International Conference on Computational Mechanics (CM13) 25-27 March 2013, Durham, UK

Strain Concentration Analysis of Biaxialy Loaded Countersunk Hole in an

Orthotropic Plate

Feras Darwish a,*, Ahmad Al-Shyyaba, Mohammad Hayajneh b

a
Aeronautical Engineering Department, Jordan University of Science and Technology, P.O. Box 3030, Irbid 22110, Jordan
b
Mechanical Engineering Department, Jordan University of Science and Technology, P.O. Box 3030, Irbid 22110, Jordan

Abstract
This research investigates the strain concentration factors (SCF’s) around a centered countersunk hole in an infinite
orthotropic laminate subjected to a biaxial static load. ANSYS - finite element software is used to build the finite
element (FE) model and to perform the analysis. The effect of the biaxial remote strain ratio, the geometric parameters
of the hole, and the material orthotropy represented by the ply angle of the laminate on the maximum SCF’s at the hole
is studied. Based on the FE results, it is found that the SCF’s are directly proportional to the parametric dimensions of
the hole and also dependent of the biaxial remote strain ratio and the ply angle of the laminate. It is found that a
compromised minimum SCF’s can be attained by selecting a ply angle between 70 and 80-degrees for remote strain
ratios less than 1.0 and a ply of 45o for a remote strain ratio of 1.0.In this paper we show the basic formatting required
for papers submitted to the CM13 conference to be held in March 2013 at Durham University. Please use this file as the
initial template for your own submission and follow the guidelines included. This will allow the rapid inclusion of your
work within the conference and associated proceedings.
Keywords: Strain concentration, Biaxial load, Orthotropic material

1. Introduction
Flush rivets are commonly used in joining thin structural components such as the fuselage and the external surfaces of
airplanes and marine vessels. Upon their application, flush rivets form countersunk holes through the thickness of the
joint and act like stress and strain risers. Therefore, stress and strain analyses around such holes must be carefully
considered in the design of the joint to avoid any structural failures at low load levels.
Numerous results for the stress concentration factor (SCF) around holes of different geometries and under different
types of loading conditions are reported in the literature [1]. Experimental failure results of riveted fuselage specimens
under uniaxial and biaxial loading conditions were reported by the FAA [2]. The failure due to biaxial tensile loading of
a quasi-isotropic composite plate with a circular hole was experimentally investigated [3, 4]. The limit load of plates
with different hole shapes under different loading conditions was analytically estimated and compared with the FE
results [5]. Analytical stress solutions were obtained for the SCF around cutouts in infinite composite laminates
subjected to biaxial loading [6]. A finite element investigation was carried out on a stiffened plate with a square cutout
under various combinations of biaxial loading conditions [7]. A general analytical solution was also obtained for the
stresses and deformations around a traction free elliptic hole in an infinite plate subjected to a biaxial load [8]. The size
effect of a countersunk hole on the SCF of infinite plates subjected to different loading conditions was investigated
through finite element analysis (FEA) [9]. Parametric equations for the stress and strain concentration factors around a
countersunk hole in an isotropic plate under uniaxial tension were formulated [10, 11]. A modified equation for the
uniaxial SCF around a countersunk hole was reported as well [12]. A new equation for the SCF around countersunk
rivet holes in orthotropic laminates under uniaxial tension was expressed as a function of the hole geometry and the
material orthotropy [13]. The effect of the plate thickness and the notch geometry on the SCF, the stress constraint
factor and the strain energy density was investigated through an analytical solution of the stress domain around the
notch [14].
The main objective of this study is to examine through FEA the effect of the remote strain ratio (R= xo/yo), the
dimensions of the countersunk hole and the material orthotropy of the laminated plate on the in-plane strain
concentration factors in the x- and y-directions, Ktx and Kty respectively, which are defined as the ratio between the
maximum local strain at the hole and the applied remote strain at the edge of the plate as expressed in equations (1).
 x ,max  y ,max
Kt  , x  , Kt  , y  (1)
 xo  yo

*
Corresponding author
Email address: fhdarwish@just.edu.jo ( Feras Darwish )

Extended abstract submitted to CM13 September 12, 2012

2. The configuration of the plate
The geometry of the plate with a countersunk hole and the defining geometric parameters are shown in figure 1.
The plate is considered relatively infinite with the plate’s half width to hole radius ratio (w/r) and half-length to radius
ratio (L/r) greater than 15. A biaxial static tension load represented by the in-plane remote strain in the x- and y-
directions, xo and yo respectively, is applied at the edges of the plate. Due to the symmetry in the plate and the applied
loads across the x- and y-axes, only one quarter of the plate will be considered for the FE modeling to reduce the
number of generated elements and hence reduce the analysis run time. Accordingly, new generated boundary conditions
will appear at the cut lines (lines of symmetry) in order to restrain the displacement in the x-direction at the cut surface
at x=0 and the displacement in the y-direction at the cut surface at y=0.

Figure 1: The geometry and boundary conditions of a plate with a centered countersunk hole

An orthotropic material system of eight plies of carbon/epoxy (AS4/3501-6) with stacking sequence [±θp]2s is
considered for the analysis as shown in figure 2. The mechanical properties of a single lamina of the unidirectional
(AS4/3501-6) in the principal directions are: E1 = 149 GPa, E2= 10.3 GPa, E3 = 10.3 GPa, G12 = 6.9 GPa, G23 = 3.7 GPa,
G13 = 6.9 GPa, ν12=0.27, ν23=0.54 and ν13=0.27 [15].

Figure 2: Orthotropic laminate with stacking sequence of [±θp]2s

The value of the ply angle (θp) describes the angle between the fibers and the y-axis, (i.e. θp=0 represents fibers aligned
with the y-axis and θp=90 represents fibers aligned with the x-axis), and determines the homogenized mechanical
properties of the orthotropic laminate (Ex, Ey, Ez, Gxy, Gyz, Gxz, νxy, νyz and νxz). The homogenization of the mechanical
properties of a laminate was adopted by several researchers in the literature including Whitney and Nuismer [16], Pipes
et al. [17], Tan [18], Soutis and Hu [19], and Darwish and Shivakumar [20].

3. Finite element modelling

ANSYS Parametric Design Language (APDL) was used to build the FE code that was used to define the material
properties and the element type, create the volumes, mesh the geometry, assign the load and boundary conditions and
perform the analysis. Three dimensional hexahedral brick element (solid 45) was used to map mesh the volumes of the
geometry. Mesh gradation and mesh refinement studies were conducted to create fine mesh at the hole and course mesh
elsewhere and to optimize the FE solution. Figure 3 shows the meshed FE model of the quarter plate.

2
 Figure 3: The FE model of the quarter plate

The FEA of the present study was verified by using the adopted modeling strategy to reproduce the solutions and
compare with the results of several problems from the literature. The verification study has shown high credibility in the
present modeling strategy and it was published in previous studies [12, 13].

4. FE results and discussion

In order to study the effect of the biaxial remote strain ratio, the material orthotropy, and the hole dimensions on
the maximum strain concentration factors Ktx and Kty, several runs were performed at different load values, different
ply angles (p), and different hole dimensions. The geometric parameters that control the dimensions of the hole are the
plate’s thickness to hole radius ratio (t/r), the countersink depth of the hole to the plate’s thickness ratio (Cs/t), and the
countersink angle of the hole (c).

4.1 The effect of the hole dimensions on the strain concentration factors

To examine the effect of the geometric parameters of the countersunk hole (t/r, Cs/t, c) on Ktx and Kty, the remote
strain ratio and the ply angle were maintained constant at 0.4 and 30o respectively. The FE results have shown that the
strain concentration factor in both directions increases with increasing t/r, or Cs/t, or c as shown in table 1.

Cs/t=0.5, θc=100o t/r=0.5, θc=100o Cs/t=0.5, t/r=0.5

t/r Ktε,x Ktε,y Cs/t Ktε,x Ktε,y θc Ktε,x Ktε,y
0.2 7.24 3.96 0.00 6.64 3.86 80 7.33 4.10
0.5 7.54 4.21 0.25 7.04 4.01 90 7.44 4.14
1.0 7.74 4.43 0.50 7.54 4.21 100 7.58 4.21
2.0 9.42 4.85 0.75 7.34 4.03 110 7.69 4.28
3.0 10.05 5.11 1.00 7.71 4.24 120 7.82 4.37
Table 1: The effect of the hole’s geometric parameters on the strain concentration factor at R=0.4, θp =30o

The direct proportionality between the strain concentration factors and the geometric parameters of the hole is referred
to the fact that when increasing t/r or Cs/t while maintaining the other parameters fixed, leads to an increase in the
partial thickness of the plate accommodating the sinking surface of the hole. This change in geometry pushes more
force lines down toward the root of the sinking surface yielding to an increase in the strain concentration factors around
the hole. In addition, when increasing the value of the countersink angle θc while maintaining all other parameters fixed,
the outer circle of the sinking hole at the surface of the plate becomes larger. This change in geometry permits more
force lines to redirect their flow downward toward the root of the sinking hole to turn around the smaller circle. This
would absolutely increase the magnitude of the strain concentration factor.

4.2 The effect of θp and R on the strain concentration factors

As mentioned earlier, the ply angle θp represents the fiber orientation of the orthotropic laminate and in this paper it is
measured from the y-axis. It is well known that θp determines the values of the mechanical properties of the laminate
which in turn affect the values of the local stresses and strains around the hole. In addition to the ply angle, the biaxial
remote strain ratio, R, directly affects the local stress and strain distributions around the hole. In other words, the local
strain value at any point in the orthotropic laminate is function of the stiffness matrix which is dictated by the ply angle,
and the local stresses which are dependent of the remote strain ratio. The effect of the ply angle on the strain
concentration factor at two values of R (0.4 and 1) is shown in figure 4.

3
 8
R=0.4
7

6
R=1.0
5

Ktε
4

3
R=0.4
2
Ktε,x
1
Ktε,y
0
0 10 20 30 40 50 60 70 80 90
Ply Angle (θp)
Figure 4: The effect of θp on Ktε (t/r=0.5, Cs/t=0.5, θc=100o)

It can be seen in figure 4 that when R=1 (xo=yo), Ktx and Kty have perfectly opposite trends. This is due to the
application of unbiased remote loads while varying the ply angle from 0 o to 90o. In this case scenario, the Ktx curve has
in general an ascending trend and Kty curve has a descending trend as the material becomes stiffer in the x-direction
and weaker in the y-direction while changing θp from 0o to 90o. On the other hand, when R=0.4, the influence of the
remote load becomes biased and dominated by the load in the y-direction. Therefore, Kty fairly maintains its
descending trend with a steeper drop at θp between 30o and 50o, and Ktx becomes more dominated by R which dictates
together with θp the value of the maximum local strain in the x-direction and hence Ktx. It can also be noticed in figure
4 that for R=1, a ply angle around 45o provides a compromised minimum values of the strain concentration factors,
whereas for R=0.4, a ply angle of around 70o would be the right choice. For other R values (0.2, 0.6, 0.8), it was found
that a ply angle between 70 and 80-degrees provides a compromised minimum values for Ktx and Kty.
In this context, it is important to shed light on the influence of varying θp and R on the ratio of the maximum local strain
in the x- and y-directions (x,max/y,max). Figure 5 shows the relationship between (x,max/y,max) and θp for R values varying
from 0.2 to 1. It can be seen in figure 5 that x,max is always less than y,max (x,max/y,max < 1)for R=0.2 and that x,max
becomes greater than y,max (x,max/y,max > 1) when R ≥ 0.4 and θp > 70o. It is also noticed that at constant θp, increasing
the remote strain ratio leads to an increase in the local strain ratio, and that at constant R value, the local strain ratio
curve has in general an ascending trend with increasing the ply angle. In addition, for θp values between 40o and 50o the
slope of the curve becomes near zero for R≥0.4 and slightly negative for R=0.2.
R=1.0
2.5 R=0.8
R=0.6
2.0
R=0.4
1.5
εx/εy

1.0

R=0.2
0.5

0.0
0 10 20 30 40 50 60 70 80 90
Ply Angle (θp)

Figure 5: The effect of θp and R on the local strain ratio (εx,max/εy,max)

4.3 The location of the maximum strain concentration factor

The FE results have shown that the maximum strain concentration factor in either the x- or y-direction is always located
at the edge between the sinking and the straight shank surfaces of the countersunk hole. This part of the study aims to
specify the circumferential location around the edge of the hole at which the maximum strain concentrations are found
to occur at different ply angles of the laminate. For this purpose, two angles x and y are defined as the angles
between the location of x,max and the x-axis, and the location of y,max and the y-axis respectively. Figure 6 shows the
variation of x and y versus θp for R=0.4.

4
 Figure 6: Location of the maximum strain concentrations around the hole edge (t/r=0.2, Cs/t=0.5, θc=100o, R=0.4)

As shown in figure 6, the circumferential locations of the points of maximum local strains x,max and y,max are dependent
of the ply angle of the laminate. For example, for θp between 0o and 30o, the laminate is stiffer in the y-direction than in
the x-direction, hence, the location where y is maximum will be dominated by the resulting y-component stress (y)
which is always maximum at y =90o. For θp increased beyond 30o, the effect of y on y becomes weaker as the
laminate starts to gain strength in the x-direction and to be more compliant in the y-direction. This gradual increase of
θp gives a gradual degrading effect of y and an increasing effect of x on y, therefore, y starts to decrease until it
reaches its minimum at θp values between 70o and 90o. An opposite behavior is found for x which describes the
circumferential location of x,max. It is shown in figure 6 that the two curves do not show perfect opposite trends, that is
because of the biased remote strain ratio, R=0.4.

Cs/t=0.5, θc=100o t/r=0.5, θc=100o Cs/t=0.5, t/r=0.5

t/r Ktε,x Ktε,y Cs/t Ktε,x Ktε,y θc Ktε,x Ktε,y
0.2 7.24 3.96 0.00 6.64 3.86 80 7.33 4.10
0.5 7.54 4.21 0.25 7.04 4.01 90 7.44 4.14
1.0 7.74 4.43 0.50 7.54 4.21 100 7.58 4.21
2.0 9.42 4.85 0.75 7.34 4.03 110 7.69 4.28
3.0 10.05 5.11 1.00 7.71 4.24 120 7.82 4.37
Table 1: The effect of the hole’s geometric parameters on the strain concentration factor at R=0.4, θp =30o

5. Conclusions
Three-dimensional FEA was performed on an orthotropic laminate with a centered countersunk hole and subjected to a
biaxial remote strain. The research aimed to investigate the effect of the geometric parameters of the hole, the remote
strain ratio, and the ply angle on the strain concentration factors around the edge of the hole. Based on the FE results, it
was found that as any of the geometric parameters of the hole (t/r, Cs/t, c) increases, both Ktx and Kty increase. It was
also shown that there is a combined effect of the remote strain ratio and the fiber orientation on the values and the
locations around the edge of the hole of the Ktx and Kty. In general, and based on the obtained results, it is
recommended to use a ply angle of 45o with a remote strain ratio of 1.0 and a ply angle between 70 and 80-degrees with
remote strain ratios less than 1.0 that is to attain compromised minimum values of the SCF’s and to increase the static
strength capacity of the laminate.

Acknowledgements

5
References
1. Pilkey, Walter D. and Pilkey, Deborah F., Peterson's stress concentration factors, 3rd Edition, John Wiley &
Sons, 2008
2. Vlieger, H., Ottens, H.H., Uniaxial and biaxial tests on riveted fuselage lap joint specimens, FAA final report,
DOT/FAA/AR-98/33, 1998
3. Daniel, Isaac M., Behavior of graphite/epoxy plates with holes under biaxial loading, Experimental Mechanics,
Vol. 20(1), 1-8, 1980
4. Daniel, Isaac M., Biaxial testing of [O2/±45], graphite/epoxy plates with holes, Experimental Mechanics, Vol.
22(5), 188-195, 1982
5. Rahimi, G.H., Alashti, R.A., Limit load analyses of plates with a hole under in-plane loads, Scientia Iranica,
Vol. 12(4), 442-454, 2005
6. Sharma, Dharmendra S., Stress concentration around circular/elliptical/triangular cutouts in infinite composite
plate, Proceedings of the world congress on engineering, Vol. III, London, UK, 2011
7. Srivastava, A.K.L., Datta, P.K., Elsatic stability of square stiffened plates with cutouts under biaxial loading,
Int. J. of Applied Mechanics and Engineering, Vol. 11(2), 421-433, 2006
8. Gao, Xin-Lin, A general solution of an infinite elastic plate with an elliptic hole under biaxialloading,
International Journal of Pressure Vessels and Piping, Vol. 67(1), 95-104, 1996
9. Shivakumar, K.N., Newman, J.C., Stress concentrations for straight-shank and countersunk holes in plates
subjected to tension, bending, and pin loading, NASA technical Report Paper 3192, 1992
10. Bhargava, A., Shivakumar, K.N., Three-dimensional tensile stress concentration in countersunk rivet holes,
The Aeronautical Journal , RAeS, 111, 1126, 777-786, 2007
11. Bhargava, A., Shivakumar, K.N., A three dimensional strain concentration equation for countersunk holes,
Journal of Strain Analysis for Engineering Design, Vol. 43(2), 75-85, 2008
12. Darwish, F., Gharaibeh, M., Tashtoush, G., A modified equation for the stress concentration factor in
countersunk holes, European Journal of Mechanics-A/Solids, Vol. 36, 94-103, 2012
13. Darwish, F., Tashtoush, G., Gharaibeh, M., Stress concentration analysis for countersunk rivet holes in
orthotropic plates, European Journal of Mechanics-A/Solids, Vol. 37, 69-78, 2012
14. Betro, F., Lazzarin, P., Wang, C.H., Three-dimensional linear elastic distributions of stress and strain energy
density ahead of V-shaped notches in plates of arbitrary thickness, International Journal of Fracture, Vol. 127,
265-282, 2004