Вы находитесь на странице: 1из 20

See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/303896046

An Improved solid-shell element based on ANS


and EAS concepts

Research · June 2016


DOI: 10.13140/RG.2.1.1585.3681

CITATIONS READS

0 123

1 author:

Mohammadreza Mostafa
University of Colorado Boulder
6 PUBLICATIONS 35 CITATIONS

SEE PROFILE

All content following this page was uploaded by Mohammadreza Mostafa on 10 June 2016.

The user has requested enhancement of the downloaded file.


INTERNATIONAL JOURNAL FOR NUMERICAL METHODS IN ENGINEERING
Int. J. Numer. Meth. Engng (2016)
Published online in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/nme.5260

An improved solid-shell element based on ANS and EAS concepts

M. Mostafa*,†
Department of Civil, Environmental and Architectural Engineering, University of Colorado Boulder, 428 UCB, Boulder,
CO 80309, USA

SUMMARY
This paper presents an eight-node nonlinear solid-shell element for static problems. The main goal of this
work is to develop a solid-shell formulation with improved membrane response compared with the previous
solid-shell element (MOS2013), presented in [1]. Assumed natural strain concept is implemented to account
for the transverse shear and thickness strains to circumvent the curvature thickness and transverse shear
locking problems. The enhanced assumed strain approach based on the Hu–Washizu variational principle
with six enhanced assumed strain degrees of freedom is applied. Five extra degrees of freedom are applied
on the in-plane strains to improve the membrane response and one on the thickness strain to alleviate the
volumetric and Poisson’s thickness locking problems. The ensuing element performs well in both in-plane
and out-of-plane responses, besides the simplicity of implementation. The element formulation yields exact
solutions for both the membrane and bending patch tests. The formulation is extended to the geometrically
nonlinear regime using the corotational approach, explained in [2]. Numerical results from benchmarks show
the robustness of the formulation in geometrically linear and nonlinear problems. Copyright © 2016 John
Wiley & Sons, Ltd.

Received 7 July 2015; Revised 26 March 2016; Accepted 30 March 2016

KEY WORDS: solid shell; EAS; ANS; membrane; nonlinear; corotational

1. INTRODUCTION

Shell elements have been widely used to model the behavior of thin structures in 3D for several
decades. These elements can be classified as degenerated shell elements with plane stress assump-
tion (as examples: [3–8]) and classical shell elements in which the thickness stretch is also taken
into account (as examples: [9–15]).
During the past decade, another class of finite elements, called solid shell, has gained popu-
larity among researchers. Solid-shell finite elements behave like shell elements but contain only
translational degrees of freedom (DOFs). The early works on solid shell are brought in [16–19].
The trilinear isoparametric shape functions are not capable of capturing the physical behavior of
thin-walled structures. This issue, identified in the literature as locking problem, causes the numer-
ical results to diverge from the analytical solutions. There are many locking problems that the
solid element with trilinear shape functions undergoes, which are summarized in the succeeding
discussions:
The Poisson’s thickness locking happens when the displacement is assumed to vary linearly in
the thickness direction, and as a result, the thickness strain becomes constant. However, because of
Poisson’s effect, the thickness strain is coupled with the in-plane strains that vary linearly across
the thickness. This discrepancy results in Poisson’s thickness locking. Potential remedies to address
this defect are as follows: (i) assuming a quadratic displacement distribution across the thickness,
which produces linear thickness strain in thickness direction [16]; (ii) using enhanced assumed strain

*Correspondence to: M. Mostafa, 428 UCB, University of Colorado Boulder, Boulder, CO 80309, USA.
† E-mail: mohammadreza.mostafa@colorado.edu

Copyright © 2016 John Wiley & Sons, Ltd.


M. MOSTAFA

(EAS) DOFs and enhancing the thickness strain to vary linearly across the thickness [20]; or (iii)
dividing the strain in the thickness direction into a membrane (constant) and bending (linear) part
and enforcing the bending stress to be zero [18].
The transverse shear locking occurs when the element thickness tends toward zero; physically
speaking, the element has a high aspect ratio in terms of length versus thickness. The EAS method
is one way to reduce the effect of parasitic shear terms (see references [21, 22] for small strains
and [23, 24] for large strains). Another remedy is the assumed natural strain (ANS) concept (see
references [25–27] for shell and [18, 20, 28–35] for solid-shell elements).
Another locking that occurs in solid-shell elements is curvature locking, referred to by some
authors as trapezoidal locking. This occurs when the element edges in the thickness direction are not
perpendicular to the element mid-plane. This type of locking happens in modeling curved structures
with solid-shell elements, where the thickness is tapered. The ANS concept has shown to circumvent
this issue as described in [10, 20, 36].
The volumetric locking occurs when the material is nearly incompressible. The EAS approach
[37–39], the B-bar method [40], and selective reduced integration method [41] have been applied to
remedy the volumetric locking of solid-shell elements in the literature. In another publication [42],
the EAS method was blended with the element-free Galerkin formulation in order to increase the
flexibility for nearly incompressible materials.
The last locking problem called the membrane locking happens when the element is subjected to
in-plane longitudinal or transverse (shear) loads and the low-order shape functions are not capable
of modeling the physical behavior of the element. As an example, the author of this paper used
the assumed natural deviatoric strain (ANDES) method, explained in reference [43], to alleviate the
membrane locking of the proposed solid-shell formulation in [1]. The ANDES method was applied
to the mid-surface with four nodes. If the mid-surface became warped, which was the case for many
curved structures with coarse mesh arrangement, the stiffness matrix would be polluted; that is, the
rigid body motions would produce nonzero in-plane strains (xx ; yy ; xy ). In order to circumvent
that problem, the projection operator, presented in [1], was used.
In this work, a full-integration solid-shell element is presented. Similar to the author’s previous
work (MOS2013), presented in [1], the ANS scheme is applied to alleviate the thickness and curva-
ture locking problems. The EAS scheme based on the Hu–Washizu variational principle [37] with
six enhanced DOFs is used. In lieu of using the ANDES method (presented in [1]), in this paper, five
enhanced DOFs are used to improve the element’s in-plane response. One DOF is used to circum-
vent the volumetric and Poisson’s thickness locking problems (similar to [1, 20]). The advantage of
EAS for the membrane response (as compared with the previous publication) is that there is no need
to apply the projection operator to the stiffness matrix to eliminate the parasitic strains. This will
enable the application of various kinematic descriptions to extend the formulation to the nonlinear
regime.
The following is the summary of the major differences between the linear element presented in
this paper and the solid-shell formulation (MOS2013) presented in [1]:
(i) In the current paper, the EAS method with five extra DOFs has been used to alleviate the
membrane locking problem. The main contribution of this paper is developing a formulation
that works satisfactorily in both in-plane and out-of-plane problems, especially when an in-
plane load is applied to the structure (Cook’s skew plate problem).
(ii) In the previous formulation, the in-plane response was decomposed from the out-of-plane
response, and the ANDES method was used for the in-plane response. The ANDES formu-
lation was applied on the mid-surface and was linearly interpolated across the thickness of
the element in the local coordinates. In the current work, the trilinear shape functions have
been used as the starting point, and no decomposition between the in-plane and out-of-plane
responses has been performed.
(iii) In the previous work, in the case of a warped mid-surface (relatively common for arched thin-
shell structures), the stiffness matrix would become polluted; that is, the rigid body motions
would produce nonzero membrane strains. In order to clean up the stiffness matrix, the

Copyright © 2016 John Wiley & Sons, Ltd. Int. J. Numer. Meth. Engng (2016)
DOI: 10.1002/nme
AN IMPROVED SOLID-SHELL ELEMENT BASED ON ANS AND EAS

projection operator was applied. In the current work, there is no need for the projection, because
the warped mid-surface does not produce any parasitic strains.
(iv) In the previous work, because the projection operator was applied in developing the linear
formulation, it was not possible to extend the formulation to the geometrically nonlinear regime
using the total Lagrangian (TL) approach, while it is feasible to use the TL method for the new
element to account for large deformations.
Many different approaches have been applied to formulating solid-shell elements with nonlinear
kinematics in the literature. For example, Schwarze and Reese [44] developed a reduced integra-
tion geometrically nonlinear element based on TL kinematics, while Meraim and Combescure [45,
46] used an updated Lagrangian (UL) approach. In the previous work in [1], a corotational (CR)
kinematic description was used to extend the element to the nonlinear regime. Other examples of
geometrically nonlinear solid-shell formulations are brought in [33, 47–49]. In this paper, the solid-
shell element formulation is extended to the geometrically nonlinear regime using a CR kinematic
description. The CR kinematic description was pioneered by Wempner [50] and Belytschko and co-
workers [51]. Among different CR formulations in the literature (as an example, see [52] for the
history of the formulations), the element-independent approach known as ‘EICR’ is selected, which
was first introduced by Rankin and Brogan [53]. This approach is based on the use of projection
operators. This approach has been widely used on shells (as examples: [54–57]) and continua (as
examples: [1, 58, 59]). The EICR scheme is of interest because the linear element kernel is reused
in developing the geometrically nonlinear element and the element library need not to be drastically
modified.
Similar to [2], in the current work, the corotated coordinate is defined by minimizing the defor-
mations within the frame. The orientation of the corotated frame is defined using quaternions and
is posed as finding the smallest eigenvalue of a positive semi-definite 4  4 matrix. The element
internal force vector and consistent tangent stiffness matrix are derived by taking variations of the
internal strain energy with respect to nodal DOFs to preserve the consistency requirement. The ensu-
ing exact geometric stiffness matrix does not involve the second derivative of the rotation function
and is always symmetric for elements with translational DOFs.
In the next sections, the element linear and nonlinear formulations are explained, and several
numerical examples are presented. The results are compared with MOS2013 and other solid-shell
and shell formulations in the literature.

2. LINEAR ELEMENT FORMULATION

The element geometry is shown in Figure 1. One of the directions is defined as the shell thickness
direction or wall direction. Surfaces extending along two other directions are in-plane surfaces, and
directions contained in an in-plane surface are in-plane directions.

2.1. Global to local coordinate transformation (R)


Referring to Figure 1, we denote the global and local coordinates as X and x; respectively. Point C
is chosen to be the origin of the local coordinate system, whose global coordinates XC are given by
the average of the nodal coordinates.

y
Top 8
Surface Thickness
5 4 7 direction
Y
3
6 x
X 1
Z Bottom
2
Surface

Figure 1. The solid-shell element geometry.

Copyright © 2016 John Wiley & Sons, Ltd. Int. J. Numer. Meth. Engng (2016)
DOI: 10.1002/nme
M. MOSTAFA

The local coordinates are defined using the methodology explained in [2]. The local coordinate
system is defined such that the local z-axis is along the thickness direction. This is accomplished
by minimizing the sum of the squared distances of the nodes from the x–y plane. If Q́ were the
component in the global coordinate system of a unit vector along the local z-axis, then the local
z-coordinate of node n would be Q́ T .Xn  XC /. Thus, Q́ is given by
Q́ D argmin T A subject to  T  D 1; (1)


where
Nnd D8
X
AD .Xn  XC /.Xn  XC /T : (2)
nD1

Q́ is the normalized eigenvector corresponding to the smallest eigenvalue of A. The other two
eigenvectors of A give the local x and y axes in such a way that x:.
Q yQ  Q́ / becomes positive. The R
matrix takes the following form:
 
R D xQ yQ Q́ : (3)
Node positions and displacements are then transformed from global coordinates to local coordinates
by
xn D RT .Xn  XC / and un D RT Un : (4)

2.2. Stiffness matrix computation


The Cartesian position vectors (x) and displacement vectors (u) are approximated using trilinear
ansatz functions. The natural (contravariant) base is chosen to approximate the displacement and
(covariant) compatible strains.
The Cartesian and natural strains in Voigt notation are as follows:
 T  T
e C D exx ; eyy ; xy ; e´´ ; y´ ; x´ and e n D e ; e ;  ; e  ;  ;  : (5)
The Cartesian strains (e C ) and natural strains (e n ) are related by the following well-known
transformation:
e C D T .; ; /e n
with
2 2 2 2 3
j11 j21 j11 j21 j31 j21 j31 j11 j31
6 j122 2
j22 j12 j22 2
j32 j22 j32 j12 j32 7
6 7
6 2j11 j12 2j21 j22 j12 j21 C j11 j22 2j31 j32 j22 j31 C j21 j32 j12 j31 C j11 j32 7
6
T D6 7: (6)
2 2 2 7
6 j13 j23 j13 j23 j33 j23 j33 j13 j33 7
4 2j12 j13 2j22 j23 j13 j22 C j12 j23 2j32 j33 j23 j32 C j22 j33 j13 j32 C j12 j33 5
2j11 j13 2j21 j23 j13 j21 C j11 j23 2j31 j33 j23 j31 C j21 j33 j13 j31 C j11 j33
The T matrix
 h contains the coefficients i of the inverse of the Jacobian matrix
J.; ; / D @x.;;
@
/ @x.;; / @x.;; /
; @
; @
[60].
In order to avoid thickness curvature locking, the ANS concept is applied, which was also used by
Schwarze et al. [20]. e  is evaluated at four collocation points (K D A,B,C,D shown in Figure 2),
and then the evaluated strains at the collocation points are interpolated by the use of bilinear shape
functions within the element as follows:
D
X
eANS
 D NN A eA C NN B eB C NN C eC C NN D eD D NN K eK
KDA (7)
1
with NN K D .1 C K /.1 C K /
4
Copyright © 2016 John Wiley & Sons, Ltd. Int. J. Numer. Meth. Engng (2016)
DOI: 10.1002/nme
AN IMPROVED SOLID-SHELL ELEMENT BASED ON ANS AND EAS

H
D L

K
C
A E
M G

F
B
J

Figure 2. Collocation points for the assumed natural strain approach.

K , K , and K are the natural coordinates corresponding to each of the considered collocation
points.
To alleviate transverse shear locking, the work of Cardoso et al. [33] is followed in which the
natural compatible shear strains ( ;  ) are evaluated at four collocation points and interpolated
with bilinear shape functions as follows:
H
X
ANS E F G H K
 D NN E  C NN F  C NN G  C NN H  D NN K 
KDE
1
with NN K D .1 C K /.1 C K /
4
and (8)
M
X
ANS J K L M K
 D NN J  C NN K  C NN L  C NN M  D NN K 
KDJ
1
with NN K D .1 C K /.1 C K /:
4
For  , the collocation points are E, F, G, and H, and for  , the collocation points are J, K, L,
and M, as shown in Figure 2.
One extra DOF is used to enhance the thickness strain (´´ ) to alleviate the Poisson’s thickness
locking and five to enhance the in-plane strains.
The total Cartesian strain vector is obtained by
   .e/ 
  u.e/ .e/ u
e C D B C Benh C
D B
˛ .e/ ˛ .e/
2 3
 0 0 0  0
6 0  0 0  07 (9)
6 7
C 6 0 0    2  2 07
with Benh D T 0 Benh
n n
and Benh D6 7;
60 0 0 0 0 7
40 0 0 0 0 05
0000 0 0

where ˛ .e/ is the vector containing the six extra DOFs, T 0 is the transformation matrix (6) evaluated
at  D  D  D 0; and B C is the compatible strain–displacement operator derived from the trilinear
functions.
n
The first three rows of Benh , which correspond to the in-plane response, are picked from Simo’s
EAS quadrilateral element in [37]. Following Simo’s work, the formulation is stable owing to the
n
fact that the columns of Benh are linearly independent, and it passes the constant strain patch test
because the following relation is satisfied:
Z
n
Benh d D 0: (10)


Copyright © 2016 John Wiley & Sons, Ltd. Int. J. Numer. Meth. Engng (2016)
DOI: 10.1002/nme
M. MOSTAFA

Figure 3. Summary of the assumed strain fields. ANS, assumed natural strain.

Equation (10) guarantees that the subspace of approximation contains piecewise constant stress
fields, which insures the satisfaction of the constant strain patch test [37]. The reason that the formu-
lation yields excellent convergence for in-plane bending problems with distorted mesh arrangement
is that the derivation of the first three rows of the matrix in Equation (9) has started from the original
Wilson et al. incompatible modes [61].
The
R local stiffness matrix K.e/
L for the element is obtained by applying the kinematic equation (9)
in V .B / EB.e/ dV and doing the static condensation to eliminate the rows and columns
.e/ T

corresponding to the extra DOFs (˛ .e/ ). E is the linear elastic material constitutive matrix.
The development of the assumed and enhanced strain fields for the linear element is summarized
in Figure 3.
2.3. Local to global stiffness matrix
In this part, the local stiffness matrix (K.e/
L ) is transformed to the global stiffness matrix by

K.e/ .e/ T
Gs D diag.R/KL diag.R / (11)

3. NONLINEAR ELEMENT FORMULATION

In the next step, the solid-shell element formulation is extended to the geometrically nonlinear
regime using the EICR [53]. This approach is based on decomposing the motion of an element
into rigid body and deformation components. An attractive feature of the EICR is that elements
that perform well in geometrically linear problems can be extended to nonlinear problems in an
element-independent fashion.
Deformations are described with reference to a coordinate frame that translates and rotates with
the element referred to as ‘corotated frame’. There are several methods for defining the coro-
tated frame (see [2] for the literature review). Similar to the method explained in [2], the ‘best-fit’

Copyright © 2016 John Wiley & Sons, Ltd. Int. J. Numer. Meth. Engng (2016)
DOI: 10.1002/nme
AN IMPROVED SOLID-SHELL ELEMENT BASED ON ANS AND EAS

corotated frame is defined by minimizing the deformations within the frame. The orientation of
the corotated frame is defined using quaternions and is posed as finding the smallest eigenvalue
of a positive semi-definite 4  4 matrix. The ensuing exact geometric stiffness matrix does not
involve the second derivative of the rotation function and is always symmetric for elements with
translational DOFs.

4. NUMERICAL EXAMPLES

In this section, some well-known numerical benchmarks are analyzed to investigate the performance
of the proposed solid-shell element for linear and nonlinear problems.

4.1. Linear examples


4.1.1. Patch tests. As explained in the element formulation, the enhanced strains have been defined
in such a way that the orthogonality of the assumed stress field to the enhanced strain field is guar-
anteed, which is the requirement for passing the patch test [37]. The element passes both membrane
and bending patch tests proposed by MacNeal and Harder [62] as explained as follows.

4.1.1.1. Membrane patch test. Referring to Macneal and Harder [62], a solid element of length
L D 0:24 mm, width B D 0:12 mm, and thickness t D 0:001 mm shown in Figure 4 is meshed with
five distorted elements.
A displacement state is applied on the structure that leads to a constant in-plane strain distri-
bution. The material behavior is modeled as isotropic linear elastic with material properties as
E D 106 N/mm2 and  D 0:25. The displacements of the exterior nodes are prescribed by linear
functions:
 
Y X
UX D 103 X C ; UY D 103 Y C ; (12)
2 2
and the UZ D 0 is prescribed for the bottom layer of nodes. Assuming the aforementioned dis-
placements at exterior nodes, a constant strain distribution leading to a plane stress state is obtained.
The displacements of the nodes are shown in Table I. The results agree exactly with the solution of
Equation (12), and the stresses at Gauss points are
XX D Y Y D 1333:33 N=mm2 ; XY D 400:0 N=mm2 ; (13)

Nodal Coordinates (mm):


1:(0.04,0.02,0.0005)
4 2:(0.18,0.03,0.0005)
3 3:(0.16,0.08,0.0005)
Z
8 4:(0.08,0.08,0.0005)
Y 1 7 5:(0.04,0.02,-0.0005)
2 6:(0.18,0.03,-0.0005)
( , , )
5
X 6 7:(0.16,0.08,-0.0005)
B
8:(0.08,0.08,-0.0005)

Figure 4. Patch test geometry.

Table I. Membrane patch test, the


finite element solution results.

Node UX (mm) UY (mm)


1 5.000e5 4.000e5
2 1.950e4 1.200e4
3 2.000e4 1.600e4
4 1.200e4 1.200e4

Copyright © 2016 John Wiley & Sons, Ltd. Int. J. Numer. Meth. Engng (2016)
DOI: 10.1002/nme
M. MOSTAFA

which agrees with the analytical solution. All other stresses are zero in this plane stress problem.
Hence, the membrane patch test is passed.

4.1.1.2. Bending patch test. In this test, the patch shown in Figure 4 is tested to reproduce a constant
bending stress state. The displacements of exterior nodes are
 
103 Y 103 X 103
2
UX D t XC ; UY D t Y C ; UZ D X C X Y C Y 2 ; (14)
2 2 2 2 2

which shows a linear strain distribution in thickness direction. The displacement of the internal
nodes is shown in Table II. The numerical results produce the analytical solutions exactly.
In order to analyze the stresses at top and bottom surfaces, the values at Gauss points are
extrapolated across the thickness. The resulting stresses at top and bottom surfaces are as follows:
XX D Y Y D ˙0:6667 N=mm2 ; XY D ˙0:200 N=mm2 ; (15)
which conforms with the analytical solution. Hence the bending patch test is also passed.

4.1.2. Pinched cylinder with end diaphragms. The test assesses the performance of the element for
inextensional bending and complex membrane modes. The cylinder is pinched by two diametrically
opposite point loads as shown in Figure 5(a). The analytical solution for the vertical displacement
under the point load is UY D 1:8248  105 mm [63].

Table II. Bending patch test, the finite element


solution results.

Node UX (mm) UY (mm) UZ (mm)


1 2.500e8 2.000e8 1.400e6
2 9.750e8 6.000e8 1.935e5
3 1.000e7 8.000e8 2.240e5
4 6.000e8 6.000e8 9.600e6

Figure 5. Pinched cylinder with end diaphragms. (a) Geometry and mesh arrangement and (b) convergence
study considering regular mesh.

Table III. Normalized deflections for the pinched cylinder (distorted mesh).

Distortion angle D 0ı D 10ı D 20ı D 30ı D 40ı


Present element 0:965 0:962 0:926 0:925 0:915
MOS2013 [1] 0.970 0.963 0.926 0.925 0.924

Copyright © 2016 John Wiley & Sons, Ltd. Int. J. Numer. Meth. Engng (2016)
DOI: 10.1002/nme
AN IMPROVED SOLID-SHELL ELEMENT BASED ON ANS AND EAS

In Figure 5(b), the vertical displacements under the point load are normalized to the analyti-
cal solution and compared with some cited solid-shell (MRESS [33], H1=ME9 [64], RESS [65],
HCiS12 [21], ANS  , and ANS  -HS [28]) and shell (ANS -EAS) [10] formulations. Compared
with other existing solid-shell and shell elements, the current element has a better rate of conver-
gence followed by MRESS, which is an ANS element.
The sensitivity of the results to the mesh distortion is also studied and presented in Table III along
with the results for MOS2013 solid-shell formulation presented in [1] . The distortion is identified
as angle ; which is shown in Figure 5(a). The mesh arrangement of 20  20 is used to do the
sensitivity test. In this test, the distorted mesh arrangement causes the membrane locking problem,
but using the EAS scheme for the membrane response of the element alleviates the membrane
locking problem. The results show that both formulations perform well in case of distorted mesh
arrangements; however, MOS2013 performs slightly better than the present formulation.

4.1.3. Scordelis–Lo roof. The goal of this test is to assess the membrane and bending behaviors of
the proposed solid-shell element formulation. As shown in Figure 6(a), the roof is mounted on two
rigid diaphragms at two curved edges and loaded under its own weight. Owing to symmetry, only
a quarter of the cylinder is modeled. The boundary conditions of the rigid diaphragms are defined
by UX D UY D 0. The analytical solution for the vertical displacement of point A under the
volume load is UY D 0:3024 mm [62]. The vertical displacements of point A are normalized to
the reference analytical value and presented in Table IV along with the results for MOS2013 and
some existing elements.
Table IV shows that the proposed element converges from above and MOS2013 converges slightly
better than the present formulation.

4.1.4. Hemispherical shell with 18ı hole. The goal is to assess the performance of the formulation
for inextensional membrane and bending modes. The geometry of the model is shown in Figure 6(b).
The equator accounts for a free edge, and the XZ and YZ planes are the symmetry planes.

Figure 6. (a) Scordelis–Lo roof and (b) hemisphere with 18ı hole.

Table IV. Scordelis–Lo roof, normalized displacements of point A.

ne Current MOS2013 Schw09 Alv05 Kim05 Are03


work [1] (Q1STs) [20] (RESS) [65] (XSolid85) [30] [23]
44 1:049 1.047 0.997 0.995 0.960 1.029
88 1:013 1.011 0.994 0.986 0.984 1.001
16  16 0:996 0.996 1.000 0.993 0.999 0.992
32  32 0:998 0.997 1.001 0.996 — 0.991

Copyright © 2016 John Wiley & Sons, Ltd. Int. J. Numer. Meth. Engng (2016)
DOI: 10.1002/nme
M. MOSTAFA

Table V. Hemispherical shell with hole, normalized displacements of point A.

Current work MOS2013 Schw09 (Q1STs) Ree07 (Q1SPs) Kim05 (XSolid85) Are03
ne [1] [20] [66] [30] [23]
44 1:019 1.018 1.043 0.062 1.058 0.040
88 0:995 0.997 1.002 0.723 1.005 0.756
16  16 0:985 0.986 0.993 0.919 — 0.991
32  32 0:987 0.990 0.994 0.969 — 0.999

Figure 7. Full hemisphere problem, geometry.

Table VI. Full hemispherical shell, Normalized displacement of point A.

ne Current work MOS2013 [1] Schw09 (Q1STs) Ree07 (Q1SPs) Kim05 (XSolid85) Gru05 Car02
4 0:793 0.663 0.418 0.104 — 0.573 0.680
8 0:972 0.962 0.956 0.630 1.079 0.971 0.980
16 0:998 0.997 0.996 0.907 1.014 1.002 0.990
32 0:999 1.002 0.999 0.970 1.000 1.000 —

The displacements of node A are normalized to the reference value UX D 0:0940 mm [62]
and presented in Table V along with some other existing formulations. The proposed element and
MOS2013 yield the best results for coarse mesh arrangement as compared with other cited elements.

4.1.5. Full hemispherical shell. In this example, the previous test is investigated without the 18ı
hole as shown in Figure 7. This is a more demanding test on the formulation because the element is
bending about the diagonal plane. In Table VI, the displacements of point A are normalized to the
reference value UX D 0:0924 mm [62] along with some other existing solid-shell (Q1STs, Q1SPs,
XSolid85, and MOS2013 [1]) and shell (Gru05 and Car02) formulations.
The results show that the proposed element exhibits a faster convergence rate compared with
MOS2013 solid-shell and other cited elements.

4.1.6. Twisted thin beam. The twisted beam benchmark shown in Figure 8 proposed by MacNeal
and Harder [62] is to assess the warping performance of the element. In this work, the thin beam
with thickness t D 0:05 mm proposed by Simo [5] is investigated. The tip displacements in loading
direction are normalized to the reference analytical values and presented in Tables VII and VIII
along with the results for MOS2013 and some existing formulations.
Numerical results show that the proposed element converges slightly faster than MOS2013 and
other cited formulations.

Copyright © 2016 John Wiley & Sons, Ltd. Int. J. Numer. Meth. Engng (2016)
DOI: 10.1002/nme
AN IMPROVED SOLID-SHELL ELEMENT BASED ON ANS AND EAS

Figure 8. Twisted thin beam.

Table VII. Twisted beam (reference solution for in-plane load D 1:387 mm).

ne Current work MOS2013 Schw09 (Q1STs) Car08 (MRESS) Alv05 (RESS) Simo89
[1] [20] [33] [65] [5]
61 1:004 1.005 1.002 — — 0.993
12  2 1:000 1.001 0.998 0.965 0.998 1.000
24  4 0:999 1.000 0.999 1.000 1.000 1.001
48  8 0:999 1.000 1.000 1.001 1.000 1.002

Table VIII. Twisted beam (reference solution for out-of-plane load D 0:343 mm).

ne Current work MOS2013 Schw09 (Q1STs) Car08 (MRESS) Alv05 (RESS) Simo89
[1] [20] [33] [65] [5]
61 0:951 0.953 0.942 — — 0.951
12  2 0:989 0.988 0.983 0.958 0.985 0.986
24  4 0:995 0.995 0.995 0.995 0.996 0.997
48  8 0:998 1.000 0.999 0.999 0.998 1.000

4.1.7. Clamped square plate. This test is to assess the element sensitivity to distorted mesh. The
geometry is shown in Figure 9(a). The side length is L D 100 mm, and the thickness is t D 1 mm.
The material behavior is modeled as isotropic linear elastic with material properties as Young’s
modulus E D 1:0104 N=mm2 and Poisson’s ratio  D 0:30. A concentrated force with magnitude
F D 16:3527 N is applied at the center of the plate in the negative Z direction. Owing to symmetry,
only a quarter of the plate is modeled. Point A is moved by e (0 6 e 6 12 mm) as depicted in the
Figure 9(a). In this example, the model is discretized by 2  2 elements. The analytical solution
corresponding to the explained geometry and applied load is provided in [67] as
12.1   2 /
U´ D 0:00560FL2 D 1:000 mm. (16)
Et 3
The center displacements are normalized with respect to the analytical solution and are plot-
ted versus the e values along with the results from some existing solid-shell formulations
(MOS2013 [1], Shwarz [20], and Kli06 [31]) in Figure 9(b). As is shown in the figure, all elements
have satisfactory responses for highly distorted meshes. It is noticeable that among the reported
elements, the element proposed in [20] behaves slightly better for very distorted mesh arrange-
ment. Moreover, MOS2013 performs slightly better than the current work for the distorted mesh
arrangement.

4.1.8. Cook’s skew plate. This test was first proposed by Cook [68] and applied on shell formu-
lation by Simo in [5]. The goal is to assess the membrane behavior of the element. As shown in
Figure 10(a), the trapezoidal beam is clamped at one end and subjected to distributed shear force

Copyright © 2016 John Wiley & Sons, Ltd. Int. J. Numer. Meth. Engng (2016)
DOI: 10.1002/nme
M. MOSTAFA

Figure 9. Clamped square plate. (a) Geometry and (b) sensitivity study on distorted mesh.

Figure 10. Cook’s skew plate. (a) Geometry and (b) convergence study.

at the tip. The analytical solution for this problem is UYA D 23:91. In Figure 10(b), the verti-
cal displacements of point A are presented along with the results from some existing solid-shell
(MOS2013 [1], MRESS [33], and RESS [65]) and shell (iCYSE-E4 [69]) formulations. For coarse
mesh arrangement, the current formulation has better performance compared with MOS2013 ele-
ment and other cited solid-shell formulations. The reason that the formulation performs better than
other cited formulations is that the first three rows of the matrix in Equation (9) that correspond to
the membrane strains have been developed starting from the incompatible modes of Wilson et al.
[61]. iCYSE shell with four enhanced strain variables has the same convergence rate, but it does not
pass the membrane patch test.

4.1.9. MacNeal’s cantilever thin beam with distorted mesh. This benchmark tests the performance
of the formulation in case of distorted mesh arrangement across the thickness of the structure. As
shown in Figure 11, the cantilever beam is modeled using mesh arrangement with high aspect ratio
together with distortion in the form of a trapezium. A shear force with magnitude of 1:0 N is applied
at the end. The material behavior is modeled as isotropic linear elastic with Young’s modulus E D
106 N=mm2 and Poisson’s ratio values of  D 0:0 and  D 0:3. The depth of the beam is d D 0:2 m.
Numerical solutions obtained using different elements are presented in Table IX. It is observed
that using ANS approach with four collocation points circumvents the trapezoidal locking and the
element works as well as the cited elements. This test also shows that the additional enhancing DOF
in the thickness direction helps alleviate the Poisson’s thickness locking problem.

Copyright © 2016 John Wiley & Sons, Ltd. Int. J. Numer. Meth. Engng (2016)
DOI: 10.1002/nme
AN IMPROVED SOLID-SHELL ELEMENT BASED ON ANS AND EAS

Figure 11. MacNeal’s thin beam.

Table IX. Tip deflection for MacNeal’s thin beam problem.

Displacement D 0o D 45o
 D 0:0  D 0:3  D 0:0  D 0:3
ANS3Dq-at [70] 0.1073 0.1071 0.1074 0.1071
M19k3DEAS-at [70] 0.1081 0.1065 0.1076 0.1061
Present work 0.1081 0.1066 0.1076 0.1063

Figure 12. Twisted beam under out-of-plane loading (geometrically nonlinear problem). (a) Deformed shape
and (b) load displacement of point A.

4.2. Geometrically nonlinear examples


In all the nonlinear examples, MOS2013 in [1] is extended to geometrically nonlinear regime,
using the approach presented in [2]. Hence, in the following examples, MOS2013 refers to the CR
nonlinear solid-shell formulation.

4.2.1. Nonlinear twisted beam under out-of-plane loading. In this test, the performance of the non-
linear formulation is assessed in case of warped geometry. The geometric and material parameters as
well as the boundary conditions are the same as example 4.1.6. The geometry is shown in Figure 8.
An out-of-plane load with magnitude of F D 60 N is applied at the free end of the beam (point A). In
this example, the twisted beam with thickness t D 0:05 mm is investigated because the convergence
study of the linear formulation is presented for the beam with the same thickness.
Similar to [2], a 24  4 mesh is considered, and the SC8R of ABAQUS with mesh of 96  8 is
chosen as the reference solution. The maximum load is applied in 15 equal increments.
The displacements UX , UY , and UZ of the point A are plotted versus the load in Figure 12(b)
along with the reference solution and the results presented in [2] for nonlinear MOS2013. It is
noted that the only difference between the current work and the work in [2] is the linear element
formulation that has been used for the material stiffness matrix.
Considering the CR kinematics, both nonlinear elements yield excellent agreement with the ref-
erence solution (SC8R) with UL kinematics; however, the current formulation yields slightly better
results than the previous formulation presented in [2]. The deformed shape of the twisted beam is
shown in Figure 12(a).

Copyright © 2016 John Wiley & Sons, Ltd. Int. J. Numer. Meth. Engng (2016)
DOI: 10.1002/nme
M. MOSTAFA

Figure 13. Slit annular plate subjected to lifting line force. (a) Geometry, (b) deformed shape, and (c)
-
displacement curves for points A and B.

4.2.2. Slit annular plate subjected to lifting line force. This test is another benchmark to assess the
performance of the thin-shell structures under finite rotations. The internal and external radii of the
structure are Ri D 6 mm and Re D 10, respectively. The shell thickness is t D 0:03 mm. The
material behavior is modeled as isotropic linear elastic with Young’s modulus E D 21106 N=mm2
and Poisson’s ratio  D 0:0.
The undeformed geometry of the structure is shown in Figure 13(a). The circular ring has a slit
cut along the radial direction A–B, where a vertical lifting line force F D 0:8 N/mm is applied on
one edge of the slit and the other edge is fully clamped. Similar to [71], the reference solution is
selected to be 10  80 using S4R shell element of ABAQUS. In the current work, the intermediate
mesh of 8  64 is used to model the structure; 20 equal increments are used to plot the response
curve; however, the same results are achieved by only 10 increments.
The vertical displacements of points A and B are plotted versus the load multiplier
in
Figure 13(c) along with the reference solution and the results from the nonlinear MOS2013. Both
nonlinear solid-shell formulations with CR kinematics conform well with the reference solution of
ABAQUS with UL kinematic description.
The deformed shape of the ring is also shown in Figure 13(b), which shows the large displacement
and rotation of the free edge.

Copyright © 2016 John Wiley & Sons, Ltd. Int. J. Numer. Meth. Engng (2016)
DOI: 10.1002/nme
AN IMPROVED SOLID-SHELL ELEMENT BASED ON ANS AND EAS

Figure 14. Pinching of a clamped cylinder. (a) Geometry, (b) deformed shape, and (c) load–displacement
curves for point under the load.

4.2.3. Pinching of a clamped cylinder. In this test, a cantilevered cylinder is subjected to two oppo-
site loads at its open end. Owing to symmetry, only a quarter of the structure is modeled as shown
in Figure 14(a). The radius of the structure is R D 1:016 mm, the length is L D 3:048 mm, and the
thickness is t D 0:03 mm. This is a demanding test for the solid-shell elements because they suf-
fer from trapezoidal locking in modeling the curved structures. The material behavior is modeled
as isotropic linear elastic with Young’s modulus E D 2:0685  107 N=mm2 and Poisson’s ratio
 D 0:30.
The total load P D
P0 (load multiplier
and P0 D 1600 N) is applied in 20 equal increments,
and the tip displacement is controlled up to a total displacement of 1.6 times the radius of the
cylinder. It is noted that, in this problem, the largest physically possible displacements of the points
under the load are equal to the radius of the shell.
The results of current work are compared with those of the nonlinear MOS2013 and Stander’s
element [72]. The results are shown in Figure 14(c). The force–displacement curves for a regular
2424 mesh arrangement show that both nonlinear element formulations have very good agreement
with the results from Stander et al., with regular 32  32 mesh arrangement. The deformed finite
element mesh at the maximum load value is shown in Figure 14(b).

4.2.4. Clamped-hinged deep circular arch subjected to a concentrated load. In this problem, a deep
circular arch undergoes a point load at its apex as is shown in Figure 15(a). The radius of the arch is
R D 100 mm, the thickness is t D 1 mm, and the width is w D 24 mm. The structure is hinged at
one end and fixed at the other end; this allows for non-symmetric buckling. The material behavior is
modeled as isotropic linear elastic with Young’s modulus E D 0:5105 N=mm2 and Poisson’s ratio
 D 0:0. The total load value applied on the apex is F D 1000 N. This example has been addressed

Copyright © 2016 John Wiley & Sons, Ltd. Int. J. Numer. Meth. Engng (2016)
DOI: 10.1002/nme
M. MOSTAFA

Figure 15. Clamped-hinged deep circular arch subjected to a concentrated load. (a) Geometry, (b) deformed
shape, and (c) load–displacement curves for point under the load.

in [1, 46, 73]. Because of the lack of symmetry, the entire structure is modeled with mesh of 40  1
following the cited references. In Figure 15(c), the horizontal (UX ) and vertical (UY ) displacements
of the point under the load are plotted versus the incremental load, and the results are compared
with the nonlinear MOS2013 and SHB8PS in [46]. For this problem, 50 equal increments are used
to produce the response curve.
The responses of current CR solid-shell element and the nonlinear MOS2013 are identical, and
both conform with those of cited element with UL kinematic description; however, the plots show
that the post-buckling regime starts slightly later than the cited element (at load value of F D 990 N
compared with F D 950 N for SHB8PC in [46]). The schematic deformed shape of the arch is
shown in Figure 15(b).

5. CONCLUDING REMARKS

In the present work, a geometrically nonlinear solid-shell element, suitable for capturing the non-
linear response of thin-shell structures is presented. A CR kinematic description is used to extend
the formulation to the geometrically nonlinear regime. The advantage of the CR approach is the
applicability of the linear element library in the nonlinear formulation.
For the linear element formulation, in order to circumvent the transverse shear locking and curva-
ture locking problems in the thickness direction, the ANS concept has been used. The EAS concept
with six extra DOFs has been used in the element, one to alleviate the Poisson’s thickness and
volumetric locking problems and five to improve the in-plane response of the element.
In Section 4, the linear element formulation has been assessed by a number of famous benchmarks
proposed for shell-like finite elements. The element passes both membrane and bending patch tests.
Moreover, the element’s performance has been tested for mesh distortion, warped mesh arrange-

Copyright © 2016 John Wiley & Sons, Ltd. Int. J. Numer. Meth. Engng (2016)
DOI: 10.1002/nme
AN IMPROVED SOLID-SHELL ELEMENT BASED ON ANS AND EAS

ment, and structure geometries with single and double curvature. It was evident from the pinched
cylinder example with end diaphragms that the thickness-tapered element has a poor convergence
rate for inextensional bending problems; however, the pinched cylinder example shows that the cur-
rent linear element formulation has higher convergence rate compared with other cited shell and
solid-shell formulations.
Cook’s skew plate example shows that the EAS approach with five extra strain variables improves
the element response in the case of in-plane (membrane) loads compared with the author’s previous
work (MOS2013 [1]). The MacNeal’s thin beam test shows that the element performs well in the
case of distorted mesh arrangements across the thickness direction. The other advantage of the
EAS approach compared with the ANDES method [1, 43] is that there is no need for applying
the projection operator to clean up the stiffness matrix when the mid-surface is warped.
Compared with other cited linear solid-shell elements that pass the patch tests (membrane
and bending), the proposed formulation yields remarkable results for both out-of-plane and in-
plane responses, which is due to applying the EAS and ANS approaches together in developing
the element.
The nonlinear examples prove that once the linear formulation yields satisfactory results for linear
problems, using a CR kinematic description, it will also show satisfactory results for the nonlinear
problems if the strains remain within the small strain range. The comparison between the current
nonlinear element and nonlinear MOS2013 shows that both formulations generate very close results
for different nonlinear problems. This is because both of them yield very similar results for linear
problems that involve the out-of-plane response of the element.
Although linear elasticity and geometrically nonlinear elasticity have been assumed in this work,
extension to materially nonlinear regime such as elasto-plasticity is also possible. This development
is under investigation and will be presented in a separate publication.

ACKNOWLEDGEMENTS

The author of this paper gratefully acknowledges the financial support from the National Sci-
ence Foundation through the grant CMMI-0847053. The author is also grateful to professor
Mettupalayam Sivaselvan for the guidance, support, and instruction as his PhD adviser.

REFERENCES
1. Mostafa M, Sivaselvan M, Felippa C. A solid-shell corotational element based on andes, ANS and EAS for
geometrically nonlinear structural analysis. International Journal for Numerical Methods in Engineering 2013;
95(2):145–180.
2. Mostafa M, Sivaselvan M. On best-fit corotated frames for 3D continuum finite elements. International Journal for
Numerical Methods in Engineering 2014; 98(2):105–130.
3. Belytschko T, Lin JI, Tsay CS. Explicit algorithms for the nonlinear dynamics of shells. Computer Methods in
Applied Mechanics and Engineering 1984; 42(2):225–51.
4. Bathe KJ, Dvorkin EN. A four-node plate bending element based on Mindlin/Reissner plate theory and a mixed
interpolation. International Journal for Numerical Methods in Engineering 1985; 21(2):367–83.
5. Simo JC, Fox DD. On a stress resultant geometrically exact shell model. I. Formulation and optimal parametrization.
Computer Methods in Applied Mechanics and Engineering 1989; 72(3):267–304.
6. Simo JC, Rifai MS, Fox DD. On a stress resultant geometrically exact shell model. IV. Variable thickness shells with
through-the-thickness stretching. Computer Methods in Applied Mechanics and Engineering 1990; 81(1):91–126.
7. Simo JC, Kennedy JG. On a stress resultant geometrically exact shell model. V. Nonlinear plasticity: formulation and
integration algorithms. Computer Methods in Applied Mechanics and Engineering 1992; 96(2):133–71.
8. Ćesar de Sá JMA, Natal Jorge RM, Fontes Valente RA, Almeida Areias PM. Development of shear locking-free
shell elements using an enhanced assumed strain formulation. International Journal for Numerical Methods in
Engineering 2002; 53(7):1721–50.
9. Betsch P, Gruttmann F, Stein E. A 4-node finite shell element for the implementation of general hyperelastic 3D-
elasticity at finite strains. Computer Methods in Applied Mechanics and Engineering 1996; 130(1-2):57–79.
10. Bischoff M, Ramm E. Shear deformable shell elements for large strains and rotations. International Journal for
Numerical Methods in Engineering 1997; 40(23):4427–49.
11. Brank B, Korelc J, Ibrahimbegovic A. Nonlinear Shell Problem Formulation Accounting for Through-the-thickness
Stretching and Its Finite Element Implementation. Saxe-coburg Publications: Stirling, UK, 2002 33–62.
12. Cardoso RPR, Yoon JW. One point quadrature shell element with through-thickness stretch. Computer Methods in
Applied Mechanics and Engineering 2005; 194(9-11):1161–99.

Copyright © 2016 John Wiley & Sons, Ltd. Int. J. Numer. Meth. Engng (2016)
DOI: 10.1002/nme
M. MOSTAFA

13. Lu H, Ito K, Kazama K, Namura S. Development of a new quadratic shell element considering the normal stress
in the thickness direction for simulating sheet metal forming. Journal of Materials Processing Technology 2006;
171(3):341–7.
14. Sansour C. A theory and finite element formulation of shells at finite deformations involving thickness change:
circumventing the use of a rotation tensor. Archive of Applied Mechanics 1995; 65:194–216.
15. Sansour C, Kollmann F. Families of 4-node and 9-node finite elements for a finite deformation shell theory. An
assesment of hybrid stress, hybrid strain and enhanced strain elements. Computational Mechanics 2000; 24:435–447.
16. Parisch H. A continuum-based shell theory for non-linear applications. International Journal for Numerical Methods
in Engineering 1995; 38(11):1855–83.
17. Miehe C. A theoretical and computational model for isotropic elastoplastic stress analysis in shells at large strains.
Computer Methods in Applied Mechanics and Engineering 1998; 155(3-4):193–233.
18. Hauptmann R, Schweizerhof K. Systematic development of ‘solid-shell’ element formulations for linear and non-
linear analyses employing only displacement degrees of freedom. International Journal for Numerical Methods in
Engineering 1998; 42(1):49–69.
19. Hauptmann R, Schweizerhof K, Doll S. Extension of the solid-shell concept for application to large elastic and large
elastoplastic deformations. International Journal for Numerical Methods in Engineering 2000; 49(9):1121–41.
20. Schwarze M, Reese S. A reduced integration solid-shell finite element based on the EAS and the ANS concept-
geometrically linear problems. International Journal for Numerical Methods in Engineering 2009; 80(10):1322–55.
21. Alves de Sousa RJ, Natal Jorge RM, Fontes Valente RA, Ćesar de Sá JMA. A new volumetric and shear locking-free
3D enhanced strain element. Engineering Computations 2003; 20(7):896–925.
22. Andelfinger U, Ramm E. EAS-elements for two-dimensional, three-dimensional, plate and shell structures and their
equivalence to HR-elements. International Journal for Numerical Methods in Engineering 1993; 36(8):1311–1337.
23. Areias PMA, Ćesar de Sá JMA, Conceicão António CA, Fernandes AA. Analysis of 3D problems using a
new enhanced strain hexahedral element. International Journal for Numerical Methods in Engineering 2003;
58(11):1637–82.
24. Fontes Valente RA, Alves De Sousa RJ, Natal Jorge RM. An enhanced strain 3D element for large deformation
elastoplastic thin-shell applications. Computational Mechanics 2004; 34(1):38–52.
25. Hughes TJR, Tezduyar TE. Finite elements based upon Mindlin plate theory with particular reference to the four-node
bilinear isoparametric element. Journal of Applied Mechanics 1981; 48(3):587–96.
26. Wempner G, Talaslidis D, Hwang CM. A simple and efficient approximation of shells via finite quadrilateral
elements. Journal of Applied Mechanics 1982; 49(1):115–20.
27. Dvorkin EN, Bathe KJ. Continuum mechanics based four-node shell element for general non-linear analysis.
Engineering Computations 1984; 1(1):77–88.
28. Sze KY, Yao LQ. A hybrid stress ANS solid-shell element and its generalization for smart structure modelling.
Solid-shell element formulation. International Journal for Numerical Methods in Engineering 2000; 48(4):545–64.
29. Vu-Quoc L, Tan XG. Optimal solid shells for non-linear analyses of multilayer composites. I. Statics. Computer
Methods in Applied Mechanics and Engineering 2003; 192(9-10):975–1016.
30. Kim KD, Liu GZ, Han SC. A resultant 8-node solid-shell element for geometrically nonlinear analysis. Computa-
tional Mechanics 2005; 35(5):315–331.
31. Wagner W, Klinkel S, Gruttmann F. A robust non-linear solid shell element based on a mixed variational formulation.
Computer Methods in Applied Mechanics and Engineering 2006; 195(1-3):179–201.
32. Felippa CA. The SS8 solid shell element a Fortran implementation, March 2002.
33. Cardoso RPR, Yoon JW, Mahardika M, Choudhry S, Alves de Sousa RJ, Fontes Valente RA. Enhanced assumed
strain (EAS) and assumed natural strain (ANS) methods for one-point quadrature solid-shell elements. International
Journal for Numerical Methods in Engineering 2008; 75(2):156–187.
34. Edem I, Gosling P. One-point quadrature ans solid-shell element based on a displacement variational formulation
part I geometrically linear assessment. Computer methods in Applied Mechanics and Engineering 2012; 237-240:
177–191.
35. Norachan P, Suthasupradit S, Kim K. A co-rotational 8-node degenerated thin-walled element with assumed natural
strain and enhanced assumed strain. Finite Elements in Analysis and Design 2012; 50:70–85.
36. Betsch P, Stein E. An assumed strain approach avoiding artificial thickness straining for a non-linear 4-node shell
element. Communications in Numerical Methods in Engineering 1995; 11(11):899–909.
37. Simo JC, Rifai MS. Class of mixed assumed strain methods and the method of incompatible modes. International
Journal for Numerical Methods in Engineering 1990; 29(8):1595–1638.
38. Simo JC, Armero F. Geometrically non-linear enhanced strain mixed methods and the method of incompatible
modes. International Journal for Numerical Methods in Engineering 1992; 33(7):1413–1449.
39. Simo JC, Armero F, Taylor RL. Improved versions of assumed enhanced strain tri-linear elements for 3D finite
deformation problems. Computer Methods in Applied Mechanics and Engineering 1993; 110(3-4):359–386.
40. Hughes TJR. Generalization of selective integration procedures to anisotropic and nonlinear media. International
Journal for Numerical Methods in Engineering 1980; 15(9):1413–1418.
41. Hughes TJR. The Finite Element Method: Linear Static and Dynamic Finite Element Analysis. Dover Publications:
Mineola, NY, 2000.
42. Garca A, Cardoso R, Yoon J. Subspace analysis to alleviate the volumetric locking in the 3D solid-shell efg method.
Journal of Computational and Applied Mathematics 2013; 246:185–194.

Copyright © 2016 John Wiley & Sons, Ltd. Int. J. Numer. Meth. Engng (2016)
DOI: 10.1002/nme
AN IMPROVED SOLID-SHELL ELEMENT BASED ON ANS AND EAS

43. Mostafa M, Sivaselvan M, Felippa C. Reusing linear finite elements in material and geometrically nonlinear analysis
application to plane stress problems. Finite elements in analysis and design 2013; 69(2):62–72.
44. Schwarze M, Reese S. A reduced integration solid-shell finite element based on the EAS and the ANS concept-large
deformation problems. International Journal for Numerical Methods in Engineering; 85(3):289–329.
45. Abed-Meraim F, Combescure A. SHB8PS-A new adaptative, assumed-strain continuum mechanics shell element for
impact analysis. Computers and Structures 2002; 80(9-10):791–803.
46. Abed-Meraim F, Combescure A. An improved assumed strain solid-shell element formulation with physical stabiliza-
tion for geometric non-linear applications and elastic-plastic stability analysis. International Journal for Numerical
Methods in Engineering 2009; 80(13):1640–1686.
47. Alves de Sousa RJ, Yoon JW, Cardoso RPR, Fontes Valente RA, Grácio JJ. On the use of a reduced enhanced
solid-shell (RESS) element for sheet forming simulations. International Journal of Plasticity 2007; 23(3):490–515.
48. Alves de Sousa RJ, Cardosa RPR, Fontes Valente RA, Yoon JW, Grácio JJ, Natal Jorge RM. New one-point quadra-
ture enhanced assumed strain (EAS) solid-shell element with multiple integration points along thickness - part II.
Nonlinear applications. International Journal for Numerical Methods in Engineering 2006; 67(2):160–88.
49. Sze KY, Chan WK, Pian THH. An eight-node hybrid-stress solid-shell element for geometric non-linear analysis of
elastic shells. International Journal for Numerical Methods in Engineering 2002; 55(7):853–78.
50. Wempner G. Finite elements, finite rotations and small strains of flexible shells. International Journal of Solids and
Structures 1969; 5(2):117–54.
51. Belytschko T, Hsieh BJ. Non-linear transient finite element analysis with convected co-ordinates. International
Journal for Numerical Methods in Engineering 1973; 7(3):255–71.
52. Felippa CA, Haugen B. A unified formulation of small-strain corotational finite elements: I. Theory. Computer
Methods in Applied Mechanics and Engineering 2005; 194(21-24 SPEC. ISS.):2285–2335.
53. Rankin CC, Brogan FA. An element independent corotational procedure for the treatment of large rotations.
Transactions of the ASME. Journal of Pressure Vessel Technology 1986; 108(2):165–74.
54. Haugen B. Buckling and stability problems for thin shell structures using high performance finite elements. PhD
Thesis, Colorado At Boulder, 1994.
55. Jiang L, Chernuka MW. Simple four-noded corotational shell element for arbitrarily large rotations. Computers and
Structures 1994; 53(5):1123–1132.
56. Stolarski H, Belytschko T, Sang-Ho L. A review of shell finite elements and corotational theories. Computational
Mechanics Advances 1995; 2(2):125–212.
57. Battini JM. A modified corotational framework for triangular shell elements. Computer Methods in Applied
Mechanics and Engineering 2007; 196(13-16):1905–14.
58. Crisfield MA, Moita GF. A co-rotational formulation for 2-D continua including incompatible modes. International
Journal for Numerical Methods in Engineering 1996; 39(15):2619–33.
59. Moita GF, Crisfield MA. A finite element formulation for 3-D continua using the co-rotational technique.
International Journal for Numerical Methods in Engineering 1996; 39(22):3775–92.
60. Bathe KJ. Finite Element Procedures. Prentice Hall: Englewood Cliffs, N.J., 1996.
61. Wilson E, Taylor R, Doherty W, Ghaboussi J. Incompatible displacement models. Technical Report, 1973. Numerical
and computer models in structural mechanics, Academic press New York.
62. MacNeal RH, Harder RL. Proposed standard set of problems to test finite element accuracy. Finite Elements in
Analysis and Design 1985; 1(1):3–20.
63. Belytschko T, Liu WK. Test problems and anomalies in shell finite elements. Reliability of Methods for Engineering
Analysis Proceedings of the International Conference, Pineridge Press, Swansea, UK, 1986; 393–406.
64. Kasper EP, Taylor RL. A mixed-enhanced strain method. I. Geometrically linear problems. Computers and Structures
2000; 75(3):237–50.
65. Alves de Sousa RJ, Cardoso RPR, Fontes Valente RA, Yoon JW, Grácio JJ, Natal Jorge RM. New one-point
quadrature enhanced assumed strain (EAS) solid-shell element with multiple integration points along thick-
ness: part 1-geometrically linear applications. International Journal for Numerical Methods in Engineering 2005;
62(7):952–77.
66. Reese S. A large deformation solid-shell concept based on reduced integration with hourglass stabilization.
International Journal for Numerical Methods in Engineering 2007; 69(8):1671–716.
67. Timoshenko S, Woinowsky-Krieger S. Theory of Plates and Shells (2nd edn.) McGraw-Hill: New York, 1959.
68. Cook R, Malkus D, Plesha M. Concepts and Applications of Finite Element Analysis (3rd edn.) John Wiley and Sons:
New York, 1989.
69. Cardoso RPR, Yoon JW, Fontes Valente RA. A new approach to reduce membrane and transverse shear locking for
one-point quadrature shell elements: linear formulation. International Journal for Numerical Methods in Engineering
2005; 66(2):214–49.
70. Hauptmann R, Doll S, Schweizerhof K. Solid-shell elements with linear and quadratic shape functions at large
deformations with nearly incompressible materials. Computers and Structures 2001; 79(1):1671–1685.
71. Sze KY, Liu XH, Lo SH. Popular benchmark problems for geometric nonlinear analysis of shells. Finite Elements in
Analysis and Design 2004; 40(11):1551–69.
72. Stander N, Matzenmiller A, Ramm E. Assessment of assumed strain methods in finite rotation shell analysis.
Engineering Computations 1989; 6(1):58–66.
73. Klinkel S, Wagner W. Geometrical non-linear brick element based on the EAS-method. International Journal for
Numerical Methods in Engineering 1997; 40(24):4529–4545.

Copyright © 2016 John Wiley & Sons, Ltd. Int. J. Numer. Meth. Engng (2016)
DOI: 10.1002/nme

View publication stats

Вам также может понравиться