Vous êtes sur la page 1sur 12

7/31/2017

Stability of Portal Frames ~~~&:;:(-


- I
In-plane behaviour of Portal Frames
r~~~ II Shape of portal frame under load differs from the un-deformed shape.

----
I I I
Deflection causes axial loads in members to act along different lines from
I those assumed in the analysis
Buckling out of the plane of the frame is checked similar to

n
beam-columns between lateral restraints and torsional
restraints provided by bracings P-delta ,.
effects

Bracings make the effective lengths identifiable. ....
,;,
1

In plane, \
restraint to any column depends on stiffness of rafters and Vertical loads are eccentric to bases causing further deflection
other columns. Apex drops, reducing arching action
Applied moments curve members; axial compression in curved members
restraint to rafter depends on stiffness ofcolumns and increase curvature, reducing stiffness
other rafters
Frame is less stable (nearer collapse) than a first-order analysis suggests.

What are Second order effects? - 2 types Frame Stability


Effects of displacements of the intersections of members, usually called
P-l\ effects. Effect of in-plane deflections on frame
SLS EN 1993-1-1 describes this as the effect of deformed geometry.

If deflections are small, consequences are very small and


a first-order analysis (neglecting the effect of the

Tr-~1 t,.....-., I
I" I deflected shape) is sufficiently accurate.

If deflections are such that effects of the axial load on the


deflected shape cause significant additional moments
and further deflection, the frame is sensitive to second
order effects.
;
I
Second order effects, or P-delta effects, can reduce resistance
of frame.
Effects of deflections within the length of members, usually called P-O
effects.

1
l
7/31/2017

Consequence of P-6 and P-fl effects? What is Second-order analysis?


it reduces the stiffness of the frames and its elements to Methods in which effects of increasing deflection under increasing load is
below that calculated by first-order analysis. considered explicitly in the solution, so that the results include the P-/i and
P-L'>effects.

Single-storey portals are sensitive to the effects of the axial


compression forces in the rafters and columns. First-order or second-order analysis?
For either plastic analysis of frames, or elastic analysis of frames,
These axial forces are commonly of the order of 10% of the
elastic critical buckling loads of the rafters and columns,
around which level the reduction in effective stiffness
becomes important. ~. ",/

---- .. .--
J;
,
, Choice of first-order analysis or second-order analysis depends on the in
plane flexibility of the frame, characterised by the calculation of the ocr
factor- buckling amplification factor
n

How sensitive are P-delta effects? Di 1993-1-1 51.1 (4) B gives:

Limitations to the use of first-order analysis are defined in SLS EN


1993-1-1, Section 5.2.1 and the SLS National Annex Section
aq = l'
HEd
~

,q
h
C)H.Ed
1

NA.2.9 as: However, this can only be applied when the axial load ill the rafter is not
Significant. Note 28 of 5.2.J(4)B describes significant as when

For Elastic Analysis X;a O.3\rA/ T


winch mav be rearranzed to indicate that the axial load is not
\'
'Eol

- .
significant when 'VEd :SO.09.Vcr

Where:
For Plastic Analysis
N is the elasnc critical buckling load for the complete span ofrhe rafter
. .'EI
pair. r.e. ,Vtr = 7
where:
L is:the developed length of the rafter pair from COIUlllH (0 column.
Ocr is the buckling amplification factor taken liS span Cos 8 (8 is the roof slope).
Fcr is elastic critical buckling load for global instability mode, If tho! axial load HI the ratter exceeds this lunit. the expressrou 111 EN 1993-1-1
based on initial elastic stiffness cannot be used
Fed is the design load on the structure.

2
7/31/2017

a." factor
Expre ..
ron 'i ~ 01 N )09J-I-I ~ 511('"')8 ~I\t"':) {(cr~<'

If the axial load in the rafter exceeds tlus limit, tile expression 111 E 1993-1-1
canuor be used.
Xotc 1B and ;-':llt~:!B of that clause Iuuit the application of Expressrou
roof \\ nh suallow roof <;,1\)1~C;
and where the ~'(I~I force III the rafter An alternative expression, accounnng for tile axial force in the rafter. has been
stgmficam. 111lb
developed by J. Lilli and C Kingl61 and is detailed below.
a roof slope l~ considered as shallow at slopes 110 steeper thau :!6~
For frames with pitched rafters.
axial fOI;:"( III the rafter may be- assumed to be- significant if ~ ~ 0,\ :!:d .
acr,ffi = uun (aCf.s.~: a(fJ.~\I)
A couvcmcnt \\(1) IQ express the huutauou Oil the axial force b thdl the a.\.tal
force is not sigmficant If where:
NEd fO.09Ncr
is the estnuate of a" for sway buckling mode
WIlen:.'
is the estimate of a", for rafter snap-through buckling mode.
NCT i!> the Clll':)11C CI irical bucklina load for the complete Spall of the rafter
. . :r1EI - This mode need only be checked when there are three or more
patr.j c Ncr-~ spans. or if the rafter is horizontal. or when the columns are not
vertical.
is the developed length of ibe rafter pair from column to column.
takeu as span/Cos () (8 is the roof slope)
If the Iinuts Me satisfied. then Expression 5.1111<1)' be used to calculate acr In
most practical portal frames. the axial load in the rafter will be significnnr and
EXprl!SSIOIl 5 2 canner be IISed.

In plane stability
Snap through stability (Clause 554.3)
First order analysis (Ocr> 10)
This occurs in multi bay frames where the effects of
continuity could result in the use of slender rafters. In Amplified moments method (3 < Ocr < 10)
such frames the ridge could drop below the valley. Second order analysis (ocr < 3)

~
~
.... " - -.0--.::.-:.
::::'=-=-~ -J_ -~
r 'I, '.

f
I d
I :'
,
f<.. ,..... ~jr I

:, "
4.

r ,/
-

3
7/31/2017

Table 3.1 Second order effects: assessment and amplification factors


When 2nd order effects are significant, 2 options
Restrictions Elastic analysis Plastic analysis
are possible: Measure of shallow siopes, and au a",
3::; . ::;10 sensitivity to second
order effects
rafter axial force not
significant
steep slopes, and
acr.e~t acr.~
rafter axial force
Rigorous 2 nd order analysis
Approximate 2 order analysis nd significant

appropriate
(i.e. hand calculations using
(i.e. in practice, using an first-order
second order appropriate
analysis
allowance
with
for
Regular frames

C-~aJ L-~aJ or or

software) second order effects). Amplifier to allow to(


(l-V~aJ C-V~aJ
D- second order effects Irregular frames, but
excluding tied portals
C-l~aJ Of
( --- 1,1 ) or

1st J -I/a"
In 2 nd method, (modified order
analysis), applied actions are amplified
by a factor, to allow for second order
I
{_\ 1 (1- V~a,,,) ( 1.1 )
l-Vaa .

effects while using 1st order calculations. : Amplifier applied to: Horizontal loads All loads
only

n'
Equivalent Horizontal Forces
1mperfections
Equivalent Horizontal Forces (EHF) to allow for effects of initial sway
Frame imperfections are addressed in EN 1993-1-1 imperfections is allowed by CI. 5.3.2(7). Initial imperfections are given
CI. 5.3.2. Generally, frame imperfections must be I I by Expression 5.5, where the initial imperfection <jl (indicated as an
modelled. The frame may be modelled out-of-plumb, I I inclination from the vertical) is given as: <t> = <jlo o, am
or alternatively, a system of equivalent horizontal
forces (EHF) may be applied to the frame to allow where:
for imperfections. The use of EHF is recommended <jlo is the basic value: <jlo = 1/200
as the simpler approach.

It.....
System of equivalent forces replacing out-of-plumb h is height of structure in metres,

,1"" am =~O.5(1 + ~) for a portal m is number of columns in a single frame

V
tN , t
~~ For single span portal frames, h is the height of the column, and m = 2.

It is conservative to set o, = am = 1.

4
7/31/2017

1. Elastic analysis of a single bay portal frame


EHF may be calculated as <pmultiplied by the vertical reaction at the base of TIllS example covel'S the design ofa portal frame for ::I single-storey building.
the column (including crane loads as appropriate). The EHF are applied H<;;ingthe elastic method of global analysis. Only gravity loads are covered ill
horizontally, in the same direction, at the top of each column.

r.-v
tlns example. TIle frame uses hot rolled I sections for rafters and columns.
Svstern of equivalent forces replacing cot-ct-cfumt

2. Frame geometry
'i

CI 532(4) states that


Sway imperfections may be disregarded when HEd
~'n~
w !
-.:

~ 0.15 VEd.
~ ~
'---r-=---="3-~::~-=!.-~'~~~===-'=-='::",--C-
:
8~ I
~ I
~
I<--
---~-:o--

300ll
----==-==-'0--,

Spacing of portal frames = 7,2 Ul

Compare net total horizontal reaction at the base with net total vertical
reaction. In many cases, EHF may not be required in combinations of actions The cladding to the roof and walls is supported by purlins and side rails.
that include wind. However, EHF will need to be included in combinations of The purlins have been provisionally located at intervals of between 1500 nun
only gravity actions. and 1800 nun as shown, The side rails are provisionally located at intervals of
110 more than 2000 nun. The rafter and column verifications may require these
locations to be modified.

~ G-2.16ON.m.Hi1-wc

3. Loads ;::::I:::CL=::;~~;'-1
ii ~~i
30Z

i
~~~
13'5 1345

,
1700

:
~
1700

I
troo
g
1700 1700 rreo
_~ r-_~_~-:,...L.:-+-~-#-~->t""--"'!,;"",-
~~

I 1 I '
WOO

"
,1SS

,
3.1.
G
Permanent
= G~df.wc'ighf + Groof
loads
1 1
-- - ---- ~~~ G..elf-weight- self-weight of the beams
~ Groof: roofing with purlins Groof= 0.30 kN/m2
725 eoo l'il-==---=~
A. => for an internal frame: G,,,,,f~ 0,30 x 7 .20 ~ 2.16 kNfm
!

1
1475
Y
~ 3.2. Snow loads
6000
S27S
-soo I 7313

The characteristic value for 5110\\" loading 011 the roof for a specific location in
X a given country at certain altitude has been calculated as:
1900 I
y
Sk ~ 0.618 kNflll'
Y y
= 4.45 kN/1ll
f< 3020
:.k
I 11960
o-
,
=> for an internal frame: s = 0.618 x 7.20

3.3. Imposed load on roof


I 15000 "
Characteristic values for loading on the roof (type H: not accessible).
qk = 0.4 kN/m2
=) for all internal frame: (jk = 0,4 x 7.20 = 2.88 kN/m

5
7/31/2017

3.4. Load combinations 4. Preliminary sizing


ISing te-srorey STeel buildings. Parr ]. Concept design [11provides a table of
For simplicity, the wind actions are not considered in this example. preliminary member sizes. according to the rafter load find the height to
eaves.
Therefore, the critical design combination for choosing the member size is: Yo
IRafter load = 1.35( 2.16 + selfweight )+1.5 x 4.45 = 9.6 kl'J/m- self weight
G+~Q Say 10 k.''l/UlIO include self weight.

Where: ITIle section chosen for the rafter is an IPE 450. S355
The section chosen for The COlUIlUl is an IPE 500, 5355
Q is the maximum of the snow load and the imposed load,

-
T~a.1 MM1berlaetklf~~fnnte.ith6toolphch
ro= 1,35 (pennanent actions) e-....,..I ,.
JpMof ___

I ft.-r .
,. t ,.
rQ= 1,50 (variable actions) ,,
" ~~ 0,,,,,
1"- "
FE:l
P!:<l
PE:.a
1"""
"'l3O

"'" ."'.
.....
""lOI
"'''''
ee ee

i=" ..." I
The snow loads are greater than the imposed loads on the roof. therefore
Q = 4.45 kN/m ; .,
-,,,,
.,,""
.,"" ..I''''''
"'"." ....
"",. I
".,.,
"'''''
~e~~

.... .....
I~ . I " I "'.., ....."'" .------ e ec
""""
..~ I "' ee

I~ " I I "''''' I I .....,,'"..,.,.


I PE"~. tV

.
'" ..
"''''
...~ I
sa
zrc
Plr-
JIE:!33
PElJO

'1'''''
PEe
""
I~ .'"
I ., I .. I ..... q:.::o
>E,.,
H~ """
l>.:g \PE see ee ec
"" """"
'-- ~""

.1. Axial compression in the rafter


According: to the code, if the axial compression in the rafter is significant then
5. Buckling amplification factor lXcr ris not applicable. In such sirunnous. Appendix B of this document
'ecommends the use of ac;r.,,~ instead,
In order to evaluate the sensitivity ofthe frame to 2nd order effects. the
buckling amplification factor. a", has to be calculated. This calculation he axial compression is significant If J.;;:: O.3~ :~
requires the deflections of the frame to be known under a given load
combination.
r if NEd 2: 0,09 .Vcr. winch is all equivalent expression
An elastic analysis is performed to calculate the reactions under vertical loads '''Ed is the design axial load at LLS III the rafter. noted .VR,EdIII tfus example.
at ULS. which provides the following information:
Lcr is the developed length of the rafter pair Iiom column to COhUlUl
The vertical reaction at each base: VEd = 168 k:N

The horizontal reaction at each base: Hal = 116 k:N Lcr= ~ = 30.1m
cos S"
The maximum axial force in the rafters: NR.Ed = 130 k:N
.Va = ,T! Ell = )7 ~ x 210000'( 337-l0x 10-1 x 10-) = 77'2 k.N
La' (30.1x10)'

0.09 Na = 0.09 x 772 ~ 69 L'\

SJUd ~ 130 L'\ > 69 kN


the axial compresstou
Therefore III The rafter IS sigmficaut and aa from
8\ 1993)1 15 not applicable
Following the guidance [10m Appendix B. frame stabrhty is tls,~;)~ed based
en Cta-.d.l. U1 Sect lOll 5.2

L
7/31/2017

.1. Axial compression in the rafter 5.2. Calculation of ac.r.e51

According if the axial compression in the rafter is significant


10 the code. then For a pitched roof frame G",.<:'>t2 1Il1ll(G(Cr ...... aa.t.f.")
I~ 1'11101applicable. In such situations. Appendix B of this document act"'...
t only needs 10 be checked for porn I fnm,t'c; \11':t 0, urore ~Jl;lII<;

'ecouuueuds the use of acr. h! instead


When assessing frame stability. allowance can be made 101 the base suffness

I";~-'"1-~--
III 11mexample. a base stiffness equal to 1000 of the column snffue, .lme been
The axial compression is sianificant if;' ~03 i AI.. assumed to allow for the nominally pinned bases _ ~
, IX"
10 calculate a.".. a uouonal horizontal force IS applied tothe flame and the -1... ...,
I if NEd ~ 0.09 Noto which is all equivalent expression borizonral deflection of the top of the COIUIIUlS ,50 deteruuned under thisload,

"Ed lOSthe design axial load at ULS 1Il the rafter, noted NR.Ed in this example. The notional horizontal force IS ,

H-:-'"HF""~VEd= ..!.....>c168=O.S4k.l>.f .
La IS the developed length of the rafter pair froui COIUUUlIO column. , 200 200

Lcr=~ =30.1m
The houzontal deflection of the top or the COIUIIIII under t.!1I<; force l~ obtamed
c055 from the elastic analysis as 1.6 uun
The maximulIl "xi,,1 force in the rafters: SR.f.d = 130 kN
2
Ncr= 172.1z = ,7 x210000x33740xlO~ x 10-) =772 kN

l '~w) }{f--" ~
~o.sL-(
L,,2 (30,lxlO))2 a.". .s, c:u Not _ If: EI. _ -7:" 2100)< 33710 . IO~ .<10-1 _ 772 kN
J N~a ~ _000"",
0.09 Na ~ 0.09 x 772 ~ 69 Ie'l L,,' (,0.1.10')'
N"E<! ~ 130 kN > 69 kN
~08rl-r~)}{...!....
l .712
6000} -12.5
200 1.6
U.t..N,'\'.,. - 0.09)( 772 - 6Y \,J\

NR.LJ - IJU k.\I ....c.y k.\I


Therefore the axial compression in the rafter is significant and Ocr from
Thus acu;;t = Clc:r.s..esl- 12.5> 10
EK 1993-1-1 is nOI applicable.
First order elastic analysis may be used and second order effects do 1101 need
Following the guidance from Appendix B. frame stability is assessed based 10 be allowed fer.
allac:r.~, in Section 5.2.

6. Frame imperfections Applying equivalent horizontal forces is the preferred option and the method
that is used in this worked example. The equivalent horizontal forces are
The global initial sway imperfection may be determined from
calculated as:
= tA> au am
HFJfF = VEd
tA> = 11200
However sway imperfections may be disregarded where HEd ~ 0,15 VEd.
= ~=-2....=0.82
au
,I" '/6.0 Table I shows tile total reactions for the structure to determine HEd and VEd.

Table 1 Vertical and horizontal reactions


f 1
au> = JO,5(1+~) =0.87 10_5(1+-) =0.87 Right column Total reaction 0,15 VEd
, 2 Left column (kN)
(kN) (kN) (kN)
m = 2 (number of COIUllUlS) ' HEd VEd HEd VEd HEd VEd

Reactions 116 168 -116 168 0 336 50


= _1-xO,S2xO,87=3.56xlO-3
200
Initial sway imperfections may be considered in two ways: HEd = 0 < 0,15 VEd

By modeling the frame out of plumb Therefore the initial sway imperfections have to be taken into account.
By applying equivalent horizontal forces (EHF).

7
7/31/2017

Frame geometry
IThe equivalent horizontal forces:
The propo.sed trame is defined ~s shown below.

EHF = VEd.colurnn= 3,56 x 10-3 x 168 = 0,60 kN The nodal heiCht of the columns i& estabilaMd from !I requir&me<'lt for 12 m clear hei:nt

betweert the fw1ished rlCXlf level (ffI) and the undersoc!e of the haunch. The base IS assumed

IThis force is applied at the top of each column, in combination with the to be 300 mm below m. The rafter is 8$SUmed to be 500 mm deep, and the ceieoce from

ennanenr and variable actions. the underside of the haunch to the node is assumed to be 1.5)( ratter dep(h.

Nodal heil/lt - 300 + 12000 + 1.5 500 13050. say 13100 mm


or the ULS analysis, the bases are modeled as pinned. Otherwise the base The haunch length is nsumed to be 10".Jf. of the span.

details and foundation would need to be designed for the resulting moment. Rafter: 533 " 210 x 101 UKB, Grade S355
The portal frames are spaced lit 8 m centres.
Column: 686 x 254 x 125 UKB, Grade S355
'The following figure shows the internal forces on the frame subject to the
ULS loads including the equivalent horizontal forces. ...."'"
~.]m
Aiw" I.5II

1+->1

J".
~~.1~1oN
Nto-1JOIIN '1'~.111 ilN V,._lPk'" z;::~~~::~:...
VIli:O.l~~ VT.,SOtN =fF
MI~.MW"" Z;::~~Nm~''::~~ N.-tl-l
M ~o_
lie. ~ 111"""

'[';:~2~:~~~~~
v. -HelIN t , Nr OIo,.",,",

=~~,,-'-'~:~:~::-,:~
;-:';:~:::"
Z::;:;~I:~\
. ~~~o~. ):,-

~<f/el
~
\
~ - ~ ~ 1'7k."l
M,".~k.'*t1
111-1010,.0{
'-':1I;;-11e~N
101 " l:>1 k.Nm
~";",Jl.1
<, -, SQ11- -...

I
i y"" .."....
f \...-/'4t.o~ ~ +;

f /'1". f The cladd;nt to the roof and walls is supported by purl;n, imd side rails. The specina:;
of the rails and purlins, and tl'le locations of restraints to the ;n!olde "8l1ges, will be

defined at the design stage.

2.4 Wind actions


2 Actions
The peak velocity pressure qp has b~en calculated for the chosen location. and
2.1 Permanent actions internal and external pressure coefficients, all determined from 8S EN 1991-1-4
and the UK National Annex.
The characteristic value of load due to self weight is given by:

Two load cases have been adopted as shown in Figure 0.5 and Figure 0.6 to determine
gl: = g lf !iII + groor (in Figure 0.5) the most onerous uplift condition and (in Figure 0.6) the most onerous
wind loading to be used in combination with gravity loads. Figure 0.5 and Figure 0.6
gH'll'''~PI is the self weight of the rafters show the characteristic values of loading on each frame due to wind pressures and
g.ooC is the self weight of the roofing with purlins and services. taken as 0.55 kN/m2 represent the situation mid-way along the building.

Therefore for an internal frame: ~


fn'l~~tllt
r-- 10..'1 ~~
g,." = 0.55 x 8.0 = 4.4 kNlm

""'~::
.192 ~ !!!.l.~!l...!)-!!!J!.'"~
.I.~l

I ~~

f '*
ri'n 1 "3'
~
~I
~
~~"
~ Load'Q/'o'InlrlWm ..

~
:: r ::,
/itr,

8
7/31/2017

C()fI,!BINATION PERMANENT VARIABLE ACTtaNS


ACT}()NS
lEADING ACTION ACCOMPANYING ACTION
PAFHlAl ACTION PAJITl ACTION PARTIAL COOBINATK>N
FACTOR, FACTOR. FACTOR, FACTOR, Y'D
3 Combinations of actions
Permanent
'0 Yo Yo

1 1.35 Imposed 1.5


&. imposed
According to 8S EN 19911. imposed actions on a roof are not considered with either
Perr"'lanent,
2 1.35 Snow 1.5 W;od 1.5 05
snow & wind
the wind actions or the snow load.
3
P~-, 1.35 Wind 1.5 Snow 1.5 as
wmd&STIO'N'
The critical combinations of actions for preliminary sizing are likely to be from:
4
Po,"""",, 1.0 Wood 1.5
&
1. Permanent actions with imposed roof loads, or

2 Permanent actions with snow loads. or


3 Permanent actions wnh snow loads and wind actions. AXlALFORCE
BASE REACTIONS (kN)
IN RAFTER (kN)
COMBINATION
As the imposed roof loads are greater than the snow loads. combination 2 can be
F,'J.. F,';>. FfL). F N"
ignored. Combination 3 must be considered with both the wind actions and the snow
1 241 241 74 -74 97
load as the leading variable action. in turn.
2 '227 227 91 -81 70
For preliminary sizing, it is assumed that the combination considering wind uplift will 3 213 213 108 -108 43
not be critical. 4 -5 91 -93 52 46

The results of the initial analysis are shown in Table 0.2. The base reactions at the left
and right column in Combination 4 (subscript land R) differ due to the asymmetric
wind actions included in that combination.

6 Sensitivity to second order effects 6.1 Axial compression In the rafter

In order to evaluate the sensitivity of the frame to effects of deformed geometry, the The axial compression is significant if ;. ~ OJ
vN.:w" which may be rearranged to show

factor an has to be calculated. If a.is greater than 10. second-order effects are small
that the compression is significant if NUi;;>-. 0.09Ncr'
enough to be ignored. Hand calculation methods to determine au are generally more
conservative than software solutions. N~ is the design value of the compression force in the rafter at UlS.

BS EN 1993-1-1 offers a simple approximation to calculate ao' but this can only be used
when the roof slope is less than 260 and the axial force in the rafter is not significant-

If the axial compression in the rafter is significant, Section 6.6 of this publication
Lu is taken as the developed length of the rafter pair between columns.
describes an alternative approach to estimate an'
Hence. L = ~ = 30.165 m
cr cos e"

n'EJ
n' x210000x6150~xlO' x 10-' = 1401 k.'1
= L(Xl r
(30.165xlO') 0.09N. = 0.09 x 1401 = 126 kN

N", =97kN.< 126kN

Therefore the axial compression in the rafter is not significant and Expression 5.2 of
BS EN 1993-1-1 may be used to calculate a.

9
7/31/2017

COfroIBINATION PERMANENT
VARIABLE ACTIONS
ACTIONS

lEADING ACTION ACCOhlPA.NYlNG ACTION

PARTIAL ACTION PARTIAl ACTION PARTIAL COtAerNATK)N


FACTOR. FACTOR. FACTOR, FACTOR. '1',
3 Combinations of actions Y. Y, Y,
1 Permanent
A imposed I." Imp<>"'" 1.'
According to 8S EN 1991-1. imposed actions on a roof are not considered with either
2 Penr.anent, Snow
1.3S 1.' Wind 1.' O.S
anow & wind
the wind actions or the snow load.
3 Permanant. 1.5
wind & snow 1." Wind 1.' Snow 0.5
The critical combinations of actions for preliminary sizing are likely to be from:
-4 Permanent 1.0 Wind 1.5
& wind
1. Permanent actions with imposed roof loads, or

2. Permanent actions with snow loads. or


3. Permanent actions with snow loads and wind actions, AXIAL FORCE
BASE REACTIONS (kN)
IN RAFTER (kN)
COMBINATION
As the imposed roof loads are greater than the snow loads, combination 2 can be
r; F\'}t F.. Fu NEd
ignored. Combination 3 must be considered with both the wind actions and the snow
1 241 241 74 -74 97
load as the leading variable action. in turn. -81
2 '227 227 91 70
For preliminary sizing. it is assumed that the combination considering wind uplift will 3 213 213 108 -108 43

not be critical. 4 -5 91 -83 52 46

The results of the initial analysis are shown in Table 0.2. The base reactions at the left
and right column in Combination 4 (subscript Land R) differ due to the asymmetric
wind actions included in that combination.

6 Sensitivity to second order effects 6.1 Axial compression In the rafter

In order to evaluate the sensitivity of the frame to effects of deformed geometry. the The axial compression is significant if ;. z OJ W, , which may be rearranged to show

factor acr has to be calculated. If acr is greater than 10. second-order effects are small
vN::
that the compression is Significant if N'fA ~ O.09Na'
enough to be ignored. Hand calculation methods to determine a.are &!enerally more

conservative than software solutions. NEd is the design value of the compression force in the rafter at ULS.

BS EN 1993-1-1 offers a simple approximation to calculate aa' but this can only be used _ yt2/y

when the roof slope is less than 26 and the axial force In the rafter is not significant . - L.'
If the axial compression in the rafter is significant. Section 6.6 ofthis publication
L", is taken as the developed length of the rafter pair between columns.
describes an alternative approach to estimate acor
Hence.L = ~ =30.165m
cr cos e"

. _ n'El,
- Lo'
0' x21ooo0x61500xlO'
(30.165xI0')'
x 10-' = 1401 k.'<
0.09,va =0.09 X 1401 = 126 kN

,v.. = 97 kN. < 126 kN

Therefore the axial compression in the rafter is not significant and Expression 5.2 of
BS EN 1993-1-1 may be used to calculate a.

9
7/31/2017

Combination 1

In combination 1, the notional horizontal force at the top of each column is:

6.2 Calculation of aa H_ =~F =~x241 =1.205kN


200 v 200
The expression for aD" offered in Section 5.2 of BS EN 1993-1-1 is not immediately
applicable to portal frames. as the vertical loads dominate the calculated horizontal Following the guidance given in Section 6.4.2 for nominally pinned bases, a base

deflections at the eaves. In Section 6.5 of this publication it is recommended that for stiffness equal to 10% of the column stiffness has been assumed in the analysis.

portal frames. aa should be calculated from the expression: In this example. the nominally pinned base has been modelled using an additional
member. with an inertia set as 10% of the column member. and a length equal to 75%

a. of the column length. The end of the dummy member is pinned.


200XDNlW
The horizontal deflection of the top of the column under this force. allowing for nominal

,t~:J
base stiffness, is obtained from an elastic analysis as 7.2 mm.
where:

h is the height to the eevee


is the horizontal deflection at the eaves. under a notional horizontal force applied
""'"
at each eaves node. equal to 1/200 of the factored vertical base reaction.

7 Frame imperfections
!The global initial sway imperfection may be determined from:

:
200XDNur
~.

ilt
= ,a"a.
= 1/200
= 13.1 m (height to eaves)
III = 2 (number of columns)
=~=9.l 2 2
200x7.2 a" = iil =7tTi=O.55 but a" > 213. hence cr,= 0.67

Because au = 9.1 < 10, second order effects cannot be ignored.


am =
r--T
10.5(1 +-)
111
=
R 0.5(1 +-)
2
= 0.87
Second order effects may be allowed for by amplifying the results of the first order
analysis by the factor: = ~xO.67xO.87 =2.9xIO-1
200

Initial sway imperfections may be conveniently included in the analysis by applying


1
1 = 1 12 equivalent horizontal forces (EHF).
= -I
I-~ 1-9.l Sway imperfections may be disregarded where HEd ~ O.15VEd
a"
Table D.4 shews the total reactions for the structure to determine Hu and VE4
for Combination 1.

iH", = 0 kl'i and 0.15V" = 72.3 kN

10
7/31/2017

If AnalysIs results
LEFT COLUMN RIGHT COLUMN TOTAL REACTION O.lSV" For the UlS analysis. the bases are modelled as truly pinned. Otherwise. the bases and
(kN) (kN) (kN) (kN)
foundations would need to be designed for the resulting moment.

Fa t; Fa r; H v 8.1 Final analysis results


For the load combinations In which Q'cr is less than 10. second-order effects need to
Reactions 74 241 -74 241 482 72.3
be allowed for. In this example the amplified moment method has been used: the

amplification factors are shown in Tebje 0.5.

Because Hf4 < 0.l5H4 the initial sway imperfections need to be taken into account
COMBINATION COMBINATION COMBINATION COMBINATION
in Combination 1.
1 2 3 4

The equivalent horizontal forces are taken as a proportion of the design base
Calculated Clcr (l'cr=9.1 a.= 9.8 acr= lOA acr = 46.8
vertical reactions:
Amplification 1.12 1.11 does not apply does not apply
factor

H"" = V = 2.9 X 10-' x 241 = 0.7 kN

'The following diagrams show the results of the frame analysis for Combinations 1 and 2
This force is applied horizontally at the top of each column. in the same direction. including amplification and equivalent horizontal forces.

in combination with the permanent and variable actions.


Figure 0.10 shows the bending moment diagram for Combination 3. including the EHF.

It can be demonstrated that equivalent horizontal forces must also be included in Figure 0.11 shows the bending moment diagram for Combination 4. Note that there

is no need to include the EHF in Combination 4. and second order effects are small
Combinations 2 and 3. In Combination 4. initial sway imperfections may be disregarded.
enough to be ignored.

RilIft:
axt~ll09kN
ShN,18J kN

FIgure D.8
Design bentJin,g
moment diagram
for Combination 1
illduaing EHF

11

Vous aimerez peut-être aussi