Internatiqnal Journal of Fracture Mechanics, Vol. 8, No.

3, September 1972 Wolters-Noordhoff Publishing- Groningen Printed in the Netherlands

267

Plane Elastostatic Analysis of V-Notched Plates

B. G R O S S A N D A. M E N D E L S O N
Lewis Research Center, National Aeronautics and Space Administration, Cleveland, Ohio
(Received February 19, 1971 ; in revised form July 9, 1971)

ABSTRACT Solutions are given for several plane elastostatic problems of plates having a V-notch on one edge, and subjected to a variety of boundary conditions. The effect of the magnitude of the V-notch angle and specimen geometry on stress intensity factors K l and K H are obtained for unloaded notch surfaces. There is less than one percent difference in opening mode I stress intensity factor in going from a zero degree notch angle to a 30 degree notch angle. Notch opening displacements at the plate edge were measured experimentally and the results obtained were in excellent agreement with the computed results.

Symbols The following list contains the commonly used symbols and their definitions. In general, all symbols are defined when introduced. coefficients of the stress function series associated with the n th eigen value Ai,n specimen dimensions as shown in figures 1, 2 and 3 a ,b, h, s, w modulus of elasticity E K, KI, KII stress intensity factor at the notch tip, subscript I, II refers to mode of notch extension bending moment M applied load P displacement in y-direction Uy 2 . = q . + i ~ . n th complex eigen value Poissons ratio v equivalent stress (7e components of the stress tensor O'r, O'q~, Trq~ stress function Z 1. Introduction For evaluating the resistance to quasi-brittle fracture of structural materials, many different types of specimens have been used in fracture toughness tests. However, for several reasons connected with optimization of specimen size and load requirements [1] single edge crack plate specimens loaded in tension or bending are of current interest from the standpoint of fracture test method development [2]. An analytical solution for a finite width single edge crack specimen loaded in tension was given by Gross, Srawley and Brown [3], using boundary collocation techniques applied to an appropriate stress function derived by Brahtz [4] and Westergaard [5] and independently by Williams [6, 7]. Solutions for other finite width edge cracked specimens subject to bending or combinations of tension and bending were subsequently obtained by Gross and Srawley [8]. The results are in good agreement with those obtained experimentally [9-12] and analytically by other methods [13-15], as shown in detail in reference 16. The specimens recommended by ASTM Committee E-24 for plane strain fracture toughness tests [2] have machined notches tipped with fatigue cracks, Fig. l(b), which are intended to represent ideal plane cracks, Fig. 1(a). For practical reasons it is desirable to allow considerable
Int. Journ. of Fracture Mech., 8 (1972) 267-276

268

B. Gross, A. Mendelson

latitude in the form of the machined notch; the only important consideration is that the cracktip stress field should not be significantly different from that of the ideal plane crack which is represented. To establish how much latitude in crack-notch configuration can be allowed one must investigate the stress fields of V-notches of various angles. Any crack-notch configuration can be just bounded by a V-notch of some angle which has its apex at the crack tip, Fig. 1 (b). It is assumed that the deviation of the stress field of that V-notch from the stress field of a zeroangle notch (plane crack) is an upper bound on the deviation of the crack-notch configuration. Thus the V-notch angle is useful as a single parameter for characterizing practical crack-notch configurations in relation to ideal crack-tip stress fields. The present work is concerned with the analysis of V-notched plates as a function of notch angle.

(a) Ideal plane crack.

...../"t'~

crack

(b) Machined notch.

Figure 1. In practical specimensthe 1dealplane crack (a) is simulated by a machined notch with a fatigue crack extension, (b) The dashed lines represent a sharp, finite-angle notch which just includes the machined notch and fatigue crack extension. The first attempt at such an analysis was made in [18] for a restricted geometry, using the collocation procedures of Gross and Srawley [3, 8]. The present analysis gives the general solution over a broad range of geometries and in particular determines the influence of the notch angle on fracture toughness measurements. The elastic stress solution of a homogeneous isotropic rectangular plate of finite dimensions with a V-notch loaded by either symmetric or antisymmetric loads is given.

2. Analysis
2.1. The Plane Problem (Modes I and II) The solution presented in this paper is for a specimen having a zero radius V-notch on one edge. It applies to a homogeneous isotropic material and is in the class of plane elastostatic problems of the theory of linear elasticity. Plates containing a single edge V-notch are analyzed for symmetric loading (mode I) and antisymmetric loading (mode II). The configurations are shown in figures 2 and 3 respectively. We take the origin of coordinates at the tip of the notch as shown in figures 2 and 3. Neglecting body forces, the differential equations of the system for both plane stress and plane strain reduce to the biharmonic equation (0 2 1 0 1 ~ 2 _ ' ~ ~02 Z I~Z 1 ~zZ'~= 0

Assume a complex stress function )~ [6] whose real part satisfies Eq. (1)
Int. Journ. of FractureMech., $ (1972)267-276

Plane elastostatic analysis of V-notched plates

269

~_ D

2Yob

o Pure tension
loading P

N Pure bending lhree pointbending

loading loading (slw4) :

BI
1;

t
w -,

Doublecantileverbeam
specimen. Figure 2. Mode I specimens having a thickness b.
B C

~l~xjL_

j
(2)

w~l

Figure 3. Singleedgenotchedplate sp~cimerisubjectto antisymmetricloading(modeII), havinga thicknessb. (a) Dimensionlessequivalentstress contours for plane stress conditions. (b) Dimensionlessequivalentstresscontourstbr plane strain conditions. z(r, ~, 2 ) = rZF(2, (o) where F(A, ~b)= [al cos 2~b+a2 cos (2-2)q~+aa sin 2q5 + a4 sin(2-2)~b]. Z satisfies identically the conditions of zero notch surface boundary loads with 2, al, a2, a3 and a being complex numbers. Satisfying the homogeneous boundary conditions leads to

and

sin 2 c/2 ( 2 - 2 ) sin (2-2)a/2J

A2

(4)

cos 2 e/2

( 2 - 2 ) cos(2-2)cq2J

where A i = 2ai. The resulting eigenequations are sin ( 2 . - 1)a = _+(2.- 1)sin a . (5)

Int. Journ. of Fracture Mech., 8 (1972)267-276

270 where 2, = t/, + i~, n = 1, 2, 3.... , oo are the complex eigen values. It also follows from Eq. (4) that
-

B. Gross, A. Mendelson

At,= '

cos (2.- 2)e/2

COS 2,e/2

A2,,A3, = ' "

sin ( 2 , - 2)e/2 sin 2, e/2 A4,,.

The stress function Z can be written separating the even functions and the odd functions as follows :
Z (/', (~, "~) ~" Zeven 71- Zodd

where
Zeven = n=l

cos 2.q + &,. cos(2.- 2)4,]

(6)

and
Zodd =

rZ"[As,, sin 2,~b+A4, . sin(2,-2)qb].

n=l

Since the stresses and displacements are real, they are associated with the real parts of the stress function ~. The displacement solution is described in detail in [ 16] and [-17]. The mode I and II stress and displacement fields are associated with the even and odd solutions respectively. The corresponding stress intensity factors K~ and Ku are obtained as follows. KI = (2=) ~ lim r 2-"*%(r, 0)
r--+0

t [

(7)

K u = (2re)} lira r2-nllzro(r, 0)1 ]

r-e-0

The stress intensity factors are thus defined to cancel the singularity. As originally used by Irwin [19] the stress intensity factor K is always associated with a zero notch angle (crack). This leads to a square root singularity, i.e., th =-32. When the .notch angle is no longer zero degrees (crack) but less than 180 degrees, the singularity occurring at the notch tip is less than , (th >-32as shown in table 7). Hence 1 I ( sinthesine ~ ] K I = (2r0~rh(rh-1) - c o s e + l + c o s ~ - - / + 1 ReA~,I ~-s t/~ ~ I K n - (2r0~ i ( , 1 - 1 ) l_("1- 2) + q l \(sini----the sine

cos cz)]ReA4, t

where ql is the real part of 2t. For the limiting case of a crack (a= 360), the previous equations reduce to those obtained in [7]. Problems involving non-homogeneous boundary conditions along the crack surface can be handled by introducing a particular stress function as described in detail in [16].
3. Boundary Collocation Solution

The previous solutions satisfy exactly the boundary conditions along the notch surfaces. To satisfy the boundary conditions along the rest of the boundary, one must find appropriate values for the unknown coefficients appearing in the series obtained. This can be done by truncating the series as expressed in Eq. (6) and determining the unknown constants such that the stress function and stress function derivative satisfy the boundary conditions at a finite n u m -

Int. Journ. of Fracture Mech., 8 (1972)267-276

Plane elastostatic analysis of V-notched plates

271

ber m, of selected boundary stations. In doing this, a set of 2m simultaneous algebraic equations is obtained and the first 2m coefficients of the stress function series are determined. The truncated stress function series thus obtained is considered to be an accurate representation of the actual stress function when further increase in the number of boundary stations produce no significant change in the first coefficient of the stress function series which represents the dominant term in the vicinity of the crack tip. This boundary collocation technique is described in detail in references 3, 8, and 16. For included vertex angles e, above 300 the above method worked very well. However, as is decreased below 300 the first coefficient of the stress function oscillated about some value, bracketing this value without actually converging. This difficulty was overcome in a manner similar to that used by Hulbert [20]. More boundary stations were chosen than unknown coefficients leading to an overdetermined system and the resulting set of simultaneous equations was satisfied in the least squares sense. All of the results reported herein are based on the solution to such overdetermined systems. For included angles greater than 300 degrees the results of overdetermined systems were the same as for the ordinary system consisting of an equal number of equations and unknowns. Those early results referred to in publications by Gross et al. [3 and 8] did not use overdetermined systems as this was unnecessary. In solving the resulting set of n simultaneous equations in n unknowns, a Gauss-Jordan pivotal condensation routine was used. Since the most important quantity to be determined is the stress intensity factor K, preliminary trials were made to determine the number of equations necessary for evaluating K with sufficient accuracy for the case of pure tension, pure bending and three point loading. Over the range of geometries analyzed it was indicated that an overdetermined system of 52 equations and 40 unknowns were sufficient. For the double cantilever beam antisymmetrically loaded, it was found that a set of 64 equations and 40 unknowns were sufficient.

4. Experimental Measurement of Displacement

For comparison with the analytical results, accurate crack surface displacement measurements were made on several V-notched plate specimens of 7075T651 bare aluminum, subjected to three point bending as shown schematically in figure 6. Measurement points were symmetrically located on opposite sides of the edges of the V-notch. The bend specimens were 9-inches long by 2-inches wide by 1-inch thick with a span length-to-width ratio of 4 to 1. Nine specimens were tested corresponding to e values of 330, 300 and 270 degrees, and a/w ratios of 0.4, 0.5, and 0.6. The modulus of elasticity of the aluminum was taken as 10.4 x 106 psi. The equipment used and the experimental procedures are described in detail in [16].

5. Results and Discussion

Preliminary results indicated that for pure tension, pure bending and three point increasing the height to width ratio h/w beyond 1.2 (Fig. 1) f o r , > 300, and beyond 1.4 for 300 > ~>240, produced no change in the stress intensity factor. These h/w values were therefore used in all the subsequent calculations. Results are presented in tables 1 through 6. The value of the first term of the truncated stress function Re (A2.1) or Re (A4,1) is directly proportional to the stress intensity factor KI or K~I.Tables 1 through 5 contain the nondimensional stress intensity factors for mode I and mode II. For all mode I cases considered the value of K increased a maximum of 1 percent in going from a = 360 (crack) to 330. The square root singularity associated with the 360 crack changed from 0.5000 to 0.4996 in going from 360 (crack) to 330 degrees. One can conclude on examining the results of tables 1-4 that as long a s , is 330 or greater the difference in the stressintensity factor KI from that for a crack is very small. Table 5 contains the nondimensional mode II stress intensity factors. Clearly, small changes in ~ produce large changes in K n.
Int. Journ. of Fracture Mech., 8 (1972) 267-276

272 TABLE 1

B. Gross, A. M e n d e l s o n

Dimensionless stress intensity factors for a single edge notched specimen in pure tension
Included vertex angle c~ Notch length to plate width ratio, a/w 0.2 0.3 0.4 0.5 0.6 0.7

Ktbw~r ~ p /
360 330 300 270 240 1.085 1.097 1.169 1.366 1.804 1.614 t.630 1.724 1.978 2.593 2.369 2.389 2.520 2.888 3.766 3.539 3.569 3.756 4.297 5.630 5.537 5.579 5.859 6.736 8.934 9.422 9.491 9.979 11.515 15.551

TABLE 2

Dimensionless stress intensity factors for a single edge notched specimen in pure bending
Included vertex angle c~ Notch length to plate width ratio, a/w 0.2 0.3 0.4 0.5 0.6 0.7

K~bw~1/6M
360 330 300 270 240 0.837 0.844 0.895 1.034 1.346 1.093 1.100 1.155 1.314 1.680 1.414 1.422 1.484 1.678 2.146 1.877 1.885 1.965 2.218 2.860 2.629 2.640 2.752 3.139 4.112 4.041 4.062 4.250 4.873 6.532

TABLE3

Dimensi~n~essstressintenszty~ct~rs~rasinglee~en~tched~ec~ensu~ectedt~threep~intbendi~l~adi~
Included vertex angle c~ Notch length to plate width ratio a/w 0.2 0.3 0.4 0.5 0.6 0.7

Kxbw~r ' /6P

360 330 300 270 240 0.777 0.785 0.832 0.963 1.261 1.013 1.021 1.072 1.226 1.573 1.318 1.325 1.390 1.578 2.034 1.767 1.775 1.857 2.118 2.770 2.504 2.517 2.636 3.012 3.984 3.893 3.922 4.112 4.736 6.395

Stress ratio,

%1o
2.5

.160

16o

12o
~

OelO Z5

4.0

~ . 080
P, .040 2 0

~" ~.
~

i 4. 0

.040

0 -. 200 -. 160 -. 120 -. 080 -. 040 0 .040 . 080 . 120 . 160 Location, x/a Ib) Dimensionless equivalent stress conteurs for plane strain conditions.

(a) Dimensionless equivalent stress contours for plane stress conditions.

Fagure 4. Edge notched beam subjected to three point loading. Included vertex angle, 300 ; Poissons ratio, 0.30 ; notch length to plate width'ratio, 0.50; a = 6P/bw.

Int. Journ. of Fracture Mech., 8 (1972) 267-276

Plane elastostatic analysis of V-notched plates

TABLE 4 '~

273

Dimensionless stress intensity factors for an edge notched rectangular double cantilever beam subject to spIittino forces
Included vertex angle c~ M a x i m u m beam depth to notch depth ratio Notch depth to plate width ratio d/w 0.667 0.500 0.400 0.333 0.250 0.222 0.200

h/a
Klbh/Pa 2-~1
360 0.5 1.0 1.5 2.0 0.5 1.0 1.5 2.0 1.0 1.5 2.0 1.5 2.0 8.320 14.829 22.209 29.623 8.383 14.904 22.320 29.775 15.509 23.263 31.040 26.333 35.139 6.565 7.202 9.896 13.039 6 639 7.230 9.928 13.085 7.426 10.253 13.541 11.300 14.984 6.560 5.810 6.033 6.958 6.636 5.835 6.040 6.956 5.964 6.168 7.122 6.509 7.654 6.485 5.798 5.695 5.913 6.641 5.822 5.705 5.904 5.947 5.781 6.071 6.707 6.307 6.587 5.798 5.678 5.730 5.822 5.689 5.787 5.961 5.767 5.881

330

300

270

7.741 9.732

TABLE 5

Dimensionless stress intenstty factors for a single edge notched specimen subjected to antisymmetric loading, mode II
[ncluded vertex angle e M a x i m u m beam depth to notch depth ratio Notch depth to plate width ratio a/w 0.400 0.333 0.286 0.200

h/a
Kllbh/Pa 2-~1
360 0.5 1.0 1.5 0.5 1.0 1.5 1.0 1.5 1.0 1.5 0.353 0.500 0.614 0.186 0.401 0.543 0.278 0.456 0.124 0.348 0.353 0.500 0.612 0.186 0.401 0.540 0.278 0.454 0.124 0.346 0.353 0.500 0.612 0.186 0.401 0.540 0.278 0.454 0.124 0.346 0.352 0.500 0.612

350

340 330

TABLE 6

Comparison of the experimentally measured and computed plane stress dimensionless displacement values for the three point bendin9 (span to wdth ratio of 4 to 1) tr= 6P Ihw
Specimen number Included vertex angle c~ 330 330 330 300 300 300 270 270 270 Notch depth to plate width ratio a/w 0.4 0.5 0.6 0.4 0.5 0.6 0.4 0.5 0.6 Experimental Computed

2E uSaa

2E uy/atr

1 2 3 4 5 6 7 8 9

8.69 11.86 17.71 8.90 12.17 18.07 9.46 13.01 19.80

8.72 11.89 17.86 8.92 12.19 18.37 9.55 13.12 20.00

Int. Journ. of Fracture Mech., 8 (1972) 26%276

274
25

B. Gross, A. Mendelson

20
E~

15

~.

Notch length to plate width ratio a/w

1O
5

------~--__ ~

-I. 00 -.go -.80 -.70 -. 60 -. 50 -. 40 -. 30 -. 20 -. 10 0

Location, x/a

Figure 5. Three-pointloadingelasticnotch edgedisplacementsfor planestress conditions.Includedvertexangle,300.

_Loading

unit I crossarm vertical I displacement I control and I load recorder ]

Power supply 1 supplies excitation ~ , voltage and bridgel I [ ]balance i i i

. .

and 10adcell0utput~--~

.J
g Oisplacement / gage

gxcitation voltage ag Su~lpau~eent agde(~ m

IDmiigc%volt r t "-t mete Digital printout Standard cell output Load cell bridge excitatation voltage Displacement gage bridge excitation voltage Load ceil output Displacement gage output

~ ~__ /

Calibrated sensing compressive load unit |Power supply 2 .~Isupplies excitation |voltage and bridge | balance

Figure 6. Schematicof experimentaltest rig. Figure 4 contains typical dosed contour curves of dimensionless equivalent stress. One obtains from these contours a semi-qualitative look into possible regions of plastic flow and their shapes. Straight lines have been drawn between points of constant equivalent stress along rays emanating from the crack tip at 10 degree intervals. A typical dimensionless plane stress y displacement curve along the notch surface is plotted in figure 5 for mode I. Displacements were computed along the notch at intervals of one-tenth the relative notch depth fla. Corresponding plane strain displacements may be obtained by multiplying the plane stress y displacement by (1 - v2).
Int. Journ. of Fracture Mech., 8 (1972) 267-276

Plane elastostatic analysis of V-notched plates

TABLE 7

275

Table of first eigenvalues, th,for mode I and mode II, for various included vertex angles,
Alpha rh Mode I 1.500000 1.500053 1.500426 1.501453 1.512221 1.544484 1.615731 Mode II 1.500000 1.529355 1.562007 1.598192

360 350 340 330 300 270 240

Confidence in the correctness of the results obtained by the method of the present analysis was derived from many comparisons, both with the experimental results obtained herein as well as experimental results of other authors for the crack (e = 360) problem. Additional verification was obtained when comparisons were made with solutions to several special cases by other investigators using different analytical techniques. A detailed comparison is given in reference 16. For included vertex angles e, less than 360 degrees no experimental or analytical data has heretofore been available. Experiments, as previously described were therefore made for a three point bend specimen and the results compared with the analysis presented herein for the case of plane stress and are shown in table 6. Results are given for ~ values of 330, 300 and 270 degrees and notch depth to plate width ratios a/w of 0.4, 0.5 and 0.6. Excellent agreement is obtained over the range of a/w ratios. References
[1] W. F. Brown, Jr., and J. E. Srawley, Plane Strain Crack Toughness Testing of High Strength Metallic Materials, ASTM Special Technical Publication No. 410, (1966). [2] ASTM Committee E-24 On Fracture Testing of Metals, Proposed Method of Test for Plane-Strain Fracture Toughness of Metallic Materials, 1969 Book of ASTM Standards, Part 31, pp. 1099-1114. [3] B. Gross, J. E. Srawley, and W. F. Brown, Jr., "Stress Intensity Factors for a Single-edge-notch Tension Specimen by Boundary Collocation of a Stress Function", NASA TN D-2395, 1964. [4] J. H. A. Brahtz, "Stress Distribution in a Reentrant Corner", Journal of Applied Mechanics, 1, 2 (1933) 31-38. [5] H. M. Westergaard, "Stress at a Crack, Size of the Crack, and Bending of Reinforced Concrete", Journal of the American Concrete Institute, 5, 2 (1933) 93-102. [6] M. L. Williams, Stress Singularities Resulting From Various Boundary Conditions in Angular Corners of Plates in Extension, Journal of Applied Mechanics, 19, 4 (1952) 526-528. [7] M. L. Williams, On the Stress Distribution at the Base of a Stationary Crack, Journal of Applied Mechanics, 24, 1 (1957) 109-114. [8] B. Gross and J. E. Srawley, Stress Intensity Factors for Single-Edge-Notch Specimens in Bending or Combined and Tension by Boundary Collocation of a Stress Function, NASA TN D-2603, (1965). [9] A. M. Sullivan, New Specimen Design for Plane Strain Fracture Toughness Tests, Materials Reserach and Stan., dards, 4, 1 (1964) 20-24. [10] J. D. Lubahn, Experimental Determination of Energy Release Rate for Notched Bending and Notched Tension, Proceedings of the ASTM, 59 (1959) 885-913. [11] J. E. Srawley, M. H. Jones, and B. Gross, Experimental Determination of the Dependence of Crack Extension Force on Crack Length for a Single-Edge-Notch Tension Specimen, NASA TN D-2396, 1964. [12] M.H. Jones, R. T. Bubsey, and W. F. Brown, Jr., (proposed NASA Technical Note). [13] A. S. Kobayashi, Method of Collocation Applied to Edge Notched Finite Strip Subjected to Uniaxial and Pure Bending, Boeing Company Report D2-23551, August, 1964. [14] L.A. Wiggelsworth, Stress Distribution in a Notched Plate, Mathematika, 4 (1957) 76-96. [15] H.F. Buecker, Weight Functions for the Notched Bar, General Electric Co. Report, 69-LS-45. May 12, 1969. [16] B. Gross, Some Plane Problem Elastostatic Solutions for Plates Having a V-Notch", Case Western Reserve University, Ph.D. Thesis, 1970. [17] B. Gross, E. Roberts, Jr., and J. E. Srawley, Elastic Displacements for Various Edge Cracked Plate Specimens, The International Journal of Fracture Mechanics, 4, 3 (1968) 267-276 ; also see NASA TN D-4232, 1967. Errata relating to this paper Int. Journ. of Fracture Mech. 6 (1970) 87.

Int. Journ. of Fracture Mech., 8 (1972) 267-276

276

B. Gross, A. Mendelson

[18] A. P. Moser, Elastic SZress Fields and Stress Intensity Factor For Finite Bodies, University of Colorado, P h . D . Thesis, 1967. [19] G. R. IrwinAnaysis f Stresses and Strains Near End f a Crack Jurnal Applied M echanis 2 4 3 (1957) 36 -364. [20] L. E. Hulbert, The Numerical Solution of Two Dimensional Problems of the Theory of Elasticity, P h . D . Thesis, Ohio State Univ., (1963).

RI~SUMI~ On fournit des solutions g divers probt6mes statiques, en conditions planes et 61astiques, relatifs/t des t61es entaill6es en v6 sur un de leurs bords et soumises/t diverses conditions aux limites. O n examine l'influence de Fouverture de l'angle de l'entaille et de la g4om6trie de l'6prouvette sur les facteurs d'intensit6 de contraintes K~ et K n darts les cas oit les surfaces de l'entaille ne sont pas sollicit6es, et l'on conclut que, lorsque l'angle de l'entaille passe de z4ro/t 30 degr6s, le facteur d'intensit6 des contraintes K I se modifie de moins d'un pour cent. Des mesures de d6placement d'ouverture d'entaille ont ~t~ effectu&es, et les r~sultats obtenus sont en excellent accord avec les r6sultats des calculs.

Int. Journ. of Fracture Mech., 8 (1972) 267-276