On the Static Analysis of Sandwich Panels with Square Honeycomb Core
Hazem E. Soliman*, Dhaval P. Makhecha†, Summit Vasudeva‡, Rakesh K. Kapania§ and Owen Hughes**
Virginia Polytechnic Institute and State University, Blacksburg, VA, 24061
I. Abstract tatic analysis of sandwich panels with square honeycomb core is presented. The displacement field of the sandwich panel is obtained. The analysis of the panel is done using the finite element approach applied to
the Classical Laminated Plate theory (CLPT), the First Order Shear Deformation theory (FSDT) and the Higher Order Shear Deformation Theory (HSDT). The constitutive behavior of a continuum equivalent to the core is obtained using Hohe’s approach which is a homogenization approach based on the strain energy equivalence between a body containing cellular structure and a continuum body. The results of the displacement field of the sandwich panel are then compared to results obtained from a highly detailed FEM created in ABAQUS® in which the cell walls of the honeycomb core are modeled as shell elements, this comparison shows that the displacement results of the HSDT are in error of 7.6% when compared to the detailed ABAQUS® model. An alternative procedure for calculating the constitutive behavior of the continuum equivalent to the square honeycomb was pursued here by creating detailed FEM of the unit cell of the square honeycomb using ABAQUS® to detailed representative models and applying static loads to these models then obtaining the properties by means of the flexibility approach. Comparison between the equivalent strain energy approach and flexibility approach applied to detailed ABAQUS models of the unit cell of the square honeycomb proved the equivalent strain energy approach to be efficient. Simple formulas for calculating the different Elastic properties of equivalent continuum to the square honeycomb are introduced.
S
II. Introduction Sandwich panels have been the focus of numerous studies done in the past few decades due to their high strength
to weight ratio and high energy absorption characteristics. The use of sandwich panels is gaining momentum in different industries such as aerospace, automotive and ship-building. Sandwich panels having a honeycomb core constitute the state-of-the art of light-weight construction due to the amount of weight savings and high strength.
While it is very important to study the sandwich structures as a whole, it is equally important to analyze honeycomb core to estimate the material properties of this type of structures to identify their impact on the constitutive behavior. This is the only way in which we can design cores for enhanced energy absorption capability to resist blast and impact loads. The study of the static and dynamic response of honeycomb cores as well as sandwich panels with other types of cores has been the focus of many researchers in the past. However, based on a recent extensive literature study by Noor1 and to the author’s knowledge, studying in detail the complete constitutive behavior of sandwich panels with honeycomb core has been not attempted theoretically or experimentally.
Many researchers have conducted studies on the properties of honeycomb structures; Torquato2 et al. conducted
a 2-D study on the properties of hexagonal, square, and triangular honeycombs. Their study included two approaches, a homogenization approach and a discrete network analysis. Christensen3 has conducted a 2-D and a 3-D study on the hexagonal, triangular, combination of hexagonal and triangular honeycomb and a combination of
*GRA, Virginia Polytechnic Institute and State University, Aerospace and Ocean Engineering Department, AIAA Student member. †Research engineer, GE India Technology Center Pvt. Ltd. ‡GRA, Virginia Polytechnic Institute and State University, Aerospace and Ocean Engineering Department, AIAA Student member. §Professor, Virginia Polytechnic Institute and State University, Aerospace and Ocean Engineering Department,, AIAA Associate Fellow. **Professor, Virginia Polytechnic Institute and State University, Aerospace and Ocean Engineering Department
American Institute of Aeronautics and Astronautics
1
47th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Confere1 - 4 May 2006, Newport, Rhode Island
AIAA 2006-2168
Copyright © 2006 by the American Institute of Aeronautics and Astronautics, Inc. All rights reserved.
hexagonal and star honeycomb structures. Hayes4 et al. provided simple analytical formulae to obtain the elastic properties of the equivalent continuum to the honeycomb structures based on Timoshenko’s beam theory in addition to some experimental results and provided the response of honeycomb structures to dynamic loading. The idea of using Timoshenko’s beam theory, however, to obtain the elastic properties of cellular structures doesn’t appear to be efficient because it does not represent all the deformations that could be encountered by a cell wall of a honeycomb core. Baker5 et al. conducted an experimental study on the energy absorption of honeycomb structures made of thick-walled aluminum and stainless steel and having a sinusoidal shape with a straight wall running between every two sinusoidal waves. Liang6 et al. presented the Treble series solution of buckling mode and derived the formulae for determining the critical compressive stress of the square honeycomb. In a series of publications, Papka7-10 et al. studied the in-plane compressive response and crushing of honeycombs.
Karagiozova11 et al. employed a limit analysis to identify the plastic deformation modes of hexagonal
honeycomb for high thickness to length ratios under in-plane biaxial compression. Other publications12,13 dealt with that subject as well. Wang et al. conducted a study to find the initial yield surfaces of different honeycomb structures including the square and hexagonal honeycomb.
Hohe15-17 et al. presented a homogenization procedure based on the strain energy equivalence between two
bodies if they both have the same response to the same type of loading, subject to the same boundary conditions and have the same shape with the exception that one of the bodies contains a cellular structure while the second body is a continuum. He applied this equivalence to honeycomb structures assuming the existence of an equivalent continuum structure. The determination of the stiffness matrix started by assuming that the deformations in each cell wall of the cellular structure will follow the exact deformation pattern of its edge, i.e., the displacement field is uniform through the depth of the cell wall. Hohe then assumed the edge of the cell wall to deform as a beam. Starting with the displacement field based on the Timoshenko’s beam theory, Hohe obtained the strains in the cell wall and then assumed a state of plane stress for the cell walls and obtained the stresses in the cell walls using Hooke’s law. From the stresses and strains, the strain energy of the entire unit cell is obtained by summing up the strain energy of the individual cell walls. The stiffness matrix is then simply obtained from the strain energy expression. The procedure introduced by Hohe is the most recent and most complete as far as obtaining the constitutive behavior of the equivalent continuum to the honeycomb structure is concerned.
In this paper, the constitutive behavior of equivalent continuum to the square honeycomb is obtained using the
procedure outlined by Hohe17. We then incorporated the properties obtained in an in-house code that solve sandwich panels to obtain the stress and displacement fields using FE approach applied to the CLPT, FSDT as well as HSDT18,22. We also present a detailed FEM results for a sandwich panel with square honeycomb core, these results were obtained using ABAQUS®. A comparison between the two procedures shows that the HSDT results are in error of 7.6% compared to the detailed FEM ABAQUS® model. An investigation of the accuracy of the strain energy approach to calculate the constitutive behavior of the continuum equivalent to the square honeycomb core using detailed ABAQUS® models of the unit cells of the square honeycomb is pursued. This investigation proves the homogenization approach by means of the strain energy equivalence is accurate. We then present simple formulas that correlate the constitutive behavior of a continuum equivalent to the square honeycomb as a function of the relative density of the honeycomb and the elastic modulus of the material used to manufacture the square honeycomb.
III. Strain Energy Approach The homogenization procedure using strain energy equivalence approach was developed by Hohe15-17. A brief of
this procedure applied to the square honeycomb is presented here while more details are offered in Appendix A for the reader’s convenience. The approach is based on the concept that for any volume element containing cellular material, there is an equivalent continuum element with homogeneous properties and the same strain energies per unit surface area provided that both volume elements are subject to the same loading and boundary conditions.
Figure 1 shows an appropriate representative volume element of the square honeycomb. The total strain energy
of the volume element is calculated as the sum of the strain energy of the different cell walls contained in this volume element.
American Institute of Aeronautics and Astronautics
2
aa
Figure 1: a representative volume element of the square honeycomb
The representative volume element of the square honeycomb contains 4 cell walls. The displacement field for each of the cell walls is assumed to be uniform in the direction normal to the plane of figure 1 while a Timoshenko beam displacement field is assumed in the plane of the figure. Figure 2 shows the details of breaking up the element into cell walls and the displacement field in the 2D plane of the square honeycomb. The strain field in each cell wall is calculated by partial differentiation of the displacement field. The stress field is then calculated from the strain field by means of Hooke’s law in conjunction with the plane stress assumption in the cell walls. The total strain energy of each cell wall is then calculated as the volume integration of the strain energy density. The coefficients of the stiffness matrix Cijkl are then calculated by differentiating the sum of the total strain energy of all the cell walls in the volume element using the following expression
klij
ijklWCεε ∂∂
∂=
2
(1)
Where W is the sum of the total strain energy of all the cell walls in the volume element.
Figure 2: The square honeycomb element dissected into 4 cell walls and the assumed displacement field (adapted from Hohe)
The constitutive behavior of the continuum equivalent to the square honeycomb is noticed to be identical to
transversely isotropic materials
IV. Plate Theories In this study, the analysis of sandwich panels is carried out using an in-house code that solves sandwich plate
problems using the finite element approach applied to the CLT, FSDT and HSDT. In this section, we introduce a comparison between the kinematics of the different plate theories. The displacement field in the most general case takes the form
1
2
4
3 5 v(1)1
v(1)2
v(5)1
v(5)2
Δφ(1)
Δφ(5)v(5)2
v(4)2 v(4)1
Δφ(4)
v(5)1
Δφ(5)Δφ(5)
v(3)1
v(3)2 Δφ(
v(5)2
v(5)1 3)v(5)2
v(5)1 Δφ(5)
Δφ(2)v(2)2
v(2)1
American Institute of Aeronautics and Astronautics
3
(2)
kkkk
o
yk
yyyo
xk
xxxo
hzSyxzyxwzyxwzyxw
yxSyxzyxzyxzyxvzyxvyxSyxzyxzyxzyxuzyxu
/)1(2
),(),(),(),,(
),(),(),(),(),(),,(
),(),(),(),(),(),,(
21
32
32
−=
Γ++=
++++=
++++=
φςξψ
φςξψ
Where uo(x, y), vo(x, y), and wo(x, y) are the displacements of a point (x, y) on the midplane ψx, ψy represents the rotations of the midplane normal about the y and x axes respectively w1, Γ, ξx, ξy, ζx, and ζy are functions used only in the HSDT φx, φy are known as the zigzag functions zk is the local transverse coordinate with its origin at the center of the kthlayer
hk is the thickness of the kth layer In the CLPT20, the normal to the midplane before deformations is assumed to remain straight and normal to the
midplane after deformation. This assumption leads to the following:
01 =======Γ=∂∂
−=
∂∂
−=
yxyxyx
y
x
wywxw
φφζζξξ
ψ
ψ
(3)
In the FSDT20, it is assumed that the normal to the midplane before deformation remains straight but not
necessarily normal to the midplane after deformation. This assumption leads to the following:
01 =======Γ=
==
yxyxyx
yy
xx
w φφζζξξ
θψθψ
(4)
In the HSDT22, the in-plane displacements u and v are assumed to have a cubic distribution and therefore,
equations 2 represents the displacement field. In these equations, the functions w1, Γ, ξx, ξy, ζx, and ζy are higher order terms.
V. Finite Element Formulations The derivations of the finite element models for the plate theories used can be found in literature18, 20-22. A brief
discussion of the elements is presented here for the reader’s convenience. An eight-node serendipity quadrilateral plate element was used for all the FEM analysis. The number of degrees of freedom per node used in the CLPT and FSDT element was five while in the HSDT element, the number was thirteen. Table 1 lists the degrees of freedom for each of these elements.
Table 1: Degrees of freedom used in each of finite element formulations
THEORY DEGREES OF FREEDOM uo, vo, wo, ψx, ψyCLPT uo, vo, wo, θx, θyFSDT
uo, vo, wo, θx, θy, w1, Γ, ξx, ξy, ζx, ζy, φx, φyHSDT
American Institute of Aeronautics and Astronautics
4
The analysis was carried out for a square plate with length L = 200 mm. The face sheets thickness is 2 mm each while the core thickness is 8 mm. In this analysis, the properties used for the square honeycomb are those obtained for the continuum equivalent to the square honeycomb. Both face sheets and honeycomb core are made of Aluminum with Young’s modulus of 69 GPa and Poisson’s ratio of 0.25. The plate is subject to an out of plane pressure of 1 MPa and simply supported boundary conditions on all four sides. The boundary conditions are presented in Eq. 5.
The analysis was carried out for a square plate with length L = 200 mm. The face sheets thickness is 2 mm each while the core thickness is 8 mm. In this analysis, the properties used for the square honeycomb are those obtained for the continuum equivalent to the square honeycomb. Both face sheets and honeycomb core are made of Aluminum with Young’s modulus of 69 GPa and Poisson’s ratio of 0.25. The plate is subject to an out of plane pressure of 1 MPa and simply supported boundary conditions on all four sides. The boundary conditions are presented in Eq. 5.
0,0,0,2/@
0,0,0,2/@
===±= 0,0,0,2/@
0,0,0,2/@
===±=
± = ==
x
y
wuLywvLx
θ
=θ (5)
Figure 4: Finite element model of the unit cell of the square honeycomb using ABAQUS®
The results obtained for the transverse displacement of the center point of the plate using the finite element approach applied to the CLPT, FSDT and HSDT are then compared to those obtained from a detailed FEM created using ABAQUS®. The comparison shows that the HSDT results are 7.6% error when compared to the detailed ABAQUS® model.
VI. ABAQUS® Models A detailed ABAQUS® models was created for the sandwich plate with square honeycomb core. In this model,
the square honeycomb core is modeled in a great detail, i.e., the cell walls of the square honeycomb are modeled using shell elements, Figure 3 shows a picture of this plate model with the top and bottom face sheets removed for clarity. In this model, the face sheets and the cell walls of the honeycomb core were modeled using 2D elements in ABAQUS® of type S4R5. The plate is simply supported at all four edges. The square plate dimensions are 200×200×12 mm and the face sheets are 2 mm thick each while the core is 8 mm thick. Both the face sheets and the core are made out of Aluminum with Young’s modulus of 69 GPa and Poisson’s ratio of 0.25. The element library used in ABAQUS® was S4R5.
Figure 3: Detailed ABAQUS® model of sandwich plate with square honeycomb core. Face sheets removed for clarity
On the other hand, a detailed ABAQUS® model for a
unit cell of the square honeycomb was created for the purpose of verifying the accuracy of the strain energy based homogenization procedure followed to obtain the properties of the continuum equivalent to the square honeycomb. The unit cell of the square honeycomb was modeled at 20% relative density. Figure 4 shows the FEM of the unit cell created in ABAQUS®. The properties were obtained using the flexibility approach applied to this model. The results showed that the strain energy approach followed to obtain the properties of the continuum equivalent to the square honeycomb is highly accurate.
American Institute of Aeronautics and Astronautics
5
VII. Results The sandwich structure with square honeycomb core was studied. The face sheets and square honeycomb core
are both made of Aluminum with Young’s modulus of 69 GPa and Poisson’s ration of 0.25. The plate dimensions are 200 ×200×12 mm with the thickness of face sheets taken to be 2 mm each and the core thickness is 8 mm. The plate is subject to an out of plane pressure of 1 MPa and simply supported boundary conditions at all 4 sides. The properties of the continuum equivalent to the square honeycomb core were obtained using a strain energy based homogenization approach in which the strain energy of all the cell walls composing the representative volume element of the square honeycomb are calculated, summed and differentiated with respect to the strains to obtain the final stiffness matrix of the chosen representative volume element of the square honeycomb. These properties were then incorporated into an in-house code that solves sandwich plate problems using the finite element approach applied to the different plate theories, CLPT, FSDT and HSDT. The results for the displacement field of the sandwich plate are then obtained.
A detailed finite element model for the sandwich plate were created using ABAQUS®, the dimensions, loads and
boundary conditions are same as those used in the aforementioned finite element analysis. In this model, the cell walls of the square honeycomb were modeled as plates. The plate elements used in the ABAQUS® model in was the element of type S4R5. The displacement field obtained in these two cases was then compared to the displacement obtained from the in-house code. This comparison shows that the HSDT is in error of 7.6% of the ABAQUS® results while FSDT is in error of 15% and CLPT is in error of 15.5%
Table 2 lists the values of the transverse displacement at the center point of the plate for all the analysis
techniques mentioned above
Table 2: Results of the displacement at the center point of the sandwich plate
Equivalent Plate Theory ABAQUS®
Model S4R5 CLPT FSDT HSDT
1.528 mm 1.291 mm 1.295 mm 1.412 mm
The results listed above drove to the possibility that the homogenization procedure is not accurate enough to produce the properties of a continuum equivalent to the square honeycomb and a verification of this procedure was needed. For this purpose, a detailed finite element model of the unit cell of the square honeycomb was created in ABAQUS®, the properties of the square honeycomb were found using the flexibility approach applied to this finite element model. Table 3 shows the results obtained for the properties obtained for the continuum equivalent to the square honeycomb using the strain energy based homogenization approach against the properties obtained for the square honeycomb by means of the flexibility approach applied to the unit cell of the square honeycomb. The results show that the strain energy based approach is highly accurate; in addition, it shows that the square honeycomb behaves as a transversely isotropic material. Another interesting conclusion drawn from these results is that Poisson’s ration ν12 of the equivalent continuum vanishes.
Table 3: Strain energy based approach vs. flexibility approach applied using ABAQUS® Cijkl (MPa) Flexibility approach Strain energy approach Error
C1111 7360 7360 0% C1122 0 0 0% C1133 1840 1840 0% C2222 7360 7360 0% C2233 1840 1840 0% C3333 14260 14720 3.2% C2323 2748.64 2760 0.4% C1313 2748.64 2760 0.4% C1212 37.07 35.94 3%
American Institute of Aeronautics and Astronautics
6
In an effort to make finding the properties of the continuum equivalent to the square honeycomb easy, the coefficients Cijkl were calculated for the range of relative density between 10 – 50%. The values for each of the Cijkl normalized by Young’s modulus of the material used to manufacture the honeycomb was plotted vs. the relative density, the plots shows linear variation for all Cijkl except for C1212. Figures 5 – 9 shows these plots while equations 6 represent the corresponding formulas to obtain the coefficients Cijkl.
918.21212
23231313
22331133
3333
0563.0
2.0
13333.0
06667.1
2,1,53333.0
rc
rcc
rcc
rc
rc
iiii
EC
EC
EC
EC
ECE
C
iE
C
ρ
ρ
ρ
ρ
ρ
=
==
==
=
==
(6)
Where ρr is the relative density Ec is the Young’s modulus of the material used to manufacture the square honeycomb
0.05
0.1
0.15
0.2
0.25
0.3
10% 15% 20% 25% 30% 35% 40% 45% 50%
ρ r
C11
11/E
c, C
2222
/Ec
Figure 5: Variation of C1111 and C2222 with respect to the relative density
American Institute of Aeronautics and Astronautics
7
0
0.1
0.2
0.3
0.4
0.5
0.6
10% 15% 20% 25% 30% 35% 40% 45% 50%
ρ r
C 333
3/Ec
Figure 6: Variation of C3333 with respect to the relative density
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
10.00% 15.00% 20.00% 25.00% 30.00% 35.00% 40.00% 45.00% 50.00%
ρ r
C11
33/E
c, C
2233
/Ec
Figure 7: Variation of C1133 and C2233 with respect to the relative density
0
0.02
0.04
0.06
0.08
0.1
0.12
10.00% 15.00% 20.00% 25.00% 30.00% 35.00% 40.00% 45.00% 50.00%
ρ r
C13
13/E
c, C
2323
/Ec
Figure 8: Variation of C1313 and C2323 with respect to the relative density
American Institute of Aeronautics and Astronautics
8
0
0.001
0.002
0.003
0.004
0.005
0.006
0.007
0.008
10.00% 15.00% 20.00% 25.00% 30.00% 35.00% 40.00% 45.00% 50.00%
ρ r
C12
12/E
c
Figure 9: Variation of C1212 with respect to the relative density
VIII. Conclusions The static analysis of sandwich plates with square honeycomb core was studied. The properties of the continuum
equivalent to the honeycomb core were obtained using a strain energy based approach. These properties were then incorporated into an in-house code that solves sandwich plate problems using the finite element method applied to the different plate theories CLPT, FSDT and HSDT and the displacement field was obtained. For verification of the results, a detailed ABAQUS® model was created for the sandwich plate with square honeycomb core, in this model, the square honeycomb cell walls were modeled as plates. In both cases, the plate considered was a square plate subject to an out of plane pressure and simply supported boundary conditions at all four sides of the plate. The comparison between the results for the transverse displacement of the center point of the plate shows that HSDT results are in error of 7.6% compared to the ABAQUS® results while CLPT is in error of 15.5% and FSDT is in error of 15%.
To investigate the accuracy of the homogenization approach followed to obtain the properties of the continuum
equivalent to the square honeycomb, a detailed ABAQUS® model was created for the unit cell of the honeycomb, the properties were then obtained using the flexibility approach applied to the finite element model. The results show a very good agreement between the strain energy based homogenization approach and the flexibility approach applied to the finite element model of the unit cell of the square honeycomb. Results of the properties also show that the square honeycomb behaves identical to a transversely isotropic material and all stiffness coefficients vary linearly with the relative density except for the in-plane shear coefficient.
Acknowledgment The authors gratefully acknowledge the Office of Naval Research for their financial support for this research.
American Institute of Aeronautics and Astronautics
9
References 1Noor, A. K. Burton, W. S. Bert, C. W., “Computational models for sandwich panels and shells”, Applied Mechanics Reviews. Vol. 49, No. 3, pp. 155-199, 1996.2S. Torquato, L. V. Gibiansky, M. J. Silva, and L. J. Gibson, “Effective Mechanical and Transport Properties of Cellular Solids”, Int. J. Mech. Sci., Vol. 40, No. 1, pp. 71-82, 1998. 3R. M. Christensen, “Mechanics of Cellular and Other Low-Density Materials”, International Journal of Solids and Structures, Vol. 37, No. 1, pp.93-104, 2000. 4A. M. Hayes, A. Wang, B. M. Dempsey, and D. L. McDowell, “Mechanics of Linear Cellular Alloys”, Mechanics of Materials, Vol. 36, No. 8, pp. 691-713, 2004. 5W. E. Baker, T. C. Togami, and J. C. Weydert, “Static and Dynamic Properties of High-Density Metal Honeycombs”, Int. J. Impact Engineering, Vol. 21, No.3, pp. 149-163, 1998. 6S. Liang, and H. L. Chen, “Investigation on the Square Cell Honeycomb Structures Under Axial Loading”, Composite Structures, in press, 2005. 7S. D. Papka, and S. Kyriakides, “Biaxial Crushing of Honeycombs-Part I: Experiments”, International Journal of Solids and Structures, Vol. 36, No. 29, pp. 4367-4396, 1999. 8S. D. Papka, and S. Kyriakides, “In-Plane Biaxial Crushing of Honeycombs-Part II: Analysis”, International Journal of Solids and Structures, Vol. 36, No. 29, pp. 4397-4423, 1999. 9S. D. Papka, and S. Kyriakides, “In-Plane Compressive Response and Crushing of Honeycomb”, Journal of Mechanics and Physics of Solids, Vol. 42, No. 10, pp. 1499-1532, 1994. 10S. D. Papka, and S. Kyriakides, “In-Plane Crushing of Polycarbonate Honeycomb”, International Journal of Solids and Structures, Vol35, No. 3-4, pp. 239-267, 1998. 11D. Karagiozova, and T. X. Yu, “Plastic Deformation Modes of Regular Hexagonal Honeycombs Under In-Plane Biaxial Compression”, Int. J. of Mechanical Science, Vol. 46, pp. 1489-1515, 2004 12R. K. Mcfarland Jr., “Hexagonal Cell Structures Under Post-Buckling Axial Load”, AIAA Journal, Vol. 1, No. 6, pp. 1380-1385, 1963
13F. Cote, V. S. Deshpande, N. A. Fleck, and A. G. Evans, “The Out-of-Plane Compressive Behavior of Metallic Honeycombs”, Materials Science & Engineering, Vol. 380, pp. 272-280, 2004 14A. –J. Wang, and D. L. McDowell, “Yield Surfaces of Various Periodic Metal Honeycombs at Intermediate Relative Density”, Int. J. of plasticity, Vol. 21, pp. 285-320, 2005
15J. Hohe, and W. Becker, “Effective Elastic Properties of Triangular Grid Structures”, Composite Structures, Vol. 45, No. 2, pp. 131-145, 1999. 16J. Hohe, C. Beschorner, and W. Becker, “Effective Elastic Properties of Hexagonal and Quadrilateral Grid Structures”, Composite Structures, Vol. 46, No.1, pp. 73-89, 1999. 17J. Hohe, and W. Becker, “An Energetic Homogenisation Procedure for the Elastic Properties of General Cellular Sandwich Cores”, Composites, Part B: engineering, Vol. 32, No. 3, pp. 185-197, 2001. 18D. P. Makhecha, B. P. Patel, and M. Ganapathi, “Transient Dynamics of Thick Skew Sandwich Laminates Under Thermal/ Mechanical Loads”, Journal of Reinforced Plastics and Composites, Vol. 20, No. 17, pp. 1524-1545, 2001. 19S. P. Timoshenko and J. N. Goodier, Theory of Elasticity, 3rd Ed., McGraw-Hill, 1951, NY. 20J. N. Reddy, Mechanics of Laminated Composite Plates and Shells: Theory and Analysis, 2nd Ed., CRC Press, Boca Raton, NY, 2004 21J. G. Ren and E. Hinton, “The Finite Element Analysis of Homogeneous and Laminated Composite Plates Using a Simple Higher Order Theory”, Communications in Applied Numerical Methods, Vol. 2, pp. 217-228, 1986 22M. Ganapathi, B. P. Patel, and D. P. Makhecha, “Nonlinear Dynamic Analysis of Thick Composite/Sandwich Laminates Using an Accurate Higher-Order Theory”, Composites Part B: Engineering, Vol. 35, pp. 345-355, 2004
American Institute of Aeronautics and Astronautics
10
Appendix A Figure A1 shows the representative 2D volume element of the square honeycomb, the dissection of the that element into cell walls and the displacement field.
Figure A1: The 2D representative volume element of the square honeycomb
The total displacement field in a cell wall of length l, thickness t and height h is assumed to consist of three parts:
1. Homogenously distributed normal deformation in the x1 – x3 plane
3333
2332
11
)(
1)1(1)2(1
)(
1)1(1)2(1)1()(
xxu
xl
vvxu
xl
vvvxu
iI
iI
iI
ε
εν
ν
=
⎟⎠⎞
⎜⎝⎛ +
−−
−=
−+=
(A1)
2. Bending and shear deformation in the x1 – x2 plane
0)(21
6112)(
2112)(
3
413212
31132
213
2
31221131
=
⎟⎠⎞
⎜⎝⎛ +++=
⎟⎟⎠
⎞⎜⎜⎝
⎛+⎟
⎠⎞
⎜⎝⎛ ++−=
iII
iII
iII
xu
CxCxCxCthE
xu
xCthE
lCxCxCthE
xu α
(A2)
Where
1
2
4
3 5 v(1)1
v(1)2
v(5)1
v(5)2
Δφ(1)
Δφ(5)v(5)2
v(4)2 v(4)1
Δφ(4)
v(5)1
Δφ(5)Δφ(5)
v(3)1
v(3)2 Δφ(
v(5)2
v(5)1 3)v(5)2
v(5)1 Δφ(5)
Δφ(2)v(2)2
v(2)1
American Institute of Aeronautics and Astronautics
11
⎪⎪
⎩
⎪⎪
⎨
⎧−
=
=
⎟⎠⎞
⎜⎝⎛ Δ−Δ
++
−+
=
⎟⎠⎞
⎜⎝⎛ Δ−−Δ+−
−+
=
⎟⎠⎞
⎜⎝⎛ Δ+Δ+
−−
+=
theoryBernolli-Euler 0
theoryTimoshenko)1(5
12
2)1(121
)2(2
)1(2
22)1(2)2(112
1
)2()2()1()4(2)1(2)2(61
1121
)2()1(2)1(2)2(21
121
2
2
34
3
3
3
2
2
3
1
lt
vthEC
lvvthEC
lvv
lthEC
lvv
lthEC
ν
α
ϕαϕααα
ϕαϕαα
ϕϕα
3. Homogeneously distributed transverse shear deformation (x1 – x3 plane)
13
2
1
3)1(3)2(3)1()(
0)(
0)(
xl
vvvxu
xuxu
iIII
iIII
iIII
−+=
=
=
(A3)
Where E and ν in Eqs. A1 – A3 are the elastic constants of the cell material. The total displacement field is given by u = ui
I + uiII + ui
III. From the displacement field, the strain field of the cell wall can be obtained by partial differentiation. The stress field is then obtained from the strain field via Hooke’s law in conjunction with the plane stress assumption in the cell wall. The strain energy of the cell wall is then obtained as the volume integration of the sum of the products of the components of stress and strain. The result of this process lead to the following:
42
131212)1(
21
)1(11
)1(2 2
22
2
2
22 KKltK
ltKlthEW ννα
αν−
++⎟⎟⎠
⎞⎜⎜⎝
⎛−+
++
−= (A4)
Where
⎟⎟⎟⎟⎟⎟
⎠
⎞
⎜⎜⎜⎜⎜⎜
⎝
⎛
⎟⎟⎟
⎠
⎞
⎜⎜⎜
⎝
⎛
−−
⎟⎟⎟⎟⎟⎟
⎠
⎞
⎜⎜⎜⎜⎜⎜
⎝
⎛
=
33
T
33
1)2(
1)1(
111
-1-11)2(
1)1(
1
ενννν
εl
vl
v
lv
lv
K
American Institute of Aeronautics and Astronautics
12
⎟⎟⎟⎟
⎠
⎞
⎜⎜⎜⎜
⎝
⎛
⎟⎟⎠
⎞⎜⎜⎝
⎛−
−
⎟⎟⎟⎟
⎠
⎞
⎜⎜⎜⎜
⎝
⎛
=
⎟⎟⎠
⎞⎜⎜⎝
⎛ΔΔ
⎟⎟⎠
⎞⎜⎜⎝
⎛−
−⎟⎟⎠
⎞⎜⎜⎝
⎛ΔΔ
=
⎟⎟⎟⎟⎟⎟
⎠
⎞
⎜⎜⎜⎜⎜⎜
⎝
⎛
Δ
Δ
⎟⎟⎟⎟⎟⎟⎟⎟
⎠
⎞
⎜⎜⎜⎜⎜⎜⎜⎜
⎝
⎛
−
−−−
−
−
⎟⎟⎟⎟⎟⎟
⎠
⎞
⎜⎜⎜⎜⎜⎜
⎝
⎛
Δ
Δ=
lv
lv
lv
lv
K
K
lv
lv
lv
lv
K
3)2(
3)1(
1111
3)2(
3)1(
4
)2()1(
1111
)2()1(
3
)2(
2)2()1(
2)1(
21
21
41
21
211
211
41
21
41
21
211
211
)2(
2)2()1(
2)1(
2
T
T
T
φφ
φφ
φ
φ
φ
φ
To determine the nodal deflections, a relation between the stress resultants and the nodal deflections is needed. The stress resultants are obtained by differentiating eq. A4 with respect to the nodal deflections. The result is expressed in matrix form as follows:
(A5) 3323
3
222
2
)2()1(
)2()1(
ε⎟⎟⎠
⎞⎜⎜⎝
⎛+⎟⎟
⎠
⎞⎜⎜⎝
⎛⎟⎟⎠
⎞⎜⎜⎝
⎛=⎟⎟
⎠
⎞⎜⎜⎝
⎛KK
KKKK
FF 1
1
111
vv
Where are the nodal forces T)1)(,3)(,2)(,1)(()( iMiFiFiFi =F are the nodal deflections T))(,3)(,2)(,1)(()( iiviviviv φΔ= And the stiffness matrices are given as:
⎟⎟⎟⎟⎟⎟
⎠
⎞
⎜⎜⎜⎜⎜⎜
⎝
⎛
⎟⎟⎠
⎞⎜⎜⎝
⎛+
−−
=
12402/0
02/)1(002/00
0001
)1( 222
tll
l
lthE
ββ
νββ
ν11K
American Institute of Aeronautics and Astronautics
13
( )T2
2
22222
22221
2222
0000)1(
12402/0
02/)1(002/00
0001
)1(
12402/0
02/)1(002/00
0001
)1(
12402/0
02/)1(002/00
0001
)1(
ννν
ββ
νββ
ν
ββ
νββ
ν
ββ
νββ
ν
lll
thE
tll
l
lthE
tll
l
lthE
tll
l
lthE
−−
=⎟⎟⎠
⎞⎜⎜⎝
⎛
⎟⎟⎟⎟⎟⎟
⎠
⎞
⎜⎜⎜⎜⎜⎜
⎝
⎛
⎟⎟⎠
⎞⎜⎜⎝
⎛+−
−−
−=
⎟⎟⎟⎟⎟⎟
⎠
⎞
⎜⎜⎜⎜⎜⎜
⎝
⎛
⎟⎟⎠
⎞⎜⎜⎝
⎛−
−−−−
−
−=
⎟⎟⎟⎟⎟⎟
⎠
⎞
⎜⎜⎜⎜⎜⎜
⎝
⎛
⎟⎟⎠
⎞⎜⎜⎝
⎛−−
−−−
−
−=
3
13
1
KK
K
K
K
The aforementioned relations provide a linear system of equations, the solution of which provides the nodal deflections of the entire volume element.
American Institute of Aeronautics and Astronautics
14