
0.03
0.06
0.09
0.12
0.15
0.18
0.21
0.02
0.04
0.06
0.08
0.10
0.12
0.14
Y
in.
11.25
11.25
11.25
11.25
11.25
11.25
11.25
5.63
5.63
5.63
5.63
5.63
5.63
3.75
3.75
3.75
3.75
3.75
2.81
2.81
2.81
2.81
2.81
2.81
2.81
2.25
2.25
2.25
2.25
2.25
2.25
2.25
1.88
1.88
1.88
1.88
1.88
1.88
1.88
1.61
1.61
1.61
1.61
1.61
1.61
1.61
11.25
11.25
11.25
11.25
11.25
11.25
11.25
5.63
5.63
5.63
5.63
5.63
5.63
3.75
3.75
3.75
3.75
3.75
2.81
2.81
2.81
2.81
2.81
2.81
2.81
2.25
2.25
2.25
2.25
2.25
2.25
2.25
1.88
1.88
1.88
1.88
1.88
1.88
1.88
1.61
1.61
1.61
1.61
1.61
1.61
1.61
fcrl
ksi
1.48
1.49
1.50
1.51
1.53
1.54
1.56
5.20
5.22
5.33
5.52
5.77
6.04
9.97
9.62
9.66
10.09
11.07
15.53
15.14
14.83
14.59
14.42
14.30
14.24
23.86
23.01
22.29
21.66
21.13
20.68
20.32
37.44
35.83
34.40
33.14
32.01
31.01
30.11
61.27
58.47
55.92
53.62
51.55
49.58
47.81
1.48
1.47
1.45
1.44
1.43
1.40
1.36
5.19
5.25
5.31
5.17
4.62
3.86
9.97
10.79
11.83
10.02
7.24
16.58
17.27
18.11
19.13
20.35
21.75
21.79
25.99
27.34
28.92
30.82
33.09
35.84
38.21
41.35
43.75
46.52
49.74
53.51
56.68
56.66
67.67
71.25
74.96
76.63
75.11
72.38
68.78
9.2.6 Unstiffenedelementinuniaxialcompression,
transverselycenteredholes
Model
number
hole type
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
ang. slotted
ang. slotted
ang. slotted
ang. slotted
ang. slotted
ang. slotted
ang. slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
square
square
square
square
square
square
square
circular
circular
circular
circular
circular
circular
circular
hhole
in.
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
h
in.
15.00
7.50
5.00
3.75
3.00
2.50
2.14
15.00
7.50
5.00
3.75
3.00
2.50
2.14
15.00
7.50
5.00
3.75
3.00
2.50
2.14
15.00
7.50
5.00
3.75
3.00
2.50
2.14
15.00
7.50
5.00
3.75
3.00
2.50
2.14
15.00
7.50
5.00
3.75
3.00
2.50
2.14
15.00
7.50
5.00
3.75
3.00
2.50
2.14
7.89
7.89
7.89
7.89
7.89
7.89
7.89
5.77
5.77
5.77
5.77
5.77
5.77
5.77
3.41
3.41
3.41
3.41
3.41
3.41
3.41
2.27
2.27
2.27
2.27
2.27
2.27
2.27
15.00
7.50
5.00
3.75
3.00
2.50
2.14
15.00
7.50
5.00
3.75
3.00
2.50
2.14
Lhole
in.
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
6.00
6.00
6.00
6.00
6.00
6.00
6.00
8.00
8.00
8.00
8.00
8.00
8.00
8.00
12.00
12.00
12.00
12.00
12.00
12.00
12.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
433
S
in.
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
96.00
48.00
32.00
24.00
16.00
12.00
8.00
96.00
48.00
32.00
24.00
16.00
12.00
8.00
96.00
48.00
32.00
24.00
16.00
12.00
8.00
96.00
48.00
32.00
24.00
16.00
12.00
8.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
L
in.
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
96.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
t
in.
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0692
0.0692
0.0692
0.0692
0.0692
0.0692
0.0692
0.1038
0.1038
0.1038
0.1038
0.1038
0.1038
0.1038
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
hole
in.
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
fcrl
ksi
0.06
0.23
0.48
0.76
0.98
1.08
1.03
0.06
0.23
0.48
0.76
0.98
1.08
1.03
0.06
0.22
0.45
0.67
0.81
0.85
0.77
0.06
0.21
0.43
0.61
0.63
0.58
0.49
0.06
0.20
0.31
0.34
0.32
0.28
0.23
0.25
0.91
1.93
3.04
3.88
4.22
3.96
0.55
2.05
4.33
6.79
8.59
9.23
8.52
0.22
0.22
0.22
0.21
0.21
0.20
0.19
0.40
0.38
0.38
0.38
0.37
0.36
0.34
0.91
0.87
0.87
0.87
0.86
0.84
0.78
1.16
1.08
1.08
1.08
1.08
1.08
1.06
0.06
0.24
0.52
0.88
1.26
1.57
1.71
0.06
0.24
0.52
0.91
1.34
1.77
2.11
9.2.7 Unstiffenedelementinuniaxialcompression,offset
holes
Model
number
hole type
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
slotted
hhole
Lhole
hole
fcrl
in.
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
1.50
in.
15.00
15.00
15.00
15.00
15.00
15.00
15.00
7.50
7.50
7.50
7.50
7.50
7.50
5.00
5.00
5.00
5.00
5.00
3.75
3.75
3.75
3.75
3.75
3.75
3.75
3.00
3.00
3.00
3.00
3.00
3.00
3.00
2.50
2.50
2.50
2.50
2.50
2.50
2.50
2.14
2.14
2.14
2.14
2.14
2.14
2.14
15.00
15.00
15.00
15.00
15.00
15.00
15.00
7.50
7.50
7.50
7.50
7.50
7.50
5.00
5.00
5.00
5.00
5.00
3.75
3.75
3.75
3.75
3.75
3.75
3.75
3.00
3.00
3.00
3.00
3.00
3.00
3.00
2.50
2.50
2.50
2.50
2.50
2.50
2.50
2.14
2.14
2.14
2.14
2.14
2.14
2.14
in.
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
in.
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
in.
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
in.
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
0.0346
in.
0.00
-0.94
-1.88
-2.81
-3.75
-4.69
-5.63
0.00
-0.47
-0.94
-1.41
-1.88
-2.34
0.00
-0.31
-0.63
-0.94
-1.25
-0.07
-0.14
-0.20
-0.27
-0.34
-0.41
-0.47
-0.05
-0.09
-0.14
-0.18
-0.23
-0.27
-0.32
-0.03
-0.06
-0.09
-0.12
-0.15
-0.18
-0.21
-0.02
-0.04
-0.06
-0.08
-0.10
-0.12
-0.14
0.00
0.94
1.88
2.81
3.75
4.69
5.63
0.00
0.47
0.94
1.41
1.88
2.34
0.00
0.31
0.63
0.94
1.25
0.07
0.14
0.20
0.27
0.34
0.41
0.47
0.05
0.09
0.14
0.18
0.23
0.27
0.32
0.03
0.06
0.09
0.12
0.15
0.18
0.21
0.02
0.04
0.06
0.08
0.10
0.12
0.14
ksi
0.06
0.06
0.06
0.06
0.06
0.06
0.06
0.23
0.23
0.24
0.24
0.24
0.25
0.49
0.50
0.51
0.53
0.55
0.80
0.81
0.82
0.83
0.84
0.86
0.87
1.05
1.07
1.09
1.12
1.14
1.17
1.20
1.18
1.21
1.24
1.27
1.31
1.35
1.39
1.16
1.19
1.22
1.26
1.29
1.34
1.38
0.06
0.06
0.06
0.06
0.06
0.06
0.06
0.23
0.23
0.23
0.22
0.22
0.22
0.49
0.48
0.47
0.46
0.45
0.77
0.76
0.75
0.75
0.74
0.73
0.72
1.01
0.99
0.97
0.96
0.94
0.92
0.91
1.13
1.10
1.08
1.05
1.03
1.01
0.99
1.10
1.07
1.05
1.02
0.99
0.97
0.95
434
Appendix C
Derivation of elastic buckling coefficients for
unstiffened elements
9.2.8
C.1kAforanunstiffenedelementincompression
The finite strip method is employed with CUFSM (Schafer and dny 2006) to
calculate the plate buckling coefficient for an unstiffened strip in compression, kA, as a
function of unstiffened strip aspect ratio (Lhole/hA) and the compressive stress ratio (A).
TheunstiffenedelementmodelsetupinCUFSMisprovidedinFigureC.1.Theresults
from the CUFSM parameter study, where A is varied from 0 to 1, are presented in
FigureC.2.
Lhole
Simply
supported
f1
A =
hA
f2
f1
f2
free
Unstiffened Element
Section A-A
FigureC.1CUFSMfinitestripmodelingdefinitionforanunstiffenedelementincompression
ThefminsearchfunctioninMatlab(Mathworks2007)isusedtodeterminethevariables
1through5inthegeneralequationform:
435
kA =
0.578
+
A + 0.34
1 2 A
L
3 A + 4 + hole
hA
ThevariablesarechosentominimizethesumofthesquarederrorsbetweentheCUFSM
resultsinFigureC.2andthefittedcurve.Thecurvefittingresultsintheequation:
kA =
0.578
+
A + 0.34
2.70 1.76 A
L
0.024 A + 0.035 + hole
hA
10
9
plate buckling coeff., kA
8
7
6
5
4
3
2
1
0
5
Lhole/hA
10
FigureC.2TheplatebucklingcoefficientkAforanunstiffenedelementincompression(themultiplecurves
represent0A1withastepof0.1,11curvestotal)
The mean and standard deviation of the ABAQUS to predicted ratio when
0.1Lhole/yA2 is 1.14 and 0.61 respectively. As shown in Figure C.3, the accuracy of the
predictionisoftenconservativewithinthisaspectratiorange.For2<Lhole/yA10themean
andstandarddeviationoftheABAQUStopredictedratioare0.99and0.02respectively.
436
10
9
plate buckling coeff., kA
8
7
6
Fitted curve is
conservative predictor
of kA
5
4
3
2
1
0
0.2
0.4
0.6
0.8
1
1.2
Lhole/hA
1.4
1.6
1.8
FigureC.3ThefittedcurveforkAisaconservativepredictorwhenLhole/yA2
9.2.9 C.2kBforanunstiffenedelementwith
compressiononthefreeedge
The finite strip method is employed with CUFSM (Schafer and dny 2006) to
calculatetheplatebucklingcoefficientforanunstiffenedelementwithcompressionon
thefreeedgeandtensiononthesimplysupportededge,kB,asafunctionofunstiffened
stripaspectratio(Lhole/hB)andthecompressivestressratio(B).Theunstiffenedelement
modelsetupinCUFSMisprovidedinFigureC.4.
437
Lhole
A
f1
B =
free
f2
f1
hA
f2
A
Simply
supported
Unstiffened Element
Section A-A
FigureC.4CUFSMfinitestripmodelingdefinitionforanunstiffenedelementwithcompressiononthefree
edge,tensiononthesimplysupportededge
TheresultsfromtheCUFSMparameterstudy,where Bisvariedfrom0to10,are
presentedinFigureC.5C.5.Astheportionoftheplatethatisintensionincreases(i.e.,
Bincreases),thebucklingmodeswitchesfromonebuckledhalfwavetomultiplehalf
waves.
40
35
plate buckling coeff., kB
30
25
Buckles in several halfwavelength (=10 shown)
20
15
Buckles in one halfwavelength (=0 shown)
10
5
0
10
Lhole/hB
15
FigureC.5VariationinplatebucklingcoefficientkBforanunstiffenedelementwithBrangingfrom0to10
ApolynomialcurveisfittotheminimumkBasshowninFigureC.6:
k B = 0.340 B2 + 0.100 B + 0.573 .
438
ThemeanandstandarddeviationoftheCUFSMminimatofittedcurvepredictionratio
are1.03and0.11respectively.
40
CUFSM
Fitted curve
35
plate buckling coeff., kB
30
25
20
15
10
5
0
10
FigureC.6CurvefittominimumkBforBrangingfrom0to10
ThisapproximationisaccuratewhenLhole/yB>5butdoesnotcapturetheboostinkBwhen
Lhole/hBissmall.SinceLhole/hBmayoftenbelessthan1whenconsideringcommonplateand
hole sizes, it is important to have a viable estimate of kB to avoid overly conservative
predictions. A family of curves is fit to the CUFSM predictions for the case when
0.25Lhole/yB2andwhereBisvariedfrom0to10,resultinginthefollowingequation:
2
h
0.38 B + 1.6 B + 0.49
L
Lhole
, 0 hole 2, 0 B 10
kB =
0.1
hB
h
0.3
0.20 B + B + 0.14
Lhole
1.8
439
The equation provides an accurate representation of kB as demonstrated in Figure C.7.

The mean and standard deviation of the CUFSM to predicted ratio are 0.01 and 0.03
respectively.
60
50
kB
40
30
20
10
0.5
1.5
Lhole/hB
2.5
FigureC.7FamilyofcurvesusedtosimulateboostinkBwhenLhole/hB2,Brangesfrom0to10
440
Appendix D
Elastic buckling prediction method of crosssectional elements with holes
B1 Critical Elastic Buckling Stress of Elements with Perforations

B1.1 Uniformly Compressed Stiffened Element
For uniformly compressed stiffened elements with uniformly spaced perforations
satisfying the limits
S 1.5 and S 2 ,
h
L hole
the critical elastic buckling stress, fcrl, is
f crl = min[f cr , f crh ] .
(Eq. B1.1-1)
The critical elastic buckling stress, fcr, without the influence of perforations is
2
f cr = k
2E t ,

12(1 2 ) h
(Eq. B1.1-2)
where k=4 for a stiffened element with L/h>4.

The critical elastic buckling stress, fcrh, with the influence of perforations is
f crh = f crh ,net (1 h hole h ) ,
(Eq. B1.1-3)
where the critical elastic buckling stress, fcrh,net, at the location of a perforation is
f crh , net = min[f crA , f crB ] .
(Eq. B1.1-4)
The critical elastic buckling stress, fcri,of unstiffened strip i is

2
f cri
2 E t and i = A or B,

= ki
12(1 2 ) h i
(Eq. B1.1-5)
where
L hole h i 1 , k i = 0.425 +
(L hole
0.2
hi )
0.95
(Eq. B1.1-6)
0.6
441
L hole h i < 1 , k i = 0.925 , and i = A or B.
(Eq. B1.1-7)
B1.2 Stiffened Element Under Stress Gradient
For stiffened elements under a stress gradient with uniformly spaced perforations
S 1.5 and S 2 ,
h
L hole
(Eq. B1.2-1)
2
f cr = k
2E t ,

12(1 2 ) h
(Eq. B1.2-2)
where
k = 4 + 2(1 + ) + 2(1 + )
3
(Eq. B1.2-3)
and
= f 2 f1 = (h Y ) Y .
(Eq. B1.2-4)
h A , and
for hA+hhole Y, f = f
crh
crh ,net (1 + A )
Y
(Eq. B1.2-5)
h hole
h
for hA+hhole < Y, f = f
2 A hole ,
crh
crh , net 1
(Eq. B1.2-6)
where
A =
Y hA .
Y
Eq. (B1.2-7)
The critical elastic buckling stress, fcrh,net, at the location of a perforation is
f crh ,net = min[f crA , f crB ]
(Eq. B1.2-8)
Consideration of unstiffened strip A buckling is required only if hA<Y,
f crA
2E t
= kA
12(1 2 ) h A
(Eq. B1.2-9)
where
442
kA =
0.578
A + 0.34
2.70 1.76 A
2
0.024 A + 0.035 + (L hole h A )
(Eq. B1.2-10)
Consideration of unstiffened strip B buckling is required only if hA+hhole<Y,
f crB = k B
2E t
12(1 2 ) h B
Y h A h hole
(Eq. B1.2-12)
where
for Lhole/hB>2
k B = 0.340 B2 + 0.100 B + 0.573 ,
(Eq. B1.2-12)
for Lhole/hB2
2
h
+ 1.6 B + 0.49
L hole
kB =
0.1
h
0.3
0.20 B + B + 0.14
L hole
0.38 B
1.8
(Eq. B1.2-13)
and
B =
h Y
, 0 B 10 .
Y hA h hole
(Eq. B1.2-14)
443
B1.3 Unstiffened Element Under Uniform Compression with Perforations
For uniformly compressed unstiffened elements with uniformly spaced perforations

h hole
S
L hole
L
0.50 , and
2
10 , hole 10 ,
h
L hole
hA
hB
(Eq. B1.3-1)
(Eq. B1.3-2)
2
f cr = k
2E t ,

12 (1 2 ) h
(Eq. B1.3-3)
where k=0.425 for unstiffened elements with L/h>4.

2
2E t
h
, f crA 1 hole ,
f crh = min k
2
h
12 1 h
(Eq. B1.3-4)
where
L
k = 0.4251 0.062 hole
hA
(Eq. B1.3-5)
and fcrA is calculated with Eq. B1.1-5.
444
Appendix E
Derivation of global critical elastic buckling load for
a column with holes
Thisderivationdevelopstheequationfortheflexuralcriticalelasticbucklingloadof
a column with two holes spaced symmetrically about the longitudinal midline of a
column.Theconclusionsofthisderivationareusedasthefoundationfortheweighted
properties approach for approximating Pcre for columns and beams with holes as
describedinSection4.2.7.3.1.1.INHisthemomentofinertiaofthecolumncrosssection
awayfromtheholeandIHisthemomentofinertiaatthehole.
INH
IH
INH
l1
v
l2
l3
l4
L
IH
INH
FigureE.1Longcolumnwithtwoholesspacedsymmetricallyaboutthelongitudinalmidline.
A conservation of energy approach is employed in this derivation, and specifically
the RayleighRitz Method. The derivation is founded on the fundamental principle

relatingthestrainenergyandpotentialenergyofthecolumn,UandWrespectively:
445
= (U W ) = 0
where the column strain energy U is:
2
d 2v
1
U = EI 2 dx
2
dx
and the column potential energy is
2
1 dv
W = P dx .
2 dx
Applying the RaleighRitz method, we assume a shape function in the deformed
(buckled)configurationofthecolumn:
v (x ) = sin
x
L
The derivatives of this function are calculated:

dv
x
= cos
dx
L
L
2
2
d v
x
= 2 sin
2
dx
L
L
andthensubstitutedintotheequationsforUandWwhicharewrittenalongthelength
ofthecolumnas:
EI
EI H l 2 2 4
x
2 x
+
sin
dx
sin 2 dx + NH
4
0 L4
l
2 1 L
2
L
L
EI l 4 2 4
EI H L 2 4
x
2 x
sin
dx
sin 2 dx
+ H
+
4
4
2 l3 L
2 l4 L
L
L
U=
EI NH
2
l1
2 4
and
W=
1 L 2 2
P 2 2
2 x
=
P
cos
dx
.
2 0 L2
4L
L
446
l3
l2
2 4
4
x
sin 2 dx
L
ThederivativeofUistakenwithrespecttothearbitraryshapefunctionamplitude:
l3
L
dU EI NH 4 l1 2 x
x
x
sin dx + sin 2 dx + sin 2 dx
=
4
l2
l4
d
L
L
L
L
0
l4
EI H 4 l 2 2 x
x
+
+ sin dx + sin 2 dx
4
l
l
3
L
L
L
1
The definite integrals inside the derivative are then solved resulting in:
dU EI NH 4
=
d
L4
+
l 1 l 3 l 2 L l 4 L
2l 3
2l 1
2l 2
2l 4
2 + 2 2 + 2 2 + 4 sin L sin L + sin L + sin L
2l 3
EI H 4 LH
2l 2
2l 1
2l 4
L
+
+ sin
sin
+ sin
sin
4
L
L
L
L
L
2 4
Thelengthtermsfromtheintegrationsumtothelengthofcolumnwithoutahole,LNH,
and the length of column with a hole, LH. When the holes are symmetric about the
longitudinalmidlineofthecolumn,thetrigonometrictermscancelasshowninFigure
E.2andtheequationabovesimplifiesto:
dU EI NH 4 LNH EI NH 4 LH
=
+
2
2
d
L4
L4
1.5
H trig term
NH trig term
sin(2x/L)
0.5
-0.5
-1
-1.5
Trig terms cancel for holes symmetric

about longitudinal midline
0
0.1
0.2
0.3
0.4
0.5
x/L
0.6
0.7
0.8
0.9
FigureE.2Trigonometrictermsinenergysolutioncancelwhenholesaresymmetricaboutlongitudinal
midlineofcolumn.
447
Thesolutionforthepotentialenergyofthecolumnisnotdependentuponthemoment
ofinertiaandthereforethederivativecanbesolveddirectlyas:
dW P 2
=
d
2L
TheloadPthatminimizesthevariationinenergyisthecriticalelasticbucklingload,Pcre:
= (U W ) =
dU dW
= 0
d d
Equatingthevariationalenergyterms:
EI NH 4 LNH EI H 4 LH P 2
+
=0
L4
L4
2
2
2L
resultsinasolutionforPcrewherethemomentofinertiaisaweightedaverageofthenet
andgrosscrosssectionofthecolumn.
Pcre =
2 E I NH LNH + I H LH
L2
Pcreforacolumnwiththegeneralcaseofi=1..nholescanbeapproximatedas:
Pcre =
2 E I NH (LNH + TNH ) + I H (LH + TH )
L2
where
TNH =
L
2
2Lc ,i Lhole ,i
L
sin
, TH =
cos
2
i =1
L L
n
2Lc ,i Lhole ,i
sin
L L
cos
i =1
Lc,iisthedistancefromtheendofthecolumntothecenterlineofholeiandLhole,iisthe
lengthofholei.
448
Appendix F
Column experiment results
449
ColumnSpecimen362124NH
362-1-24-NH
362-1-24-NH
Flange displacement (inches)
-14
-12
-10
-8
-6
Ptest=10.48 kips
-4
-2
0
0.05
0.1
0.15
0.2
1.5
W
0.5
0
-0.5
-1
-1.5
0.05
0.1
0.15
0.2
Notes:
Loaded N to S instead of S to N. Adjusted all geometry measurements.

Lips rotated and not touching bottom platen after peak load.
450
362-2-24-NH
362-2-24-NH
-14
-12
-10
-8
-6
Ptest=10.51 kips
-4
-2
0
0.05
0.1
0.15
0.2
Notes:
451
1.5
1
0.5
0
W
-0.5
-1
-1.5
0.05
0.1
0.15
0.2
362-3-24-NH
362-3-24-NH
-14
-12
-10
-8
-6
Ptest=10.15 kips
-4
-2
0
0.05
0.1
0.15
0.2
1.5
W
0.5
0
-0.5
-1
-1.5
0.05
0.1
0.15
0.2
Notes:
Bottom lips rotated at 7 kips post-peak and are not bearing on platen.
452
ColumnSpecimen362124H
362-1-24-H
362-1-24-H
-14
-12
-10
-8
-6
Ptest=10.01 kips
-4
-2
0
0.05
0.1
0.15
0.2
1.5
1
0.5
-0.5
-1
-1.5
Visible buckling of web on either side of hole at 7 kips.
453
Notes:
0.05
0.1
0.15
0.2
Local buckling at hole
(unstiffened strip)
(a) P=0 kips
(b) P=10.4 kips

(peak load)
(c) P=7.0 kips
362-2-24-H
362-2-24-H
-14
-12
-10
-8
-6
Ptest=10.38 kips
-4
-2
0
0.05
0.1
0.15
0.2
1.5
1
0.5
0
-0.5
-1
-1.5
Notes:
Visible buckling of web on either side of hole at 5 kips.
454
0.05
0.1
0.15
0.2
362-3-24-H
362-3-24-H
-14
-12
-10
-8
-6
Ptest=9.94 kips
-4
-2
0
0.05
0.1
0.15
0.2
1.5
1
0.5
0
W
-0.5
-1
-1.5
0.05
0.1
0.15
0.2
Notes:
Accidental preload to 3 kips when adjusting specimen for test.
455
362-1-48-NH
362-1-48-NH
-14
-12
-10
-8
-6
Ptest=9.09 kips
-4
-2
0
0.05
0.1
0.15
0.2
1.5
W
0.5
0
-0.5
-1
-1.5
0.05
0.1
0.15
0.2
Notes:
Good end conditions no visible gaps.

9 kips a metallic noise yielding of west flange and increase in local wavelengths.
Column failed by global-torsional collapse.
456
Peak Load
362-2-48-NH
362-2-48-NH
-14
-12
-10
-8
-6
Ptest=9.49 kips
-4
-2
0
0.05
0.1
0.15
0.2
Notes:
Tight end conditions at 1.5 kips.

Local buckling at 6.5 kips .
No sounds for this test.
457
1.5
W
0.5
0
-0.5
-1
-1.5
0.05
0.1
0.15
0.2
Peak Load
362-3-48-NH
362-3-48-NH
-14
-12
-10
-8
-6
Ptest=9.48 kips
-4
-2
0
0.05
0.1
0.15
0.2
Notes:
Local buckling first observed at 6.5 kips.

Local wavelengths lengthen at 8.5 kips.
Yielding of flange lips at 9 kips (near peak).
458
1.5
W
0.5
0
-0.5
-1
-1.5
0.05
0.1
0.15
0.2
Peak Load
362-1-48-H
362-1-48-H
-14
-12
-10
-8
-6
Ptest=8.95 kips
-4
-2
0
0.05
0.1
0.15
0.2
Notes:
No visible gaps and ends under 1 kip.

Local buckling is visible at 7 kips.
Local half-waves merge at 8.5 kips.
Bulging of web at hole occurs near peak load.
459
1.5
W
0.5
0
-0.5
-1
-1.5
0.05
0.1
0.15
0.2
Peak Load
362-2-48-H
362-2-48-H
-14
-12
-10
-8
-6
Ptest=9.18 kips
-4
-2
0
0.05
0.1
0.15
0.2
1.5
1
0.5
0
-0.5
-1
-1.5
W
E
0.05
0.1
0.15
0.2
Notes:
End conditions tight at 4 kips.

Local buckling visible at 6.5 kips.
Distortional buckling seems to increase as load-displacement softens.
East LVDT reaches limit of range as column starts to twist.
460
Peak Load
362-3-48-H
362-3-48-H
-14
-12
-10
-8
-6
Ptest=9.37 kips
-4
-2
0
0.05
0.1
0.15
0.2
Notes:
Local buckling visible at 6.5 kips.

461
1.5
W
0.5
0
-0.5
-1
-1.5
0.05
0.1
0.15
0.2
Peak Load
600-1-24-NH
600-1-24-NH
-14
-12
-10
-8
-6
Ptest=11.93 kips
-4
-2
0
0.05
0.1
0.15
0.2
Notes:
Local and distortional waves seem to stay separate.

8 kips (post-peak) east flange buckles.
462
1.5
1
0.5
0
W
-0.5
-1
-1.5
0.05
0.1
0.15
0.2
Peak Load
600-2-24-NH
600-2-24-NH
-14
-12
-10
-8
-6
Ptest=11.95 kips
-4
-2
0
0.05
0.1
0.15
0.2
1.5
W
0.5
0
-0.5
-1
-1.5
0.05
0.1
0.15
0.2
Notes:
Web has large curve when placing specimen on bottom platen.

Visible gap between platen and specimen at top west web-flange corner - 5 kips.
10 kips (post-peak)- flanges buckle and lose contact with bottom platen.
463
Peak Load
600-3-24-NH
600-3-24-NH
-14
-12
-10
-8
-6
Ptest=12.24 kips
-4
-2
0
0.05
0.1
0.15
0.2
1.5
W
0.5
0
-0.5
-1
-1.5
0.05
0.1
0.15
0.2
Notes:
Specimen failed at bottom end condition, web rolled over and was not bearing on platen.
464
Peak Load
600-1-24-H
600-1-24-H
-14
-12
-10
-8
-6
Ptest=12.14 kips
-4
-2
0
0.05
0.1
0.15
0.2
1.5
1
0.5
0
W
-0.5
-1
-1.5
0.05
0.1
0.15
0.2
Notes:
Slight gap at east top web-flange corner - 3 kips, gap is closed at 11 kips.
East flange gives way at 11 kips with dip in load-disp. curve, may be related to above.
Loud popping sound at 8 kips (post-peak) and large change in load-displ. slope.
465
Peak Load
600-2-24-H
600-2-24-H
-14
-12
-10
-8
-6
Ptest=11.62 kips
-4
-2
0
0.05
0.1
0.15
0.2
1.5
W
0.5
0
-0.5
-1
-1.5
0.05
0.1
0.15
0.2
Notes:
Specimen failure mode similar to that of a no-hole specimen.
466
Peak Load
600-3-24-H
600-3-24-H
-14
-12
-10
-8
-6
Ptest=11.79 kips
-4
-2
0
0.05
0.1
0.15
0.2
Notes:
Good contact with platens.
467
1.5
W
0.5
0
-0.5
-1
-1.5
0.05
0.1
0.15
0.2
Peak Load
600-1-48-NH
600-1-48-NH
-14
-12
-10
-8
-6
Ptest=11.15 kips
-4
-2
0
0.05
0.1
0.15
0.2
1.5
W
0.5
0
-0.5
-1
-1.5
0.05
0.1
0.15
0.2
Notes:
Local buckling first observed at 4.5 kips (11 half-waves).

Distortional wave becomes prominent at 10 kips.
Loud noises 1 minute apart L waves turn to D waves at north, then south ends.
468
Peak Load
600-2-48-NH
600-2-48-NH
-14
-12
-10
-8
-6
Ptest=11.44 kips
-4
-2
0
0.05
0.1
0.15
0.2
1.5
W
0.5
0
-0.5
-1
-1.5
0.05
0.1
0.15
0.2
Notes:
Gap between platen and specimen at top east flange-web corner closes at 2 kips.
Can see distortional shape developing at 4.5 kips.
Local buckling visible at 5 kips.
Two loud bangs (peak load, 10.5 kips post peak) local web waves change to D waves.
Flange distortion slows at 7 kips post-peak.
469
Peak Load
600-3-48-NH
600-3-48-NH
-14
-12
-10
-8
-6
Ptest=11.29 kips
-4
-2
0
0.05
0.1
0.15
0.2
1.5
W
0.5
0
-0.5
-1
-1.5
0.05
0.1
0.15
0.2
Notes:
Gap between platen and specimen at east top flange closes at 1 kip.
Loud sound at peak load L waves change to D waves in web.
470
Peak Load
600-1-48-H
600-1-48-H
-14
-12
-10
-8
-6
Ptest=11.16 kips
-4
-2
0
0.05
0.1
0.15
0.2
1.5
W
0.5
0
-0.5
-1
-1.5
0.05
0.1
0.15
0.2
Notes:
Local buckling and DH mode visible at 5 kips.

Loud noise at 9.5 kips L waves changes to D wave in web.
471
Peak Load
600-2-48-H
600-2-48-H
-14
-12
-10
-8
-6
Ptest=11.7 kips
-4
-2
0
0.05
0.1
0.15
0.2
1.5
W
0.5
0
-0.5
-1
-1.5
0.05
0.1
0.15
0.2
Notes:
Local buckling visible at 4 kips.

D wave interrupted by large crease.
L web waves change to D waves (9 kips post-peak, 7.5 kips post-peak)
472
Peak Load
600-3-48-H
600-3-48-H
-14
-12
-10
-8
-6
Ptest=11.16 kips
-4
-2
0
0.05
0.1
0.15
0.2
Notes:
Good platen bearing conditions.

Loud noise at 7.5 kips post-peak.
Yielding in the west flange first, then east flange.
473
1.5
W
0.5
0
-0.5
-1
-1.5
0.05
0.1
0.15
0.2
Appendix G
Residual stresses backstress for kinematic
hardening implementation
Implementation of a kinematic hardening rule requires that the center of the yield
surface,instressspace,beknownforanymaterialwhichhasbeenyieldedpriortothe
loading of interest. The coordinates of the center of the yield surface (1, 2, 3),
knownasthebackstress,cannotbedirectlycalculatedfromthestressesderivedherein
because work hardening was ignored in the residual stress derivations. However, the
plastic strains developed in the manufacturing process provide a means by which the
backstressmaybeapproximated,asprovidedinthisappendix.
Thegeneralequationforeffectivestressisdefinedas
e =
1
2
( 1 2 )2 + ( 2 3 )2 + ( 3 1 )2 .
(G.1)
Giventhatthethroughthicknesssheetstressesarezero(2=0),Eq.(G.1)reducesto
e = 1 2 1 3 + 3 2
(G.2)
Consider the contribution to the backstress that develops due to coiling. From Eq.
(6.18) we know the plastic strain, pcoiling. With pcoiling and knowing the material stress
strain relation (i.e., Figure 6.23) the effective stress at that plastic strain, eycoiling maybe
determined.ConsistentwiththeresidualstressderivationofEq.(6.8),weassume =0.3
and
474
1coiling = 3coiling .
(G.3)
Finally,substitutingtheprecedingintoEq.(G.2)resultsin
coiling
3
eycoiling
Similarly for cold bending the corners, from Eq. (6.21) we know the plastic strain,
2 + 1
(G.4)
pbend. With pbend and knowing the material stressstrain relation (i.e., Figure 6.23), we
determine the effective stress at that plastic strain, eybend. Consistent with the residual
stressderivationofEq.(6.12),weassume=0.5and
3bend = 1bend ,
(G.5)
andthusfind
bend
1
eybend
2 + 1
(G.6)
Thebackstressisthendeterminedas:
1 = 1coiling + 1bend yield
2 = 0
3 = 3coiling + 3bend yield ,
(G.7)
where yield isthevirginyieldstressofthesteel.Thisestimateassumesthatthechanges

in material properties from coiling, uncoiling, and flattening and coldforming do not
influenceoneanother.
475
Appendix H
Experiment true stress-strain curves
The average true plastic stressstrain curves are provided here for each of the 24
columntestsreportedinChapter5.Foreachspecimen,threeengineeringstressstrain
curves(westflange,eastflange,andweb)wereaveragedandthentransformedintotrue
stressesandstrainswiththefollowingequations:
true = ln(1 + o )
true = o (1 + o )
true and true are the true stress and strain and o and o are the engineering stress and
strain in the above equations. The tables in this appendix provide just the plastic
componentofthetruestrainsincethisiswhatisrequiredinABAQUS:
p = true yield ,where yield =
yield
E
The true stressstrain curves presented here were modified prior to their
implementationinABAQUStoensureplasticityinitiatedattheyieldstressandnot
theproportionallimit.RefertoSection7.2.1.4fordetailsonthismodelingdecision.
476

Specimen362124NH,362224NH,362324NH
100
90
80
true stress, ksi
70
60
50
40
30
Avg. Yield Stress=55.1 ksi
20
10
0
0.02
0.04
0.06
0.08
true plastic strain
true plastic
strain, p
true stress,
true
0
0.001
0.002
0.007
0.012
0.017
0.027
0.037
0.047
0.057
0.067
ksi
33.05
46.10
51.88
60.30
64.89
68.37
73.97
78.12
81.27
83.83
86.16
477
0.1
0.12
Specimen362124H
100
90
80
true stress, ksi
70
60
50
40
30
20
10
0
0.02
0.04
0.06
0.08
true plastic strain
true plastic
strain, p
true stress,
true
0
0.001
0.002
0.007
0.012
0.017
0.027
0.037
0.047
0.057
0.067
0.077
0.087
0.097
0.107
ksi
34.2
50.0
56.1
64.4
68.3
72.0
78.6
82.5
86.2
88.7
91.0
92.9
94.6
96.2
97.5
478
0.1
0.12
Specimen362224H
100
90
80
true stress, ksi
70
60
50
40
30
20
10
0
0.02
0.04
0.06
0.08
true plastic strain
true plastic
strain, p
true stress,
true
0
0.001
0.002
0.007
0.012
0.017
0.027
0.037
0.047
0.057
0.067
0.077
0.087
0.097
0.107
ksi
27.7
45.1
53.3
62.8
67.1
71.2
76.7
81.3
84.7
87.4
89.7
91.6
93.4
94.8
96.2
479
0.1
0.12
Specimen362324H
100
90
80
true stress, ksi
70
60
50
40
30
Avg. Yield Stress=56 ksi
20
10
0
0.02
0.04
0.06
0.08
true plastic strain
true plastic
strain, p
true stress,
true
0
0.001
0.002
0.007
0.012
0.017
0.027
0.037
0.047
0.057
0.067
0.077
0.087
0.097
0.107
ksi
31.7
47.8
53.6
61.1
64.7
68.4
74.8
78.6
82.2
84.6
86.9
88.8
90.4
91.9
93.1
480
0.1
0.12
Specimen362148NH
100
90
80
true stress, ksi
70
60
50
40
30
20
10
0
0.02
0.04
0.06
0.08
true plastic strain
true plastic
strain, p
true stress,
true
0
0.001
0.002
0.007
0.012
0.017
0.027
0.037
0.047
0.057
0.067
0.077
0.087
0.097
0.107
ksi
37.8
51.6
56.6
64.3
68.2
72.0
78.1
82.1
85.9
88.3
90.8
92.5
94.3
95.7
97.1
481
0.1
0.12
Specimen362248NH
100
90
80
true stress, ksi
70
60
50
40
30
20
10
0
0.02
0.04
0.06
0.08
true plastic strain
true plastic
strain, p
true stress,
true
0
0.001
0.002
0.007
0.012
0.017
0.027
0.037
0.047
0.057
0.067
0.077
0.087
0.097
0.107
ksi
35.0
50.9
56.5
64.3
68.3
72.1
78.2
82.3
86.1
88.5
90.9
92.8
94.5
96.0
97.3
482
0.1
0.12
Specimen362348NH
100
90
80
true stress, ksi
70
60
50
40
30
20
10
0
0.02
0.04
0.06
0.08
true plastic strain
true plastic
strain, p
true stress,
true
0
0.001
0.002
0.007
0.012
0.017
0.027
0.037
0.047
0.057
0.067
0.077
0.087
0.097
0.107
ksi
34
50
56
64
68
72
78
82
86
88
91
92
94
96
97
483
0.1
0.12
Specimen362148H
100
90
80
true stress, ksi
70
60
50
40
30
20
10
0
0.02
0.04
0.06
0.08
true plastic strain
true plastic
strain, p
true stress,
true
0
0.001
0.002
0.007
0.012
0.017
0.027
0.037
0.047
0.057
0.067
0.077
0.087
0.097
0.107
ksi
33.2
50.1
56.1
64.2
68.2
72.0
78.0
82.2
85.7
88.4
90.7
92.7
94.4
95.8
97.2
484
0.1
0.12
Specimen362248H
100
90
80
true stress, ksi
70
60
50
40
30
20
10
0
0.02
0.04
0.06
0.08
true plastic strain
true plastic
strain, p
true stress,
true
0
0.001
0.002
0.007
0.012
0.017
0.027
0.037
0.047
0.057
0.067
0.077
0.087
0.097
0.107
ksi
37.3
51.5
56.7
64.2
68.2
72.0
78.1
82.4
85.8
88.5
90.8
92.7
94.4
95.9
97.2
485
0.1
0.12
Specimen362348H
100
90
80
true stress, ksi
70
60
50
40
30
20
10
0
0.02
0.04
0.06
0.08
true plastic strain
true plastic
strain, p
true stress,
true
0
0.001
0.002
0.007
0.012
0.017
0.027
0.037
0.047
0.057
0.067
0.077
0.087
0.097
0.107
ksi
37.5
50.6
55.6
62.9
67.1
70.9
76.5
80.8
84.1
86.8
89.1
90.9
92.6
94.1
95.4
486
0.1
0.12
Specimen600124NH,600224NH,600324NH
100
90
80
true stress, ksi
70
60
50
40
30
20
10
0
0.02
0.04
0.06
0.08
true plastic strain
true plastic
strain, p
true stress,
true
0
0.001
0.002
0.007
0.012
0.017
0.027
0.037
0.047
0.057
0.067
0.077
0.087
0.097
0.107
ksi
37.5
54.7
58.3
60.0
61.5
64.0
70.2
74.4
77.5
80.0
81.9
83.5
84.9
86.1
87.2
487
0.1
0.12
Specimen600124H
100
90
80
true stress, ksi
70
60
50
40
30
20
10
0
0.02
0.04
0.06
0.08
true plastic strain
true plastic
strain, p
true stress,
true
0
0.001
0.002
0.007
0.012
0.017
0.027
0.037
0.047
0.057
0.067
0.077
0.087
0.097
0.107
ksi
35.0
58.9
61.4
62.3
62.9
63.6
69.4
74.4
77.8
80.5
82.6
84.4
85.9
87.2
88.4
488
0.1
0.12
Specimen600224H
100
90
80
true stress, ksi
70
60
50
40
30
20
10
0
0.02
0.04
0.06
0.08
true plastic strain
true plastic
strain, p
true stress,
true
0
0.001
0.002
0.007
0.012
0.017
0.027
0.037
0.047
0.057
0.067
0.077
0.087
0.097
0.107
ksi
33.0
54.0
57.2
58.9
61.5
64.4
70.5
74.8
78.0
80.4
82.4
84.1
85.5
86.7
87.9
489
0.1
0.12
Specimen600324H
100
90
80
true stress, ksi
70
60
50
40
30
20
10
0
0.02
0.04
0.06
0.08
true plastic strain
true plastic
strain, p
true stress,
true
0
0.001
0.002
0.007
0.012
0.017
0.027
0.037
0.047
0.057
0.067
0.077
0.087
0.097
0.107
ksi
34.9
57.5
60.1
60.8
61.8
63.6
69.8
74.1
77.6
79.9
82.0
83.7
85.1
86.5
87.6
490
0.1
0.12
Specimen600148NH
100
90
80
true stress, ksi
70
60
50
40
30
20
10
0
0.02
0.04
0.06
0.08
true plastic strain
true plastic
strain, p
true stress,
true
0
0.001
0.002
0.007
0.012
0.017
0.027
0.037
0.047
0.057
0.067
0.077
0.087
0.097
0.107
ksi
44.4
58.9
60.3
61.0
62.0
64.8
70.6
75.0
78.3
80.8
82.8
84.5
86.0
87.3
88.5
491
0.1
0.12
Specimen600248NH
100
90
80
true stress, ksi
70
60
50
40
30
20
10
0
0.02
0.04
0.06
0.08
true plastic strain
true plastic
strain, p
true stress,
true
0
0.001
0.002
0.007
0.012
0.017
0.027
0.037
0.047
0.057
0.067
0.077
0.087
0.097
0.107
ksi
41.9
62.1
63.1
63.6
64.0
64.5
69.8
75.0
78.5
81.2
83.5
85.3
87.0
88.4
89.6
492
0.1
0.12
Specimen600348NH
100
90
80
true stress, ksi
70
60
50
40
30
20
10
0
0.02
0.04
0.06
0.08
true plastic strain
true plastic
strain, p
true stress,
true
0
0.001
0.002
0.007
0.012
0.017
0.027
0.037
0.047
0.057
0.067
0.077
0.087
0.097
0.107
ksi
36.0
55.6
60.0
61.7
62.0
62.5
68.5
73.4
76.9
79.4
81.6
83.3
84.9
86.2
87.4
493
0.1
0.12
Specimen600148H
100
90
80
true stress, ksi
70
60
50
40
30
20
10
0
0.02
0.04
0.06
0.08
true plastic strain
true plastic
strain, p
true stress,
true
0
0.001
0.002
0.007
0.012
0.017
0.027
0.037
0.047
0.057
0.067
0.077
0.087
0.097
0.107
ksi
37.5
57.7
60.9
61.9
62.4
64.3
70.4
75.0
78.4
81.0
83.1
84.8
86.3
87.7
88.9
494
0.1
0.12
Specimen600248H
100
90
80
true stress, ksi
70
60
50
40
30
20
10
0
0.02
0.04
0.06
0.08
true plastic strain
true plastic
strain, p
true stress,
true
0
0.001
0.002
0.007
0.012
0.017
0.027
0.037
0.047
0.057
0.067
0.077
0.087
0.097
0.107
ksi
37.4
57.6
61.6
62.6
63.3
64.1
69.4
74.2
77.8
80.6
82.9
84.8
86.4
87.8
89.1
495
0.1
0.12
Specimen600348H
100
90
80
true stress, ksi
70
60
50
40
30
20
10
0
0.02
0.04
0.06
0.08
true plastic strain
true plastic
strain, p
true stress,
true
0
0.001
0.002
0.007
0.012
0.017
0.027
0.037
0.047
0.057
0.067
0.077
0.087
0.097
0.107
ksi
32.4
57.1
61.3
62.2
62.6
64.5
70.4
75.2
78.6
81.1
83.2
84.9
86.5
87.8
89.0
496
0.1
0.12
Appendix I
Column experiment nonlinear FE simulation results
Specimen362124NH
14
Experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
497
0.16
0.18
0.2
Specimen362224NH
14
Experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
498
0.16
0.18
0.2

Specimen362324NH
14
Experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
499
0.16
0.18
0.2
Specimen362124H
14
Experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
500
0.16
0.18
0.2
Specimen362224H
14
Experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
501
0.16
0.18
0.2
Specimen362324H
14
Experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
502
0.16
0.18
0.2
Specimen362148NH
14
Experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
503
0.16
0.18
0.2
Specimen362248NH
14
Experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
504
0.16
0.18
0.2
Specimen362348NH
14
Experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
505
0.16
0.18
0.2
Specimen362148H
14
Experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
506
0.16
0.18
0.2
Specimen362248H
14
Experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
507
0.16
0.18
0.2
Specimen362348H
14
Experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
508
0.16
0.18
0.2
Specimen600124NH
14
Experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
509
0.16
0.18
0.2
Specimen600224NH
14
Experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
510
0.16
0.18
0.2
Specimen600324NH
14
Experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
511
0.16
0.18
0.2
Specimen600124H
14
Experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
512
0.16
0.18
0.2
Specimen600224H
14
Experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
513
0.16
0.18
0.2
Specimen600324H
14
Experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
514
0.16
0.18
0.2
Specimen600148NH
14
Experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
515
0.16
0.18
0.2
Specimen600248NH
14
Experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
516
0.16
0.18
0.2
Specimen600348NH
14
Experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
517
0.16
0.18
0.2
Specimen600148H
14
Experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
518
0.16
0.18
0.2
Specimen600248H
14
Experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
519
0.16
0.18
0.2
Specimen600348H
14
Experiment
12
axial load, kips
10
0
0
0.02
0.04
0.06
0.08
0.1
0.12 0.14
520
0.16
0.18
0.2
Appendix J
Contact simulation in ABAQUS
The frictionbearing end conditions were chosen for the experimental program
describedinChapter5becausetheyallowedforconvenientalignmentandtestingofthe
columnspecimens.Theseboundaryconditionswereexpectedtobehaveasfixedfixed,
although during the test slipping of the crosssection and lifting off of the specimens
were observed for some specimens in the postpeak region of the loaddisplacement
curve.AnonlinearFEstudywasperformedinABAQUSwheretheexperimentfriction
bearingboundaryconditionswerereplicatedusingcontactmodeling(MoenandSchafer
2007a). These end conditions allowed deformation of the crosssection at the bearing
ends under load (slipping) and lift off of the bearing ends. A master analytical rigid
surfacewasdefinedtorepresentthetopandbottomplatenasshowninFigureJ.1and
each surface was assigned a reference node. The rigid surfaces simulate fixedfixed
conditionsbyrestrainingthereferencenodedegreesoffreedom,andthespecimenwas
loadedbyapplyinganimposeddisplacementtothebottomsurfacereferencenode.Top
andbottomnodebasedslavesurfacesweredefinedtosimulatethebearingendofeach
specimen.Thetributarybearingareawasdefinedateachnodeintheslave surfaceto
ensurethatcontactstressesweresimulatedaccuratelyinABAQUS.
521
Restrain rigid surface

reference node in 1 to 6
directions
ABAQUS Analytical Rigid

Surface (Typ.)
Assign friction contact

behavior in ABAQUS
between rigid surface and
specimen
1
Apply imposed
displacement of
surface in 1 direction
5
2
Restrain rigid surface

reference node in 2 to 6
directions
6
3
FigureJ.1ContactboundaryconditionasimplementedinABAQUS
ACoulombfrictionmodelwasenforcedinABAQUSbetweenthemasterandslave
surfacesbydefiningastaticandkineticcoefficientoffriction, sandk,forsteelonsteel
contact. The assumed values for s and k were 0.7 and 0.6 in this study (Oden and
Martins 1985). Slip occurs in the model once the shear stress at the contact interface
exceedssfn,wherefnisthenormalcontactstressatthebearingsurface.
Thelocationsoftherigidsurfacesweredefinedtobeincontactwiththespecimen
endswhenthefirststepoftheanalysisbegan.Thisdoesnotguaranteeperfectcontactin
a computational sense, and so the ABAQUS command ADJUST was used to zero the
contact surface and avoid numerical instabilities during the first analysis step. The
ADJUST command modifies the geometry of the specimen to close infinitesimal gaps,
butdoesnotresultininternalforcesormomentsinthespecimen.
522
Toevaluatetheinfluenceofthecontactboundaryconditions,theloaddisplacement
response of specimen 600224NH assuming contact boundary conditions was

evaluated against an FE simulation employing the fixedfixed boundary conditions
described in Figure 7.27. The results of the two simulations are almost identical until
wellintothepostpeakrangeasshowninFigureJ.2,demonstratingthatthefixedfixed
boundaryconditionsareaviableapproximationtotheactualboundaryconditionsinan
FEsimulation.
20
Contact Surface
Fixed-Fixed
18
column axial load, kips
16
14
12
10
8
6
The failure response with frictionbearing boundary conditions is

similar to that with fixed-fixed
boundary conditions
4
2
0
0.05
0.1
0.15
0.2
FigureJ.2.AcomparisonofABAQUSnonlinearsolutionsconsideringfixedfixedandcontactboundary
conditionsforspecimen600224NH
523
Appendix K
Simulated column experiments database
Thetableprovidedinthisappendixsummarizesthedimensions,elasticbucklingloads,
andtestedstrengthsofsimulatedcolumnexperimentsdescribedinSection8.1.10.The
letters in the Study type column denote simulations considered in the DSM failure
modestudies:DforthedistortionalbucklingfailurestudyinSection8.1.2,Gforthe
globalbucklingfailurestudyinSection8.1.3,andLforthelocalbucklingfailurestudy
inSection8.1.4.
Thedatabasecontainscolumnswithslottedholesorcircularholes.Toidentifythehole
type in a specific column, use the following rule: columns with slotted holes always
haveLhole=4in,andLhole=hholeforcolumnswithcircularholes.
Thefollowingnotationisemployedtodenotesimulatedstrengths:
Ptest25
25% CDF local and distortional buckling, no global imperfections
Ptest75
75% CDF local and distortional buckling, no global imperfections
Ptest25+
25% CDF local and distortional buckling, +L/2000 global imperfection
Ptest75+
75% CDF local and distortional buckling, +L/1000 global imperfection
Ptest25-
25% CDF local and distortional buckling, -L/2000 global imperfection
Ptest75-
75% CDF local and distortional buckling, -L/1000 global imperfection
524
ID #
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
SSMA
section
600S250-97
600S162-97
800S250-97
800S200-97
800S162-97
600S137-68
1000S250-97
1000S200-97
800S162-68
1000S162-97
600S137-43
1200S250-97
1000S200-68
1000S250-54
1200S162-97
1000S162-68
800S162-43
1200S250-68
1200S200-68
1000S200-43
600S250-97
600S162-97
800S250-97
800S200-97
800S162-97
600S137-68
1000S250-97
1000S200-97
800S162-68
1000S162-97
600S137-43
1200S250-97
1000S200-68
1000S250-54
1200S162-97
1000S162-68
800S162-43
1200S250-68
1200S200-68
600S250-97
600S162-97
800S250-97
800S200-97
800S162-97
600S137-68
1000S250-97
1000S200-97
800S162-68
1000S162-97
600S137-43
1200S250-97
1000S200-68
1000S250-54
1200S162-97
1000S162-68
800S162-43
1200S200-68
1000S200-43
600S250-97
600S162-97
800S250-97
800S200-97
800S162-97
600S137-68
1000S250-97
1000S200-97
800S162-68
1000S162-97
600S137-43
1200S250-97
1000S200-68
1000S250-54
1200S162-97
1000S162-68
800S162-43
1200S250-68
1200S200-68
1000S200-43
600S250-97
600S162-97
800S250-97
800S200-97
800S162-97
600S137-68
1000S250-97
1000S200-97
800S162-68
1000S162-97
600S137-43
1200S250-97
1000S200-68
1000S250-54
1200S162-97
1000S162-68
800S162-43
1200S250-68
1200S200-68
1000S200-43
1200S200-97
L
in.
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
26
Lhole
in.
4.59
3.79
5.39
4.99
4.59
3.59
6.19
5.79
4.69
5.39
3.66
6.99
5.89
6.33
6.19
5.49
4.76
7.09
6.69
5.96
3.45
2.85
4.05
3.75
3.45
2.69
4.65
4.35
3.51
4.05
2.75
5.25
4.41
4.75
4.65
4.11
3.57
5.31
5.01
2.30
1.90
2.70
2.50
2.30
1.79
3.10
2.90
2.34
2.70
1.83
3.50
2.94
3.17
3.10
2.74
2.38
3.34
2.98
1.15
0.95
1.35
1.25
1.15
0.90
1.55
1.45
1.17
1.35
0.92
1.75
1.47
1.58
1.55
1.37
1.19
1.77
1.67
1.49
4.59
3.79
5.39
4.99
4.59
3.59
6.19
5.79
4.69
5.39
3.66
6.99
5.89
6.33
6.19
5.49
4.76
7.09
6.69
5.96
3.30
hhole
in.
4.59
3.79
5.39
4.99
4.59
3.59
6.19
5.79
4.69
5.39
3.66
6.99
5.89
6.33
6.19
5.49
4.76
7.09
6.69
5.96
3.45
2.85
4.05
3.75
3.45
2.69
4.65
4.35
3.51
4.05
2.75
5.25
4.41
4.75
4.65
4.11
3.57
5.31
5.01
2.30
1.90
2.70
2.50
2.30
1.79
3.10
2.90
2.34
2.70
1.83
3.50
2.94
3.17
3.10
2.74
2.38
3.34
2.98
1.15
0.95
1.35
1.25
1.15
0.90
1.55
1.45
1.17
1.35
0.92
1.75
1.47
1.58
1.55
1.37
1.19
1.77
1.67
1.49
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
3.30
S
in.
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
13
# of
holes
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
Pyg
kips
68.5
56.5
80.4
74.4
68.5
37.5
92.3
86.3
48.9
80.4
24.2
104.2
61.5
52.5
92.3
57.3
31.5
74.0
69.8
39.4
68.5
56.5
80.4
74.4
68.5
37.5
92.3
86.3
48.9
80.4
24.2
104.2
61.5
52.5
92.3
57.3
31.5
74.0
69.8
68.5
56.5
80.4
74.4
68.5
37.5
92.3
86.3
48.9
80.4
24.2
104.2
61.5
52.5
92.3
57.3
31.5
69.8
39.4
68.5
56.5
80.4
74.4
68.5
37.5
92.3
86.3
48.9
80.4
24.2
104.2
61.5
52.5
92.3
57.3
31.5
74.0
69.8
39.4
68.5
56.5
80.4
74.4
68.5
37.5
92.3
86.3
48.9
80.4
24.2
104.2
61.5
52.5
92.3
57.3
31.5
74.0
69.8
39.4
98.2
Pynet
kips
41.1
33.9
48.2
44.6
41.1
22.5
55.4
51.8
29.4
48.2
14.5
62.5
36.9
31.5
55.4
34.4
18.9
44.4
41.9
23.6
47.9
39.6
56.3
52.1
47.9
26.2
64.6
60.4
34.3
56.3
16.9
72.9
43.0
36.7
64.6
40.1
22.0
51.8
48.9
54.8
45.2
64.3
59.5
54.8
30.0
73.8
69.1
39.2
64.3
19.4
83.4
49.2
42.0
73.8
45.8
25.2
55.9
31.5
61.6
50.9
72.3
67.0
61.6
33.7
83.1
77.7
44.0
72.3
21.8
93.8
55.3
47.2
83.1
51.6
28.3
66.6
62.9
35.5
68.5
56.5
80.4
74.4
68.5
37.5
92.3
86.3
48.9
80.4
24.2
104.2
61.5
52.5
92.3
57.3
31.5
74.0
69.8
39.4
78.6
Pynet/Pyg
0.60
0.60
0.60
0.60
0.60
0.60
0.60
0.60
0.60
0.60
0.60
0.60
0.60
0.60
0.60
0.60
0.60
0.60
0.60
0.60
0.70
0.70
0.70
0.70
0.70
0.70
0.70
0.70
0.70
0.70
0.70
0.70
0.70
0.70
0.70
0.70
0.70
0.70
0.70
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.90
0.90
0.90
0.90
0.90
0.90
0.90
0.90
0.90
0.90
0.90
0.90
0.90
0.90
0.90
0.90
0.90
0.90
0.90
0.90
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.80
525
Pcrl
kips
52.3
43.1
34.7
32.4
27.6
13.8
25.5
23.5
10.4
19.4
3.7
19.7
8.5
4.5
14.7
7.6
2.7
7.0
6.7
2.2
52.3
43.1
34.7
32.4
27.6
13.8
25.5
23.5
10.4
19.4
3.7
19.7
8.5
4.5
14.7
7.6
2.7
7.0
6.7
52.3
43.1
34.7
32.4
27.6
13.8
25.5
23.5
10.4
19.4
3.7
19.7
8.5
4.5
14.7
7.6
2.7
6.7
2.2
52.3
43.1
34.7
32.4
27.6
13.8
25.5
23.5
10.4
19.4
3.7
19.7
8.5
4.5
14.7
7.6
2.7
7.0
6.7
2.2
52.3
43.1
34.7
32.4
27.6
13.8
25.5
23.5
10.4
19.4
3.7
19.7
8.5
4.5
14.7
7.6
2.7
7.0
6.7
2.2
17.8
Pcrd
kips
54.4
38.4
38.2
31.6
22.9
11.6
25.5
20.2
10.4
14.6
4.6
17.2
9.3
7.6
9.9
6.5
3.8
7.9
6.9
3.8
56.7
40.6
40.0
33.4
24.6
12.5
27.0
21.6
11.0
16.0
4.9
18.4
9.8
7.9
11.1
7.0
4.0
8.4
7.3
58.9
42.8
41.8
35.1
26.2
13.3
28.4
23.0
11.6
17.3
5.1
19.6
10.3
8.2
12.4
7.4
4.1
7.6
4.1
61.2
45.0
43.6
36.8
27.8
14.1
29.8
24.3
12.2
18.6
5.3
20.8
10.8
8.5
13.6
7.9
4.3
9.2
8.0
4.3
63.0
47.1
45.4
38.4
29.4
14.9
31.2
25.6
12.8
19.9
5.5
22.0
11.3
8.8
14.8
8.3
4.4
9.6
8.4
4.4
17.3
Pcre
kips
179.1
99.4
227.6
169.9
111.7
49.8
246.2
176.8
82.1
113.4
33.0
242.2
128.5
142.5
108.3
83.6
54.0
174.2
124.1
83.8
221.4
116.2
294.5
213.2
136.4
57.5
338.4
234.8
100.9
146.0
38.3
358.9
172.0
198.4
148.6
108.4
66.8
260.8
178.2
261.9
132.5
362.1
256.9
151.2
62.4
434.3
294.8
116.4
159.0
43.3
483.4
217.2
256.8
164.7
122.4
79.3
235.6
143.1
299.2
142.2
426.6
289.4
153.2
63.1
527.9
306.2
118.0
160.9
43.8
563.0
231.0
313.9
166.7
124.0
80.4
419.7
240.7
155.1
332.0
142.7
484.1
290.8
153.7
63.3
541.3
307.6
118.4
161.5
44.0
566.1
232.1
329.7
167.2
124.4
80.7
422.1
241.8
155.9
267.6
Ptest25
kips
30.4
25.6
36.7
31.7
28.3
14.9
35.1
33.7
18.7
30.1
7.7
36.7
19.4
15.3
31.2
17.9
8.9
21.9
21.0
10.2
38.0
30.3
43.0
37.8
32.4
16.9
41.0
37.2
19.3
34.4
9.1
38.4
22.0
16.3
33.9
18.6
9.8
23.4
21.6
43.8
34.3
46.0
41.7
34.2
18.1
44.9
38.9
20.3
35.2
9.5
39.7
22.9
16.0
34.5
19.7
10.0
22.1
10.5
48.1
36.4
47.6
38.8
34.8
18.2
44.8
39.3
20.5
35.7
9.9
39.8
22.6
16.1
34.7
19.9
10.1
22.5
24.2
10.5
50.4
38.7
48.7
40.8
35.1
18.3
44.5
44.3
22.9
35.4
9.6
45.7
21.9
16.6
34.7
20.0
10.3
21.9
24.2
11.1
35.7
Ptest75
kips
30.8
25.3
37.1
30
27
14.3
34.4
33.8
19.3
29.7
7.82
33.8
19.5
15.2
32.4
17.7
8.8
19.9
18.4
10.2
36
28.1
42.4
37.6
30.7
16.1
41.1
35.1
18.5
38.4
9.5
34
21.6
15.4
34.8
18.8
9.75
23.3
20.1
41.4
30.3
39.6
40.1
31.6
16.4
41
36.4
19.2
40.1
9.96
36.3
22.3
15.3
35.2
18.7
10
22.6
11
42.6
31
43.7
40.2
32
16.6
41
36.7
19.5
40.6
9.97
36.6
22.5
15.4
35.3
18.9
10.1
23.6
22.8
11.1
46.8
34
45.9
37
32.1
16.6
43.1
42.6
24.9
33.7
9.27
40.3
20.5
15.7
35.3
19.8
10
24.7
22.8
10.9
31.3
Ptest25+
kips
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
Ptest75+
kips
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
Ptest25kips
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
Ptest75kips
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
Study
Type
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
ID #
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
SSMA
section
1200S162-54
1200S250-68
1000S200-68
1000S162-54
800S137-68
1000S162-43
1200S162-54
250S137-54
250S137-54
400S162-68
600S250-97
350S162-54
250S162-33
250S137-33
362S137-43
362S137-68
250S162-54
600S137-54
250S137-33
800S137-97
800S137-97
250S137-68
250S162-68
250S162-68
250S137-54
250S137-54
400S162-68
600S250-97
250S162-33
250S137-33
362S137-43
362S137-68
250S162-54
600S137-54
250S137-33
800S137-97
800S137-97
250S162-68
250S137-68
250S162-68
250S162-68
250S162-68
250S137-68
250S162-68
250S162-68
250S137-54
250S137-54
400S162-68
600S250-97
350S162-54
250S162-33
250S137-33
362S137-43
362S137-68
250S162-54
600S137-54
250S137-33
800S137-97
800S137-97
350S162-68
1000S200-97
350S162-54
800S200-68
550S162-54
800S200-54
600S250-43
600S162-43
800S250-43
800S162-43
1000S250-43
350S162-68
1000S200-97
350S162-54
800S200-68
550S162-54
800S200-54
600S250-43
600S162-43
800S250-43
800S162-43
1000S250-43
400S162-68
250S137-33
250S137-33
362S200-43
362S137-33
800S137-54
400S162-33
600S162-43
800S250-54
250S137-68
250S137-68
250S162-68
250S137-54
250S162-68
250S137-68
362S200-68
350S162-68
250S137-68
L
in.
24
46
42
36
32
44
46
26
32
54
92
66
58
60
84
88
96
96
94
94
96
12
16
22
26
32
54
92
58
60
84
88
96
96
94
94
96
8
12
16
22
8
12
16
22
26
32
54
92
66
58
60
84
88
96
96
94
94
96
34
88
24
74
42
66
56
32
74
40
80
34
88
24
74
42
66
56
32
74
40
80
42
32
18
20
30
42
18
26
34
10
12
16
18
22
24
40
40
34
Lhole
in.
3.17
3.54
2.94
2.77
2.19
2.78
3.17
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
1.09
1.24
1.24
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
0.62
0.55
0.62
0.62
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
2.53
5.07
2.56
4.45
3.26
4.49
4.17
3.47
4.87
4.17
5.57
1.44
2.90
1.47
2.54
1.87
2.57
2.38
1.98
2.78
2.38
3.18
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
hhole
in.
3.17
3.54
2.94
2.77
2.19
2.78
3.17
1.11
1.11
1.54
2.30
1.47
1.29
1.14
1.35
1.32
1.27
1.81
1.14
2.15
2.15
1.09
1.24
1.24
0.56
0.56
0.77
1.15
0.64
0.57
0.68
0.66
0.63
0.91
0.57
1.07
1.07
0.62
0.55
0.62
0.62
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
2.53
5.07
2.56
4.45
3.26
4.49
4.17
3.47
4.87
4.17
5.57
1.44
2.90
1.47
2.54
1.87
2.57
2.38
1.98
2.78
2.38
3.18
1.54
1.14
1.14
1.71
1.36
2.21
1.59
1.98
2.77
1.09
1.09
1.24
1.11
1.24
1.09
1.67
1.44
1.09
S
in.
12
15
14
12
16
14
15
13
16
13
13
13
14
12
12
12
12
12
13
13
12
12
16
22
13
16
13
13
14
12
12
12
12
12
13
13
12
8
12
16
22
8
12
16
22
13
16
13
13
13
14
12
12
12
12
12
13
13
12
17
12
12
12
14
13
14
16
12
13
13
17
12
12
12
14
13
14
16
12
13
13
14
16
18
20
15
14
18
13
17
10
12
16
18
22
12
13
13
17
# of
holes
2
3
3
3
2
3
3
2
2
4
7
5
4
5
7
7
8
8
7
7
8
1
1
1
2
2
4
7
4
5
7
7
8
8
7
7
8
1
1
1
1
1
1
1
1
2
2
4
7
5
4
5
7
7
8
8
7
7
8
2
7
2
6
3
5
4
2
6
3
6
2
7
2
6
3
5
4
2
6
3
6
3
2
1
1
2
3
1
2
2
1
1
1
1
1
2
3
3
2
Pyg
kips
52.5
74.0
61.5
45.9
45.8
36.7
52.5
18.5
18.5
32.2
68.5
24.3
13.1
11.6
17.9
27.5
21.0
30.1
11.6
64.0
64.0
22.8
26.0
26.0
18.5
18.5
32.2
68.5
13.1
11.6
17.9
27.5
21.0
30.1
11.6
64.0
64.0
26.0
22.8
26.0
26.0
26.0
22.8
26.0
26.0
18.5
18.5
32.2
68.5
24.3
13.1
11.6
17.9
27.5
21.0
30.1
11.6
64.0
64.0
30.1
86.3
24.3
53.1
30.9
42.5
31.5
26.2
36.7
31.5
42.0
30.1
86.3
24.3
53.1
30.9
42.5
31.5
26.2
36.7
31.5
42.0
32.2
11.6
11.6
22.5
13.8
36.7
16.1
26.2
45.9
22.8
22.8
26.0
18.5
26.0
22.8
34.8
30.1
22.8
Pynet
kips
42.0
59.2
49.2
36.7
36.7
29.4
42.0
14.8
14.8
25.8
54.8
19.4
10.5
9.2
14.3
22.0
16.8
24.1
9.2
51.2
51.2
18.3
20.8
20.8
16.6
16.6
29.0
61.6
11.8
10.4
16.1
24.8
18.9
27.1
10.4
57.6
57.6
23.4
20.5
23.4
23.4
26.0
22.8
26.0
26.0
18.5
18.5
32.2
68.5
24.3
13.1
11.6
17.9
27.5
21.0
30.1
11.6
64.0
64.0
19.6
56.1
15.8
34.5
20.1
27.6
20.5
17.0
23.9
20.5
27.3
24.1
69.1
19.4
42.5
24.7
34.0
25.2
20.9
29.4
25.2
33.6
25.8
9.2
9.2
18.0
11.1
29.4
12.9
20.9
36.7
18.3
18.3
20.8
14.8
20.8
18.3
27.9
24.1
18.3
Pynet/Pyg
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.90
0.90
0.90
0.90
0.90
0.90
0.90
0.90
0.90
0.90
0.90
0.90
0.90
0.90
0.90
0.90
0.90
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.65
0.65
0.65
0.65
0.65
0.65
0.65
0.65
0.65
0.65
0.65
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.80
526
Pcrl
kips
3.1
7.0
8.5
4.0
8.8
2.0
3.1
24.9
24.9
27.4
52.3
16.7
6.3
5.7
7.4
28.8
27.4
7.2
5.7
22.9
22.9
49.5
54.7
54.7
24.9
24.9
27.4
52.3
6.3
5.7
7.4
28.5
27.4
7.0
5.7
22.9
22.9
54.7
49.5
54.7
54.7
54.7
49.5
54.7
54.7
24.9
24.9
27.4
52.3
16.7
6.3
5.7
7.4
28.8
27.4
7.2
5.7
22.9
22.9
33.5
23.5
16.7
11.4
8.8
5.8
4.6
4.0
3.1
2.7
2.3
33.5
23.5
16.7
11.4
8.8
5.8
4.6
4.0
3.1
2.7
2.3
27.4
5.7
5.7
8.9
3.3
4.8
3.2
4.0
6.1
49.5
49.5
54.7
24.9
54.7
49.5
35.3
33.5
49.5
Pcrd
kips
3.0
8.8
10.3
4.5
7.9
2.5
3.0
19.8
19.8
30.4
55.6
20.6
8.8
7.2
9.7
24.0
23.5
7.1
7.2
15.8
15.8
38.5
43.6
43.6
19.8
19.8
30.4
55.6
8.8
7.2
9.7
24.0
23.5
7.1
7.2
15.8
15.8
44.7
39.8
44.7
44.7
45.0
40.9
45.0
45.0
24.7
24.7
35.6
63.0
24.0
9.8
8.7
11.6
31.5
27.8
8.9
8.7
22.9
22.9
35.2
20.9
21.9
15.2
13.0
9.6
10.6
6.8
6.9
3.9
4.5
37.0
23.0
22.9
16.1
13.7
10.1
11.0
7.2
7.3
4.1
4.8
30.4
7.2
7.2
14.7
5.8
4.2
6.9
6.7
11.9
31.1
31.1
38.2
19.8
38.2
31.1
37.8
32.7
31.1
Pcre
kips
105.2
96.9
71.7
45.1
37.5
24.8
28.6
12.8
9.5
11.6
18.8
5.3
2.4
1.6
2.1
3.9
2.0
3.2
0.9
5.5
5.2
89.9
68.9
37.4
13.2
9.8
12.0
19.5
2.4
1.7
2.2
4.0
2.1
3.3
0.9
5.5
5.3
280.3
95.2
74.5
40.8
303.2
99.7
77.3
41.9
18.5
12.7
16.4
26.7
7.5
3.3
2.4
3.1
4.1
2.7
3.3
1.1
5.6
5.3
25.9
18.6
36.6
16.0
21.5
16.4
19.1
33.0
13.4
23.4
13.8
28.7
22.4
42.9
21.3
25.4
21.5
24.1
38.6
18.6
28.6
19.7
16.9
5.0
15.2
53.7
9.7
17.0
42.7
42.8
90.9
92.2
64.8
50.7
24.3
27.7
18.0
21.6
15.6
10.2
Ptest25
kips
13.5
22.6
21.3
12.4
16.9
8.3
10.0
11.2
10.9
13.7
24.4
7.9
3.1
3.2
2.6
3.7
3.5
2.9
1.5
5.0
4.7
16.9
19.3
18.2
12.8
11.7
15.2
25.6
3.2
2.4
2.8
3.9
4.0
3.0
1.6
5.0
4.8
24.0
19.9
23.5
22.9
24.7
20.6
23.0
22.8
15.1
11.9
16.3
25.6
7.5
5.4
2.4
3.0
4.0
4.2
3.1
1.2
5.1
4.9
13.1
15.7
11.2
12.6
10.7
10.2
9.6
8.8
8.7
8.3
8.2
19.7
17.0
15.2
13.9
12.7
11.2
11.7
10.1
8.8
8.1
8.8
17.5
5.6
5.8
11.1
5.6
8.8
6.8
10.7
17.9
14.8
15.1
17.6
11.9
16.7
14.3
20.2
16.4
12.7
Ptest75
kips
13.5
22.4
21
12.5
17.2
8.68
10.3
10.2
10.1
12.1
22.2
7.14
4.29
2.91
2.37
3.37
3.29
2.64
1.36
4.6
4.4
15.8
19.3
17.8
11.2
10.2
12.6
22.8
3.09
2.63
2.4
3.47
3.51
2.67
1.39
4.64
4.43
21.3
17
21.2
19.8
21.5
17.1
21.4
19.9
11.7
10.5
13.4
23.3
7.65
3.06
2.41
2.48
3.59
3.63
2.76
1.42
4.71
4.5
12.9
15.3
11
12.4
10.2
10.2
9.81
8.8
8.68
8.33
7.74
18
16.1
14.3
13.6
11.9
11.2
11.6
10
9.43
8.82
8.78
15.7
5.33
5.41
11.2
5.6
8.79
7.06
10.7
16.7
13.2
13.8
17.3
11.1
16.5
13.8
18.7
14.7
11.9
Ptest25+
kips
NaN
NaN
NaN
NaN
NaN
NaN
NaN
10.6
9.71
13.2
23.2
6.8
2.86
2.16
2.63
3.75
2.54
3.03
1.07
5.09
4.85
NaN
NaN
NaN
12.30
10.20
14.40
23.20
2.91
2.22
2.76
3.85
2.61
3.22
1.10
5.18
4.94
NaN
NaN
NaN
NaN
24.7
20.6
23
22.8
13.4
10.5
14.6
23.4
7
2.98
2.27
2.76
3.9
2.68
3.25
1.12
5.32
5.08
13
16.2
11.2
12.2
10.9
9.97
9.51
8.83
8.87
7.89
NaN
19.3
17.9
15.1
13.5
13
11
11.6
10.2
9.38
8.46
8.14
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
Ptest75+
kips
NaN
NaN
NaN
NaN
NaN
NaN
NaN
9.51
8.51
11.4
20.5
6.27
2.59
2.04
2.37
3.2
2.46
2.91
1.03
4.83
4.6
NaN
NaN
NaN
10.20
8.66
12.00
20.50
2.62
2.07
2.41
3.26
2.52
3.07
1.06
4.87
4.64
NaN
NaN
NaN
NaN
21.5
17.1
21.4
19.9
10.7
8.94
12.8
20.8
6.52
2.72
2.15
2.48
3.31
2.61
3.09
1.1
4.94
4.71
12.7
16.2
11
11.8
10.6
9.85
9.01
9.04
8.9
8.18
8
17.4
17.3
14.3
13
12.3
10.7
11.4
10.1
9.71
8.68
8.84
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
Ptest25kips
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
2.63
3.63
NaN
2.76
NaN
4.83
4.61
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
2.76
3.83
NaN
2.83
NaN
4.90
4.69
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
2.61
3.99
NaN
2.94
NaN
5.01
4.81
13
15.2
11.2
12.9
10.4
10.4
9.77
8.77
8.58
NaN
8.11
18.3
16.3
15.1
14.3
12.5
11.5
11.9
9.98
NaN
9.14
8.69
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
Ptest75kips
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
2.37
3.23
NaN
2.53
NaN
4.41
4.21
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
2.41
3.32
NaN
2.55
NaN
4.44
4.25
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
2.48
3.43
NaN
2.62
NaN
4.5
4.3
12.7
14.5
11
13.1
9.86
10.6
9.93
8.72
8.39
6.88
7.46
17.4
15.2
14.3
14.3
11.5
11.6
11.9
9.89
9.19
9.25
8.62
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
Study
Type
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
ID #
200
201
202
203
204
205
206
208
209
210
211
212
213
SSMA
section
362S137-68
250S162-43
350S162-68
1000S200-97
350S162-54
800S200-68
550S162-54
800S200-54
600S250-43
600S162-43
800S250-43
800S162-43
1000S250-43
L
in.
46
42
34
88
24
74
42
66
56
32
74
40
80
Lhole
in.
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
hhole
in.
1.32
1.28
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
S
in.
15
14
17
12
12
12
14
13
14
16
12
13
13
# of
holes
3
3
2
7
2
6
3
5
4
2
6
3
6
Pyg
kips
27.5
16.9
30.1
86.3
24.3
53.1
30.9
42.5
31.5
26.2
36.7
31.5
42.0
Pynet
kips
22.0
13.5
30.1
86.3
24.3
53.1
30.9
42.5
31.5
26.2
36.7
31.5
42.0
Pynet/Pyg
0.80
0.80
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
527
Pcrl
kips
28.8
13.9
33.5
23.5
16.7
11.4
8.8
5.8
4.6
4.0
3.1
2.7
2.3
Pcrd
kips
24.0
14.9
39.4
25.6
24.0
17.4
14.6
10.8
11.1
7.6
7.7
4.4
5.1
Pcre
kips
9.9
5.3
31.5
22.9
49.4
23.1
29.0
23.8
29.6
42.1
24.9
29.1
24.2
Ptest25
kips
11.9
7.2
22.6
17.8
16.9
14.1
13.0
11.6
12.2
10.1
9.7
8.7
8.9
Ptest75
kips
10.8
6.67
19.4
16.5
15
13.9
12
11.4
12
10.1
9.64
8.94
8.89
Ptest25+
kips
NaN
NaN
22.3
19.4
16.9
14
13.2
11.3
12
10.2
9.84
NaN
8.82
Ptest75+
kips
NaN
NaN
18.7
18
15
13.2
12.5
11
11.8
10.3
9.9
8.74
9.64
Ptest25kips
NaN
NaN
19.3
17
16.9
14.7
12.7
11.8
12.3
10
9.62
9.32
8.79
Ptest75kips
NaN
NaN
18.7
15.5
15
14.5
11.6
11.9
12.2
9.97
9.4
9.42
8.72
Study
Type
L
L
L
L
L
L
L
L
L
L
L
Appendix L
Simulated beam experiment database
The table provided in this appendix summarizes the dimensions, elastic buckling
moments, and tested strengths of simulated beam experiments described in Section
8.2.1.ThelettersintheStudytypecolumndenotesimulationsconsideredintheDSM
failuremodestudies:DforthedistortionalbucklingfailurestudyinSection8.2.3and
LforthelocalbucklingfailurestudyinSection8.2.2.
528
ID #
SSMA section
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
400S162-68
400S137-54
550S162-54
800S137-68
800S162-54
800S137-54
1200S250-68
1200S162-68
1200S250-54
1000S162-43
1200S162-54
400S162-68
400S137-54
550S162-54
800S137-68
800S162-54
800S137-54
1200S250-68
1200S162-68
1200S250-54
1000S162-43
1200S162-54
400S162-68
400S137-54
550S162-54
800S137-68
800S162-54
800S137-54
1200S250-68
1200S162-68
1200S250-54
1000S162-43
1200S162-54
400S162-68
400S137-54
550S162-54
800S137-68
800S162-54
800S137-54
1200S250-68
1200S162-68
1200S250-54
1000S162-43
1200S162-54
400S162-68
400S137-54
550S162-54
800S137-68
800S162-54
800S137-54
1200S250-68
1200S162-68
1200S250-54
1000S162-43
1200S162-54
400S162-68
400S137-54
550S162-54
800S137-68
800S162-54
800S137-54
1200S250-68
1200S162-68
1200S250-54
1000S162-43
1200S162-54
400S162-68
400S137-54
550S162-54
800S137-68
800S162-54
800S137-54
1200S250-68
1200S162-68
1200S250-54
1000S162-43
1200S162-54
L
in.
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
Lhole
in.
3.1
2.8
3.7
4.4
4.7
4.4
7.1
6.3
7.1
5.6
6.3
2.3
2.1
2.8
3.3
3.5
3.3
5.3
4.7
5.3
4.2
4.7
1.5
1.4
1.9
2.2
2.4
2.2
3.5
3.1
3.6
2.8
3.2
0.8
0.7
0.9
1.1
1.2
1.1
1.8
1.6
1.8
1.4
1.6
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
2.8
2.7
3.7
4.7
5.0
4.8
7.4
6.9
7.5
6.0
6.9
2.2
2.2
2.9
3.8
3.9
3.8
5.9
5.5
5.9
4.7
5.5
hhole
in.
3.1
2.8
3.7
4.4
4.7
4.4
7.1
6.3
7.1
5.6
6.3
2.3
2.1
2.8
3.3
3.5
3.3
5.3
4.7
5.3
4.2
4.7
1.5
1.4
1.9
2.2
2.4
2.2
3.5
3.1
3.6
2.8
3.2
0.8
0.7
0.9
1.1
1.2
1.1
1.8
1.6
1.8
1.4
1.6
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
2.8
2.7
3.7
4.7
5.0
4.8
7.4
6.9
7.5
6.0
6.9
2.2
2.2
2.9
3.8
3.9
3.8
5.9
5.5
5.9
4.7
5.5
# of holes
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
0
0
0
0
0
0
0
0
0
0
0
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
S
in.
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
Fy
ksi
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
529
My
kipin.
39.4
27.9
49.5
92.2
84.0
74.8
239.0
190.5
192.1
94.0
153.5
39.4
27.9
49.5
92.2
84.0
74.8
239.0
190.5
192.1
94.0
153.5
39.4
27.9
49.5
92.2
84.0
74.8
239.0
190.5
192.1
94.0
153.5
39.4
27.9
49.5
92.2
84.0
74.8
239.0
190.5
192.1
94.0
153.5
39.4
27.9
49.5
92.2
84.0
74.8
239.0
190.5
192.1
94.0
153.5
39.4
27.9
49.5
92.2
84.0
74.8
239.0
190.5
192.1
94.0
153.5
39.4
27.9
49.5
92.2
84.0
74.8
239.0
190.5
192.1
94.0
153.5
Mynet
kipin.
34.3
24.8
44.3
84.9
76.7
68.8
218.3
176.1
175.4
86.4
141.9
37.2
26.6
47.3
89.1
80.9
72.3
230.3
184.4
185.1
90.8
148.6
38.8
27.5
48.8
91.3
83.1
74.0
236.4
188.7
190.0
93.1
152.1
39.3
27.8
49.4
92.1
83.9
74.7
238.7
190.3
191.9
93.9
153.4
39.4
27.9
49.5
92.2
84.0
74.8
239.0
190.5
192.1
94.0
153.5
35.5
25.1
44.5
83.0
75.6
67.3
215.1
171.4
172.9
84.6
138.2
37.4
26.5
47.0
87.6
79.8
71.1
227.0
181.0
182.5
89.3
145.9
Mcrl
kipin.
156.3
76.5
74.8
101.6
62.7
52.2
124.9
87.6
63.3
26.6
44.6
156.3
69.6
62.7
83.0
47.1
42.1
93.3
76.9
47.1
21.3
38.7
148.4
71.8
68.6
105.7
62.7
55.4
124.9
99.6
63.3
29.0
51.7
156.3
76.5
74.8
105.7
62.7
55.4
124.9
99.6
63.3
29.0
51.7
156.3
76.5
74.8
105.7
62.7
55.4
124.9
99.6
63.3
29.0
51.7
156.3
76.5
74.8
105.7
62.7
55.4
124.9
99.6
63.3
29.0
50.9
153.4
71.1
65.7
85.0
49.9
43.3
99.0
76.9
50.1
22.1
38.9
Mcrd
kipin.
77.1
40.8
57.1
68.8
61.5
43.0
123.7
72.1
73.9
32.7
45.7
79.8
42.5
59.3
75.7
64.7
46.4
130.3
79.4
77.4
34.6
49.2
82.3
44.2
61.5
82.3
67.8
49.6
136.7
86.5
80.8
36.4
52.7
84.7
45.8
63.6
88.7
70.9
52.8
143.1
93.3
84.2
38.1
56.0
85.4
46.6
64.9
94.2
73.3
55.6
148.4
99.4
87.0
39.7
59.1
78.0
41.1
57.2
66.5
60.9
42.0
122.4
69.1
73.3
32.2
44.3
80.0
42.5
59.1
72.8
63.7
45.0
128.1
75.9
76.3
33.8
47.6
Mtest25
kipin.
35.1
24.4
41.1
73.2
62.3
55.9
155
134
108
58.1
98.7
36.7
26.3
44.4
74.8
67.4
57.1
159
137
105
53.8
99.9
39.3
26.5
44.6
75.6
68.3
57.5
162
136
114
56
100
39.5
26.5
44.9
76
68.7
57.6
163
136
111
56.9
100
39.4
25.4
44.7
76.2
65.4
57
NaN
136
NaN
59.8
100
36.5
25.3
41.3
71.7
60.8
55.2
158
133
108
57.4
97.6
38.4
26.2
44.1
74.3
66.5
56.8
158
136
104
55.2
99.4
Mtest75
kipin.
30.2
22.6
39.2
66.2
61.4
50.9
152
128
107
56.4
96.2
31.2
23.3
40.7
67.3
63.8
52.1
155
129
109
57.7
99.1
31.3
23.4
41.1
67.8
64.5
52.9
157
130
112
58.5
99.7
31.3
23.4
41.3
68
65.3
53.2
160
130
112
58.8
99.9
33.1
22.6
41.4
68.2
60.5
52.8
159
130
119
61.9
100
31.8
23.3
39.3
65.6
60.1
50.6
168
127
107
51.7
95.3
33.8
23.3
40.6
66.9
63.4
52.1
154
129
109
56.8
96.5
Study Type
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
ID #
SSMA section
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
550S162-33
600S162-33
1000S200-54
800S162-43
800S200-43
600S200-33
1200S200-54
1000S200-43
1000S250-43
800S137-33
800S162-33
800S200-33
550S162-33
600S162-33
1000S200-54
800S162-43
800S200-43
600S200-33
1200S200-54
1000S200-43
1000S250-43
800S137-33
800S162-33
800S200-33
550S162-33
600S162-33
1000S200-54
800S162-43
800S200-43
600S200-33
1200S200-54
1000S200-43
1000S250-43
800S137-33
800S162-33
800S200-33
550S162-33
600S162-33
1000S200-54
800S162-43
800S200-43
600S200-33
1200S200-54
1000S200-43
1000S250-43
800S137-33
800S162-33
800S200-33
L
in.
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
48
Lhole
in.
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
4.2
4.5
7.1
5.7
6.0
4.8
8.2
7.1
7.4
5.5
5.7
6.0
3.7
4.0
6.2
5.0
5.2
4.2
7.2
6.2
6.5
4.8
5.0
5.2
2.9
3.1
4.9
3.9
4.1
3.3
5.7
4.9
5.1
3.8
4.0
4.1
hhole
in.
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
4.2
4.5
7.1
5.7
6.0
4.8
8.2
7.1
7.4
5.5
5.7
6.0
3.7
4.0
6.2
5.0
5.2
4.2
7.2
6.2
6.5
4.8
5.0
5.2
2.9
3.1
4.9
3.9
4.1
3.3
5.7
4.9
5.1
3.8
4.0
4.1
# of holes
0
0
0
0
0
0
0
0
0
0
0
0
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
S
in.
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
Fy
ksi
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
58.6
530
My
kipin.
31.1
35.0
132.1
67.8
77.6
40.5
172.4
106.4
119.5
46.8
52.4
60.0
31.1
35.0
132.1
67.8
77.6
40.5
172.4
106.4
119.5
46.8
52.4
60.0
31.1
35.0
132.1
67.8
77.6
40.5
172.4
106.4
119.5
46.8
52.4
60.0
31.1
35.0
132.1
67.8
77.6
40.5
172.4
106.4
119.5
46.8
52.4
60.0
Mynet
kipin.
31.1
35.0
132.1
67.8
77.6
40.5
172.4
106.4
119.5
46.8
52.4
60.0
26.4
29.7
112.3
57.7
66.0
34.4
146.5
90.5
101.6
39.8
44.6
51.0
28.0
31.5
118.9
61.1
69.9
36.5
155.2
95.8
107.6
42.1
47.2
54.0
29.5
33.2
125.5
64.4
73.7
38.5
163.8
101.1
113.5
44.5
49.8
57.0
Mcrl
kipin.
17.5
16.8
63.2
32.2
35.9
18.4
58.1
32.3
35.2
13.3
14.7
16.3
17.5
16.8
63.2
32.2
35.9
18.4
58.1
32.3
35.2
13.3
14.7
16.3
17.5
16.8
63.2
32.2
35.9
18.4
58.1
32.3
35.2
13.3
14.7
16.3
16.1
15.0
50.0
25.6
30.3
17.8
44.3
25.6
29.3
10.4
12.0
14.2
Mcrd
kipin.
23.5
24.0
85.9
43.7
53.1
26.8
84.2
53.7
54.2
19.2
24.9
31.0
20.9
21.2
70.5
35.9
46.2
24.5
67.3
45.2
46.7
15.3
21.1
27.5
21.3
21.6
72.6
37.0
47.1
24.8
69.5
46.4
47.7
15.8
21.6
27.9
21.8
22.1
75.5
38.5
48.5
25.3
72.7
48.0
49.1
16.6
22.4
28.6
Mtest25
kipin.
22.2
22.9
94.4
48.7
54.2
25.7
115
67.6
66.4
26.3
31.5
33.8
19
20.2
81.9
41.7
45.6
22.2
100
57.7
61
25.5
27.6
29.4
20.6
21.6
88.8
43.8
48.6
NaN
108
62.2
63.4
27.1
29.3
30.1
22.3
22.4
92.4
46.2
49.9
23.6
113
64.5
64.4
28.2
30.9
31.7
Mtest75
kipin.
20.4
23.6
88.8
47.5
51.4
25.2
112
65.3
66.7
26.7
31.8
32.3
19
20.3
82
40.8
45.5
22.2
103
57.7
59.6
25.6
27.5
29.2
20.8
20.9
85
42.7
48
NaN
107
62.1
61.6
26.9
29.8
30.4
22.2
22.5
87.5
43.9
46.6
23.3
113
63.4
63
27.8
30.3
31
Study Type
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
Appendix M
Comparison of tested strengths to AISI S100-07
predicted strengths
The tested strengths of coldformed steel columns and beams with holes are
comparedtoAISIS10007MainSpecificationstrengthpredictionsinthisappendix.The
MainSpecificationtesttopredictedstatisticsareusefulasabaselinecomparisontothe
DSMapproachesdevelopedinChapter8.MainSpecificationtesttopredictedratiosare
obtainedfortheexperimentaldata(Table4.3andTable4.5forcolumns,Table4.11for
beams) and the simulated tests (Appendix K for columns, Appendix L for beams)
described in this report. Member strengths are calculated with custom Matlab code
written by the author. The Matlab code implements Section C4 of AISIS10007 to
predict the ultimate strength of columns with holes and Section C3 of AISIS10007 to
predictthestrengthoflaterallybracedbeams.
AISIS10007 considers two limit states for coldformed steel columns, (1) local
global buckling interaction (Section C4.1 of AISIS10007) and (2) distortional buckling
(Section C4.2 of AISIS10007). Column strength predictions for the localglobal
bucklinglimitstatearecalculatedwiththeequation:
Pn = Ae Fn
531
whereAeistheeffectiveareaandFnistheglobalcolumnstrength(stress).Aeincludes
theinfluenceoflocalbucklingattheholewiththeunstiffenedstripapproachfornon
circular holes (see Chapter 3 of this report) and employs empirical equations (Ortiz
Colberg 1981) for circular holes. Fn is determined with an empirical relationship
betweenglobalslenderness,c,tocolumnstrength,where
c =
Fy
Fe
andFyisthesteelyieldstress.Theinfluenceofholesonc,i.e.,theinfluenceofholeson
the critical global elastic buckling stress, Fe, is not taken into account in the Main
Specification.(ThisisafundamentaldifferenceoftheDSMapproachpresentedinthis
reportandtheMainSpecification.)Aeisindirectlycappedatthenetcrosssectionalarea
ofthesection,Anet,withgeometriclimitsintheMainSpecificationforbothcircularand
noncircular holes. The net section cap is consistent with the DSM approaches in this
report.Pn forthedistortionalbucklinglimitstateintheMainSpecificationiscalculated
with the AISIS10007 Appendix 1 DSM approach. Currently the Main Specification
does not provide a method to account for the influence of holes on the distortional
bucklingstrengthPnd.ThestrengthofthecolumnistakenastheminimumofPn(local
globalinteraction)andPnd(distortionalbuckling).
Forlaterallybracedbeams,AISIS10007considerstwolimitstates,(1)localbuckling
and(2)distortionalbuckling.Flexuralstrengthpredictionsforthelocalbucklinglimit
statearecalculatedwiththeequation:
532
M n = S e Fy
whereSeistheeffectivesectionmodulusderivedusingtheeffectivewidthmethod(see
AISIS10007SectionC3.1).AmethodisprovidedintheMainSpecificationtocapture
the influence of a hole in the web of a beam on the local buckling strength with the
unstiffened strip approach (see AISIS10007 Section B2.4). The distortional buckling
strength,Mnd,iscalculatedwiththeDSMequationsprovidedinAISIS10007Appendix
1. As in the case for compression members, the Main Specification does not currently
provideamethodtoaccountforthepresenceofholesonMnd.Thestrengthofabeamis
takenastheminimumofMn(localbuckling)andMnd(distortionalbuckling).
Table M.1 through Table M.2 summarize the testtopredicted statistics for the
columnexperimentsandsimulations.Thestatisticsarepresentedtoevaluateseparately
those specimens with holes that lie within code limits and the specimens outside the
codelimits(AISIS10007providesthegeometriclimitsforstiffenedelementswithholes
inSectionB2.3,forstiffenedelementwithholesunderastressgradientinSectionB2.4,
note that there are no code limits for distortional buckling controlled specimens with
holes).Themajorityoftheexperimentsarepredictedtohavealocalglobalinteraction
type failure (63 out of the 78 specimens). The testtopredicted mean is 1.20 for the
experiments within code limits and 1.06 for specimens outside the code limits,
suggestingthatcolumnstrengthissensitivetothegeometricparametersconsideredin
thecode.Thistrendisnotobservedinthesimulationtesttopredictedmean,wherethe
meanis1.04forspecimenswithincodelimitsand1.07forthespecimensoutsidecode
533
limits. The simulations tend to have larger, more closely spaced holes, resulting in a
testtopredicted mean of 0.91 for distortional buckling controlled specimens. It is
hypothesized that the specimens designated as distortional buckling failures may
indeed be failing by localglobal interaction if the Main Specification is overpredicting
the localglobal interaction strength. This low testtopredicted ratio highlights the
limitationsofthecurrentdesignapproachesinAISIS10007andconfirmsthattheDSM
approachesdevelopedinthisreport(seeTable8.1)aremoreviablepredictorsofcolumn
strengthwhenforholegeometriesoutsidethecurrentcodelimits.
TheMainSpecificationismostaccurateforstubcolumnsasshowninTableM.3,and
isbecomesmoreconservativeforintermediateandlongcolumns(i.e.,as cincreases)in
TableM.3.Theincreaslyconservativestrengthpredictiontrendwithincreasingglobal
slendernesscanbeobservedinFigureM.1.FigureM.2demonstratesasimilartrendfor
the simulated column tests, where the Main Specification prediction becomes
unconservative with decreasing c. The large distribution of distortional buckling
controlledspecimenslocalglobalbucklinginteractionandthecodeisjustmissingthe
behavior, especially because 212 of the 236 specimens are outside the Main
Specificationsgeometriclimits.FigureM.3throughFigureM.6demonstratethatlimits
imposedintheMainSpecificationonholesizeandholespacing,i.e.dh2.5inches,Lh4.5
inches,S24inches,andSend10inches,areweakindicatorsofpredictionviabilityand
that the dimensionless quantities studied in the element and member elastic buckling
534
studiesinChapter3andChapter4ofthisreport,,i.e.,S/Lhole,S/h,hhole/h,maybemore
viableasgeometricdesignlimits.
TableM.1AISIS10007testtopredictedstatisticsforallcolumndatawithholes
DataSource
Experiments
Simulations
DataRange
alldata
withincodelimits
outsidecodelimits
alldata
withincodelimits
outsidecodelimits
Mean
1.12
1.20
1.06
1.06
1.04
1.07
LocalGlobalBucklingInteraction
SD
#oftests
0.12
0.87
63
0.12
0.88
27
0.06
0.91
36
0.13
0.86
236
0.10
0.88
24
0.13
0.86
212
Mean
1.11
0.91*
SD
0.05
0.91
0.08
0.88
#oftests
15
149
*lowtesttopredictedratioresultsfromLGbucklingandDbucklingpredictionswhichdonotadequatelycapturetheinfluenceofthehole
ShadedareameansthatcodelimitsonholegeometryarenotprovidedforthedistortionalbucklinglimitstateinAISIS10007
TableM.2AISIS10007testtopredictedstatisticsforstubcolumnswithholes(c0.20)
DataSource
Experiments
DataRange
alldata
withincodelimits
outsidecodelimits
Mean
1.07
1.12
1.05
SD
#oftests
0.07
0.90
38
0.05
0.91
9
0.07
0.90
29
Mean
1.03
SD
#oftests
1
TableM.3AISIS10007testtopredictedstatisticsforintermediateandlongcolumnswithholes(c>0.20)
DataSource
Experiments
Simulations
DataRange
alldata
withincodelimits
outsidecodelimits
alldata
withincodelimits
outsidecodelimits
Mean
1.20
1.24
1.10
1.06
1.04
1.07
SD
#oftests
0.13
0.87
25
0.14
0.87
18
0.04
0.92
7
0.13
0.86
236
0.10
0.88
24
0.13
0.86
212
Mean
1.11
0.91*
ShadedareameansthatcodelimitsonholegeometryarenotprovidedforthedistortionalbucklinglimitstateinAISIS10007
535
SD
0.05
0.91
0.08
0.88
#oftests
14
149
2
AISI-S100-07 predicts local-global buckling limit state
AISI-S100-07 predicts dist. buckling limit state
1.8
1.6
1.4
Ptest/Pn
1.2
1
0.8
0.6
0.4
0.2
0
0.5
1.5
2.5
c =(Fy /Fe)0.5
FigureM.1AISIS10007testtopredictedratioforexperimentsoncolumnswithholes,predictionsbecome
increasinglyconservativewithincreasingglobalslenderness
2
1.8
1.6
1.4
Ptest/Pn
1.2
1
0.8
0.6
0.4
0.2
0
0.5
1.5
2.5
c =(Fy /Fe)0.5
FigureM.2AISIS10007testtopredictedratioforFEsimulatedtestsofcolumnswithholes
536
2
1.8
1.6
1.4
Ptest/Pn
1.2
1
0.8
0.6
0.4
0.2
0
4
dh, inches
FigureM.3AISIS10007testtopredictedratio,FEsimulatedtestsofcolumnswithholes,coderequiresthat
holedepthdh2.5inches
2
1.8
1.6
1.4
Ptest/Pn
1.2
1
0.8
0.6
0.4
0.2
0
4
Lh, inches
FigureM.4AISIS10007testtopredictedratio,FEsimulatedtestsofcolumnswithholes,coderequiresthat
holelengthLh4.5inches
537
2
1.8
1.6
1.4
Ptest/Pn
1.2
1
0.8
0.6
0.4
0.2
0
10
15
S, inches
20
25
30
FigureM.5AISIS10007testtopredictedratio,FEsimulatedtestsofcolumnswithholes,coderequireshole
spacingS24inches
2
1.8
1.6
1.4
Ptest/Pn
1.2
1
0.8
0.6
0.4
0.2
0
10
15
Send, inches
20
25
30
FigureM.6AISIS10007testtopredictedratio,FEsimulatedtestsofcolumnswithholes,coderequires
clearholespacingatcolumnendsSend10inches
The Main Specification testtopredicted comparison for the laterallybraced beam
experiments and simulated tests are summarized in Table M.4. The testtopredicted
538
meanislowforallexperiments,whichisconsistentwiththeDSMcomparisoninTable
8.5. It is still unclear what why the tested strengths are systematically lower than the
predictions,andaninvestigationisongoing.Conclusionscanstillbedrawnfromthese
results though. It is observed that the experiments with hole geometries within code
limits are predicted better than beams outside of code limits (mean testtopredicted
ratio of 0.90 versus 0.80), suggesting that the viability of the Main Specification
prediction method is related to the geometric limits. The simulated beam tests also
exhibit a low testtopredicted ratio of 0.92 for the localglobal buckling interaction
specimens,butinthiscaseforadifferentreasonthantheexperients.Itishypothesized
that the Main Specification is not as viable when the hole geometries are outside code
limits.WhencomparingtheresultsinTableM.4withthetesttopredictedstatisticsfor
theDSMapproachesinTable8.4,itbecomesevidentthatDSM(forbeamswithholes)is
amoreviablepredictorofstrengthoverawiderrangeofholegeometries.
TableM.4AISIS10007testtopredictedstatisticsforlaterallybracedbeamswithholes
DataSource
DataRange
alldata
Experiments** withincodelimits
outsidecodelimits
alldata
Simulations
withincodelimits
outsidecodelimits
Mean
0.83
0.90
0.80
0.92*
0.92*
SD
#oftests
0.12
0.84
64
0.16
0.80
21
0.07
0.88
43
0.07
0.89
20
0
0.07
0.89
20
Mean
0.90
1.07
SD
0.12
0.85
0.11
**ShanandLaBoubebeamstrengthsarelowcomparedtopredictionsevenformemberswithoutholes,Mean<1doesnotnecessarilysuggestaholeinfluence
539
0.87
#oftests
80
184

AISI DSMHoles Report Final R2

Transféré par

Informations du document

Description originale:

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

AISI DSMHoles Report Final R2

Transféré par

Droits d'auteur :

Formats disponibles

TheJohnsHopkinsUniversity

Conducted a preliminary investigation into the nonlinear solution algorithms

Completed experimental program including tensile coupon tests, elastic buckling

Papers from this research:

J Contact simulation in ABAQUS ................................................................................... 521

Figure 8.57 Test-to-predicted ratios for distortional buckling-controlled simulated laterally

buckling properties of a general coldformed steel crosssection are obtained from an

failure (high slenderness corresponds to high sensitivity, low slenderness to low

Pcre= Critical elastic global column buckling load

= gross area of the column

Global slenderness, c =(Py /P cre)0.5

Pcrl = Critical elastic local column buckling load

Nominal axial strength for global buckling

Local buckling interacts with

local slenderness, =(Pne/P cr )0.5

for d > 0.561 Pnd = 1 0.25 crd

Pcrd = Critical elastic distortional column buckling load

Distortional slenderness, d=(Py /Pcrd)0.5

Phase I research is primarily focused on elastic buckling. Chapter 2 describes

example, a stiffened element is a simplysupported plate used to model the web of a

A set of ABAQUS modeling guidelines is formalized in this chapter to provide a

2.1.1 Modeling accuracy for a stiffened element

Elastic buckling analyses of a stiffened element were performed in ABAQUS to

Plate buckling coefficients (k) are approximated in ABAQUS by performing

Comparing the number of elements required to model a buckled halfwave is a

plate buckling coefficient, k

2.1.2 Modeling accuracy for an unstiffened element

An unstiffened element is another common crosssection component considered in

Element aspect ratio is a:b

Node normal (typ.)

ABAQUS recommends 10 degrees

A minimum of two S9R5 elements per halfwavelength shall be provided in

Plate with holes

obtained with eigenbuckling analyses of plates in ABAQUS (ABAQUS 2007a). All

3.3.2 Influence of a single slotted hole

3.3.3 Influence of slotted hole spacing

fcr,holes /fcr,no holes

Buckling is dampened at the holes, halfwaves form between holes

fcr,holes /fcr,no holes

3.3.4 Approximate prediction method for use in design

f crl = min[ f cr , f crh ].

f crh ,net = min [ f crA , f crB ]

for Lhole hi < 1 ,

fcrh,net is the critical elastic buckling

f crh = f crh ,net

f crh = f crh ,net (1 hhole h ) .

Plate buckling controls

Unstiffened strip controls

Unstiffened strip controls

fcrl, ABAQUS/fcrl, predicted

fcrl, ABAQUS/fcrl, predicted

Plate buckling controls

Plate buckling controls

fcrl, ABAQUS/fcrl, predicted

fcrl, ABAQUS/fcrl, predicted

Unstiffened strip controls

fcrl, ABAQUS/fcrl, predicted

Plate buckling controls

Strips of plate adjacent to the

Plate buckling and

fcr , ABAQUS/fcr , predicted

Plate buckling controls

43 additional ABAQUS eigenbuckling analyses were performed to evaluate the