Professional Documents
Culture Documents
Adina Fea PDF
Adina Fea PDF
YNAMIC
NCREMENTAL
ONLINEAR
NALYSIS
December 2012
ADINA R & D, Inc. owns both this software program system and its documentation. Both the
program system and the documentation are copyrighted with all rights reserved by ADINA R & D,
Inc.
ADINA R & D, Inc. makes no warranty whatsoever, expressed or implied that the Program and its
documentation including any modifications and updates are free from errors and defects. In no
event shall ADINA R & D, Inc. become liable to the User or any party for any loss, including but
not limited to, loss of time, money or goodwill, which may arise from the use of the Program and its
documentation including any modifications and updates.
Trademarks
All other product names are trademarks or registered trademarks of their respective owners.
Copyright Notice
Table of contents
1. Introduction ......................................................................................................... 1
1.1 Objective of this manual.................................................................................. 1
1.2 Supported computers and operating systems .................................................. 2
1.3 Units ................................................................................................................ 2
1.4 ADINA System documentation....................................................................... 4
2. Elements ............................................................................................................... 7
2.1 Truss and cable elements................................................................................. 7
2.1.1 General considerations ............................................................................. 7
2.1.2 Material models and formulations............................................................ 9
2.1.3 Numerical integration............................................................................... 9
2.1.4 Mass matrices ......................................................................................... 10
2.1.5 Element output ....................................................................................... 11
2.1.6 Recommendations on use of elements.................................................... 12
2.1.7 Rebar elements ....................................................................................... 12
2.2 Two-dimensional solid elements................................................................... 15
2.2.1 General considerations ........................................................................... 15
2.2.2 Material models and formulations.......................................................... 23
2.2.3 Numerical integration............................................................................. 24
2.2.4 Mass matrices ......................................................................................... 26
2.2.5 Element output ....................................................................................... 27
2.2.6 Recommendations on use of elements.................................................... 33
2.3 Three-dimensional solid elements................................................................. 37
2.3.1 General considerations ............................................................................ 37
2.3.2 Material models and nonlinear formulations.......................................... 45
2.3.3 Numerical integration............................................................................. 47
2.3.4 Mass matrices ......................................................................................... 50
2.3.5 Element output ....................................................................................... 50
2.3.6 Recommendations on use of elements.................................................... 56
2.4 Beam elements .............................................................................................. 58
2.4.1 Beam geometry and cross-sections ......................................................... 59
2.4.2 Standard beam element............................................................................ 71
2.4.2.1 Kinematics........................................................................................ 71
2.4.2.2 Linear formulation........................................................................... 73
2.4.2.3 Materially-nonlinear-only formulations .......................................... 73
2.4.2.4 Large displacement formulation...................................................... 73
2.4.2.5 Mass matrices .................................................................................. 77
2.4.2.6 Elastic beam element....................................................................... 78
2.4.2.7 Elastic-plastic beam element ........................................................... 81
7.3.4 Modal damping ratios based on strain energy proportional damping .. 856
7.4 Nonlinear dynamic analysis ........................................................................ 860
7.4.1 Step-by-step implicit time integration .................................................. 861
7.4.2 Time history by mode superposition .................................................... 863
7.4.3 Global mass matrix............................................................................... 864
7.5 Choosing between implicit and explicit formulations................................. 865
7.6 Tracking solution progress .......................................................................... 868
7.6.1 .rto file ................................................................................................... 868
7.7 Solution monitoring..................................................................................... 870
8. Explicit dynamic analysis ............................................................................... 873
8.1 Formulation ................................................................................................. 875
8.1.1 Mass matrix .......................................................................................... 877
8.1.2 Damping ............................................................................................... 878
8.2 Stability ....................................................................................................... 878
8.3 Time step management................................................................................ 881
8.4 Tracking solution progress .......................................................................... 882
9. Frequency domain analysis ............................................................................ 885
9.1 Response spectrum analysis ........................................................................ 885
9.2 Fourier analysis ........................................................................................... 896
9.3 Harmonic vibration analysis........................................................................ 899
9.4 Random vibration analysis .......................................................................... 906
9.5 SDOF system response................................................................................ 911
10. Fracture mechanics ....................................................................................... 925
10.1 Overview ................................................................................................... 925
10.2 Crack geometry ......................................................................................... 926
10.3 The energy release rate.............................................................................. 929
10.4 The line contour integration method .......................................................... 930
10.5 The virtual crack extension method .......................................................... 933
10.6 Modeling of crack geometry ..................................................................... 937
10.6.1 Specification of input ......................................................................... 937
10.6.2 2-D models ......................................................................................... 938
10.6.2.1 No nodes on the line of self-similar crack advance..................... 940
10.6.2.2 Nodes on the line of self-similar crack advance.......................... 940
10.6.3 3-D models, CRACK-M features are not used................................... 941
10.6.3.1 No mapped mesh on the plane of self-similar crack advance ......... 943
10.6.3.2 Mapped mesh on the plane of self-similar crack advance............... 944
10.6.4 3-D models, CRACK-M features are used......................................... 946
10.6.5 Meshing recommendations................................................................. 956
1. Introduction
1.1 Objective of this manual
The objective of this manual is to give a concise summary and
guide for study of the theoretical basis of the finite element
computer program ADINA Solids & Structures (ADINA in short).
The program ADINA is used for displacement and stress analysis.
Since a large number of analysis options is available in this
computer program, a user might well be initially overwhelmed with
the different analysis choices and the theoretical bases of the
computer program. A significant number of publications referred
to in the text (books, papers and reports) describe in detail the finite
element analysis procedures used in the program. However, this
literature is very comprehensive and frequently provides more
detail than the user needs to consult for the effective use of
ADINA. Furthermore, it is important that a user can identify easily
which publication should be studied if more information is desired
on a specific analysis option.
The intent with this Theory and Modeling Guide is
1.3 Units
When using the ADINA system, it is important to enter all
physical quantities (lengths, forces, masses, times, etc.) using a
consistent set of units. For example, when working with the SI
system of units, enter lengths in meters, forces in Newtons, masses
in kg, times in seconds. When working with other systems of units,
all mass and mass-related units must be consistent with the length,
force and time units. For example, in the USCS system (USCS is
an abbreviation for the U.S. Customary System), when the length
unit is inches, the force unit is pound and the time unit is second,
the mass unit is lb-sec2/in, not lb.
Rotational degrees of freedom are always expressed in radians.
Most angular input parameters are expressed in degrees.
Table 1.3-1 gives some of the more commonly used units
needed for ADINA input.
1) All program versions have parallelized solvers. Only ADINA and ADINA-
Thermal have parallelized assembly.
3) The Compaq Visual Fortran 6.6 compiler is no longer available on the market.
Installation Notes
Describes the installation of the ADINA System on your
computer. Depending on the platform, the appropriate
installation notes in pdf format can be printed or downloaded
from http://www.adina.com
2. Elements
2.1 Truss and cable elements
• This chapter outlines the theory behind the different element
classes, and also provides details on how to use the elements in
modeling. This includes the materials that can be used with each
element type, their applicability to large displacement and large
strain problems, their numerical integration, etc.
Z l
l
l
l
Y
X (u, v, w) are global displacement
degrees of freedom (DOF)
Z
i vi
l
Y
(b) 1-node ring element
ref. KJB
Sections Figure 2.1-1: Truss elements available in ADINA
5.3.1,
6.3.3
• Note that the only force transmitted by the truss element is the
longitudinal force as illustrated in Fig. 2.1-2. This force is constant
in the 2-node truss and the ring element, but can vary in the 3- and
4-node truss (cable) elements.
Z P
Y
X
area A
s
Stress constant over
P cross-sectional area
Young's modulus E
Z area A
P
R
R
Z 1 radian Y
l
Y
s (hoop stress)
X
Hoop strain = v/R Elastic stiffness = EA/R (one radian)
Circumferential force P = sA
• The local node numbering and natural coordinate system for the
2-node, 3-node and 4-node truss elements are shown in Fig. 2.1-3.
r r=0 r
r=-1 r=+1 r=-1 r=+1
l l l l l
1 2 1 3 2
2-node element 3-node element
r=-1/3 r r=1/3
r=-1 r=1
l l l l
1 3 4 2
4-node element
r r
l l l l
1 1 2
One-point integration Two-point integration
r r
l l l l
1 2 3 1 2 3 4
Three-point integration Four-point integration
See Section 13.1.1 for the definitions of those variables that are not
self-explanatory.
• The AUI can generate truss elements that are connected to the
2D or 3D solid elements in which the truss elements lie. There are
several separate cases, corresponding to the type of the truss
element group and the type of the solid element group in which the
truss elements lie.
l l
l l l l l l
l
l l
l l l l l l
(a) Before data file generation (b) After data file generation
Figure 2.1-6: Rebar in 3D truss, planar 2D solid elements
l l l l
l
l l l l l l
l
l l l l l l
l
l l l l
l
3D solid element l Truss element
(a) Before data file generation (b) After data file generation
Figure 2.1-7: Rebar in 3D truss, 3D solid elements
AUI: Click the Define Element Group icon to open the Define
Element Group window. In this window, check that the Element
Type has been set to “Truss”. Click on the Advanced tab. Select
the “Use as Rebars” option from the Element Option drop-down,
and then fill in the Rebar-Line Label box below the Element
Option drop-down box.
l
l
l
l l
l
L
l
l
Recommendation:
l
l l
l
l l
l
l l l
l
l l
l l l
l l l l l
l l
z l l s xx = 0 l l e xx = 0
z
y s xy = 0 g xy = 0
x y
l l s xz = 0 x l l g xz = 0
Z l l s xx = 0 (Stresses computed
s xy = 0 in element local
X Y l l s xz = 0 coordinate system)
• The basic finite element assumptions are, see Fig. 2.2-3, for the
coordinates
q q
y = ∑ hi yi ; z = ∑ hi zi
i =1 i =1
u = xU p − x( y − y p )Φ zp + x( z − z p )Φ yp
q q
1 2 p 1 2 p
v = ∑ hi vi + x Φz w = ∑ hi wi − x Φy
i =1 2 i =1 2
where
hi(r,s) = interpolation function corresponding to node i
(r,s) = isoparametric coordinates
q = number of element nodes, 3 ≤ q ≤ 9, excluding
auxiliary node in generalized plane strain
yi, zi = nodal point coordinates
vi, wi = nodal point displacements
U p , Φ yp , Φ zp = degrees of freedom of the auxiliary node
xp, yp, zp = coordinates of the auxiliary node
Z
s
2l l
5 1
l
w4
3l l
7 4 v4
l
z4
X Y
y4
Np Fyp
Up
Z, w w4 w3
4 v4 3 v3
Y, v
w1 v w2 v2
X, u 1
1 2
ε xx = U p − ( y − y p )Φ zp + ( z − z p )Φ yp
Number of default
2-D solid element
pressure DOFs
3-node triangle 1*
4-node quadrilateral (4/1) 1
6-node triangle (6/1) 1
7-node triangle (7/3) 3
8-node quadrilateral (8/1) 1
9-node quadrilateral (9/3) 3
*
To use this element, the user must use the displacement-
based element with 1-point integration.
4-node elements,
134.5 Hz 358.0 Hz
no incompatible modes
4-node elements, 108.6 Hz 294.0 Hz
incompatible modes
107.1 Hz 279.9 Hz
8-node elements
These element types are described on pp. 369-370 of ref. KJB. The
strain singularities are obtained by collapsing one side of an 8-node
rectangular element and not correcting the interpolation functions.
The 6-node spatially isotropic triangle is obtained by correcting
the interpolation functions of the collapsed 8-node element. It then
uses the same interpolation functions for each of the 3 corner nodes
and for each of the midside nodes. Note that when the corrections
to the interpolation functions are employed, only the element side
with end nodes 1 and 4 can be collapsed.
The 3-node triangular element is obtained by collapsing one
side of the 4-node element. This element exhibits the constant
strain conditions (except that the hoop strain in axisymmetric
analysis varies over the element) but is usually not effective.
ref. KJB
• The basic continuum mechanics formulations are described in
Sections 6.2 and ref. KJB, pp. 497-537, and the finite element discretization is given
6.3.4 in ref. KJB pp. 538-542, 549-555.
ref. KJB • Note that all these formulations can be used together in a single
Section 6.8.1 finite element mesh. If the elements are initially compatible, then
they will remain compatible throughout the complete analysis.
ref. KJB
• For the calculation of all element matrices and vectors,
Sections 5.5.3, numerical Gauss integration is used. You can use from 2×2 to 6×6
5.5.4 and 5.5.5 Gauss integration. The numbering convention and the location of
the integration points are given in Fig. 2.2-5 for 2×2 and 3×3 Gauss
integration. The convention used for higher orders is analogous.
The default Gauss integration orders for rectangular elements are
2×2 for 4-node elements and 3×3 otherwise.
s
INR=2 INS=2 s INS=3
INR=3
INR=1 INR=2
INR=1 INS=2
r r
INS=1
INS=1
a) Rectangular elements
2
1
1 4
3
r
1-point 4-point
3 3 4
2 6 7 8
5 2 10
6 12
1 7
4 1 11 13
5 9
7-point 13-point
b) Triangular elements
You can request that ADINA either print or save stresses or forces.
CREEP_STRAIN(XYZ), THERMAL_STRAIN(XYZ),
ELEMENT_TEMPERATURE,
ACCUMULATED_EFFECTIVE_PLASTIC_STRAIN,
FE_EFFECTIVE_STRESS, YIELD_STRESS,
EFFECTIVE_CREEP_STRAIN
ACCUMULATED_EFFECTIVE_PLASTIC_STRAIN,
THERMAL_STRAIN(XYZ), ELEMENT_TEMPERATURE,
VOID_VOLUME_FRACTION
IRRADIATION_STRAIN(XYZ).
with similar definitions for the other abbreviations used above. But
the variable DEFORMATION_GRADIENT(XYZ) is interpreted as
follows:
DEFORMATION_GRADIENT(XYZ) =
DEFORMATION_GRADIENT-YY,
DEFORMATION_GRADIENT-ZZ,
DEFORMATION_GRADIENT-XX,
DEFORMATION_GRADIENT-YZ,
DEFORMATION_GRADIENT-ZY
• You can also request that strain energy densities be output along
with the stresses.
F = ∫ BT τ dV
V
Axis of revolution
l l l
p
l l
CL l l l
l
l
l
l l
l
1 unit
l
l
l
l
p
Mz p
My
Z
Np
Y Fp
X
1 radian 1 radian
l
l l
l l
l l
l l
l
l
l l
l l l
L
t
a) Axisymmetric idealizations
¥
P
l l l
l l l l l
l l l 1 unit l l t
l l l
L
¥
b) Plane strain idealization (long plate) c) Plane stress idealization (cantilever)
l l
l l l l l l l l
l
l l l l l l l
l
l l l
l l l l l l l
l l l l l l l
l l l l
l
l l l l l l l
l l l l l l l l l
l l l l l l l
l
l l l l l l l l l
l l l l
l l
l ll l
l l
l
l l l
l l
l l l l
l l l l
l l l l
l l l l l l
l l
l
l l
l l l
l
l l
l
l l l
l
l
l l
l l
l
l l
l l
l
l
l
l
l l l
l
l l l
l l
l
l l l
l l
l
l
l
l l
l l
l l l l
l
l
l l
l
l l
l l l
l l
l l
l l
l l l l
l l l l l
l l
l l
l
E,A
P
P
Plane sections do
B not remain plane.
A dB dA ¹ dB
L
dA
(c) 3-D model
q q q
u = ∑ hi ui v = ∑ hi vi w = ∑ hi wi
i =1 i =1 i =1
where
hi (r, s, t) = interpolation function corresponding to node i
r, s, t = isoparametric coordinates
q = number of element nodes, 4 ≤ q ≤ 27
xi, yi, zi = nodal point coordinates
ui, vi, wi = nodal point displacements
No axial displacement
should be possible Line loads One quarter of
(uniform) the structure
is modeled
l
l l
l
l
l l
l
l l
l
l
Y Z
X
Z 2-D model
Y
a) Plane strain 2-D analysis applicable
l
l l
l l l l
l l l
l l l
l
l l
l
l l
l l
Fixed end
Z 3-D model
Y
Table 2.3-1: Mixed u/p formulations available for 3-D solid elements
Number of default
3-D solid element
pressure DOFs
4-node tetrahedron 1*
10-node tetrahedron (10/1) 1
11-node tetrahedron (11/4) 4
6-node prism 1
15-node prism 1
21-node prism 4
8-node brick (8/1) 1
20-node brick (20/1) 1
21-node brick 4
27-node brick (27/4) 4
5-node pyramid 1
13-node pyramid 1
14-node pyramid 1
*
To use this element, the user must use the displacement-
based element with 1-point integration.
l l
l l l
l l 1 radian
Z l l l
l l
l l l of structure
is modeled
X
Y
Z Axisymmetric model
Y
c) Axisymmetric analysis applicable
Structure axisymmetric,
loading non-axisymmetric 3-D elements
l
l l l l l
l l
l l
l l l
l
l l
l
l l
l l
One quarter
Z
of structure
Y is modeled
d) Three-dimensional model is used
t
Z 3
l
W4 l
10
11
l
l 26 2
4 V4 19 l
l l 23
12 l l
l l 9
24 21 l 1 18
l l l
U4 l 20 22
7 l
l
l l 14
15 25 l 17 l
8l l 6
27 s
l
l 13
Z8 r 16
l
5
X8 Y
Y8
X
Figure 2.3-4: Conventions used for the nodal coordinates
and displacements of the 3-D solid element
ref. KJB • Note that all these formulations can be used in one finite
Section 6.8.1 element mesh. If the elements are initially compatible, then they
will remain compatible throughout the analysis.
INR=1,INS=1,INT=3
s
Ä
Ä
Ä
Ä Ä
Ä
Ä Ä
Ä
Ä
Ä
Ä
INR=1,INS=1,INT=1
r
INR=2,INS=1,INT=1
t t
5
1 1 4
s 2 s
1-point r r 5-point
3
16 17
4 15 5
4 13
5 5
4 s r
8 6 12 14 9 10
3 2 2 3
7 2 11
t t
with similar definitions for the other abbreviations used above. But
the variable DEFORMATION_GRADIENT(XYZ) is interpreted as
follows:
DEFORMATION_GRADIENT(XYZ) =
DEFORMATION_GRADIENT-XX,
DEFORMATION_GRADIENT-XY,
DEFORMATION_GRADIENT-XZ,
DEFORMATION_GRADIENT-YX,
DEFORMATION_GRADIENT-YY,
DEFORMATION_GRADIENT-YZ,
DEFORMATION_GRADIENT-ZX,
DEFORMATION_GRADIENT-ZY,
DEFORMATION_GRADIENT-ZZ,
See Section 13.1.1 for the definitions of those variables that are
not self-explanatory.
ref. KJB • In ADINA, the stresses are calculated using the strains at the
Section 4.2.1 point of interest. Hence, they are not spatially extrapolated or
smoothed. The AUI can be employed to calculate smoothed
stresses from the results output by ADINA.
• You can also request that strain energy densities be output along
with the stresses.
F = ∫ BT τ dV
V
The nodal forces are accessible in the AUI using the variable
names NODAL_FORCE-X, NODAL_FORCE-Y,
NODAL_FORCE-Z.
Dam
Vessel
Tunnel
shells, the use of the 3-D solid element usually results in too stiff a
model and a poor conditioning of the stiffness matrix. In this case,
the use of the shell or the plate/shell elements (see Sections 2.7 and
2.6) is more effective.
• Fig 2.4-1 shows the beam element along with its local
coordinate system (r,s,t). The r direction always lies along the
neutral line of the beam. The orientation of the s and t directions is
defined using either an auxiliary node K (as in Fig. 2.4-1(a)) or an
orientation vector (XO,YO,ZO) (as in Fig. 2.4-1(b)). If node K is
specified in the input, it is always used as the auxiliary node. If
node K is not specified, then the orientation vector is used.
When the orientation vector is used, the components
(XO,YO,ZO) are interpreted either in the skew system of local
node 1 (the default), or in the global system.
Notice that, for the Hermitian beam element, (r,s,t) are not
isoparametric coordinates, rather (r,s,t) have the same units as the
global coordinates.
X Y
a : warping DOF
X Y
Figure 2.4-1: Degrees of freedom and local axes for beam element
Fig 2.4-1 also shows the degrees of freedom at the local nodes.
These degrees of freedom are defined in the local coordinate
system. The α degree of freedom is used only for the warping
beam element, see Section 2.4.3.
s
S11
S 14
S 10
S5 S8
S7
l
S2
S1 2 S9
S4 l S12
S13
1 S3
S6 t
Z
X Y
t
Node 1
s
Node 2
r
(a) Isometric view
t
ct O
.
r cs s
C
A : Cross-sectional area
L : Beam length
(do not confuse the polar moment of inertia I rr with the Saint-
Venant torsional constant J )
I st : Product of inertia:
I st = ∫ stdA
A
If s-t axes are coincident with the principal axes of the section then
I st = 0 .
Cross-sections
Tables 2.4-1 and 2.4-2 show the beam capabilities available for the
available cross sections.
Table 2.4-1: Table of beam capabilities for the available cross-sections, standard
beam element
1. Use of the standard beam element is not recommended for these cross-sections.
Table 2.4-2: Table of beam capabilities for the available cross-sections, warping
beam element
1. The kinematics of the linear elastic warping beam element have been improved
between version 8.7 and 8.8, so the results obtained may also change.
2. In version 8.7 of the ADINA system, it was allowed to use the large displacement
elastic formulation with the warping beam option. The element works well in some
cases, but deficiencies have been identified in the element showing that the element
gives poor results in other cases. Therefore we have removed the element in version
8.8. The element can still be accessed by choosing the 8.7 element formulations
(KINEMATICS BEAM-ALGORITHM=V87).
t
t THICKNESS
HEIGHT C.G. s
C.G. s
(H2)
WIDTH H2
(H1)
DIAMETER
(H1)
a) Rectangular section b) Pipe section
THICK 2 WIDTH 2
(H4) (H3)
t
t THICK 3
(H6)
HEIGHT THICK 2
HEIGHT s s (H5)
(H2) (H2)
THICK 1 THICK 1
(H3) (H4)
WIDTH WIDTH
(H1) (H1)
THICK 2 THICK 1
(H4) (H3)
t H5=H2-H4
y
t s
HEIGHT
HEIGHT s
(H2)
(H2) x
THICK 1 THICK 2
(H3) (H4)
WIDTH WIDTH
(H1) (H1)
e) U section f) L section
1 H1 H14 3
Rectangular 1 − 0.63 1 − H1 H2, H1 ≤ H2
3 H2 12H24
π
Pipe (H14 − H24 )
32
2 H3 H4 (H1 - H3)2 (H2 - H4)2
Box
H1 H3 + H2 H4 - H32 - H42
1
I
3
( H43H1 + H63H3 + H53 (H2 - H4 - H6) )
1
U
3
( 2H43 H1 + H33 (H2 - 2H4)) )
1
L
3
( H33 H2 + H43 (H1 - H3) )
I ss ← I ss + A × SC 2
I tt ← I tt + A × TC2
I st ← I st + A × SC × TC
2
I xx + I yy I xx − I yy
I ss = + + I xy
2
2 2
2
I xx + I yy I xx − I yy
I tt = − + I xy
2
2 2
where the moments and products of inertia with respect to the local
axes x and y are:
2 2
H 4 H13 H H H
3
H
I xx = + H1 H 4 1 − C x + 5 3 + H 5 H 3 3 − C x
12 2 12 2
2 2
H H3 H H H
3
H
I yy = 1 4 + H1 H 4 4 − C y + 3 5 + H 5 H 3 H 4 + 5 − C y
12 2 12 2
H H H H
I xy = H1 H 4 1 − Cx 4 − C y + H 3 H 5 3 − Cx H 4 + 5 − C y
2 2 2 2
and where (Cx, Cy) are the centroid coordinates. In this case, the
angle between the s principal axis and the horizontal is determined
to maximize the following expression:
• When the L section is used with the warping beam, then the
I ss = I xx , I tt = I yy , I st = I xy
2.4.2.1 Kinematics
u = un − θ t s + θ s t
v = vn − θ r t
w = wn + θ r s
dwn
θs = −
dr
dv
θt = n
dr
un = L1u 1 + L2u 2
vn = H1v 1 + H 2θt1 + H 3v 2 + H 4θt 2
wn = H1w1 − H 2θ s1 + H 3 w2 − H 4θ s2
θ r = L1θ r1 + L2θ r2
dH1 1 dH 2 1 dH 3 2 dH 4 2
θs = − w + θs − w + θs
dr dr dr dr
dH dH dH dH
θt = 1 v 1 + 2 θt1 + 3 v 2 + 4 θ t 2
dr dr dr dr
r r
L1 = 1 − , L2 =
L L
r2 r3 r2 r3
H1 = 1 − 3 2 + 2 3 , H 2 = r − 2 + 2 ,
L L L L
2 3 2 3
r r r r
H3 = 3 2 − 2 3 , H4 = − + 2
L L L L
z
Vs2
Vr2
x y
2 Vt2
Vs1
Vr1
1 Vt1
a) Before deformation
Vs1
Vr1 Vs2
1
Vt1 Vr2
2 Vt2
2
y
y x
v
y
v x
-
-v
Figure 2.4-6: Coordinate system update for large displacement beam element,
in-plane bending example
KINEMATICS KBEAM-EIGENVALUE=YES
1 1
3 0 0 0 0 0 0 0 0 0 0
6
13 6 I tt 11L I 9 6 I tt 13L I tt
+ 0 0 0 + tt 0 − 0 0 0 − +
35 5 AL2 210 10 AL 70 5 AL2 240 10 AL
13 6 I ss 11L I 9 6I 13L I ss
+ 0 − − ss 0 0 0 − ss 0 − 0
35 5 AL2 210 10 AL 70 5 AL2 240 10 AL
I rr I rr
0 0 0 0 0 0 0
3A 6A
L2
2I 13L I L 2
I
+ ss 0 0 0 − + ss 0 − − ss 0
105 15 A 420 10 AL 140 30 A
L2 2I 13L I L 2
I
+ tt 0 − tt 0 0 0 − − tt
105 15 A 420 10 AL 140 30 A
M = ρ AL
1
0 0 0 0 0
3
13 6 I tt 11L I
+ 0 0 0 − − tt
35 5 AL2 210 10 AL
13 6 I ss 11L I
+ 0 + ss 0
35 5 AL2 210 10 AL
I rr
0 0
3A
2
L 2I
+ ss 0
105 15 A
2
L 2 I tt
symmetric +
105 15 A
AE
L 0 0 0 0 0 −k11 0 0 0 0 0
12 EI tt 6 EI tt
0 0 0 0 − k22 0 0 0 k26
L3 (1 + φ2 ) L2 (1 + φ2 )
12 EI ss −6 EI ss
0 0 0 0 − k33 0 k35 0
L3 (1 + φ3 ) L2 (1 + φ3 )
GJ
0 0 0 0 0 − k44 0 0
L
EI ss (4 + φ3 ) EI ss (2 − φ3 )
K= 0 0 0 − k35 0 0
L(1 + φ3 ) L(1 + φ3 )
EI tt (4 + φ2 ) EI tt (2 − φ2 )
0 − k26 0 0 0
L(1 + φ2 ) L(1 + φ2 )
k11 0 0 0 0 0
k22 0 0 0 −k26
k33 0 − k35 0
k44 0 0
k55 0
symmetric k66
24(1 +ν )
φ2 = I tt , Assh > 0
L2 Assh
= 0, Assh = 0
24(1 + ν )
φ3 = 2 sh
I ss , Atsh > 0
L At
= 0, Atsh = 0
• Note that torsional warping effects are only taken into account
by the selection of the torsional constant J. An alternative for open
thin-walled cross-sections is to use the warping beam described in
Section 2.4.3.
r r2 r r2
u = ... + −1 + 6 − 6 2 t β1 + −1 + 6 − 6 2 s β 2
L L L L
r2 r3
v = ... + 3 2 − 2 3 β 2
L L
r2 r3
w = ... − 3 2 − 2 3 β1
L L
t
grt
s
grs
u = ... + stα1 + ( s 3t − st 3 ) α 2
Nodal forces: Same as for the elastic beam element. Note that if
end forces are requested in an elastic-plastic analysis, the state of
stress in the element (whether elastic or plastic) is not indicated.
1 2 3 ... HEIGHT s
Ä
Ä
Ä
Ä Ä Ä Ä 1 2 3
r
r=0 r=L
t WIDTH
a) Integration point locations in r-direction b) Integration point locations in s-direction
t Integration points t
Ä
equally spaced
s s
Ä
ÄÄÄ
1 3
2
1
t
Integration points
Integration points equally spaced
s
equally spaced
3 Ä
...
2Ä
R 1Ä
1 2 3 ... Ä s
Ä Ä Ä Ä
r
r=0 r=L
t
t t
Integration points Integration points
equally spaced equally spaced
... Ä 3 2 ... Ä 3 2
R Ä R Ä
1 1
Ä Ä Ä
s s
Ä
b) Integration point locations in tangential direction, d) Integration point locations in tangential direction,
2-D action 3-D action
• The flexural and the torsional behavior of the beam element are
respectively described by "bending moment vs. curvature" and
"torsional moment vs. angle of twist" relationships. These
relationships are input in ADINA in the form of multilinear
functions (see Figure 2.4-10).
Figure 2.4-10: Input curves for the moment-curvature models (curves are
shown only for positive forces, moments and curvatures/twists)
• Note that the beam element does not include any shear
deformation when the moment-curvature description is used.
• Note also that warping effects must be taken into account in the
definition of the "torsional moment vs. twist" input curve if
necessary.
Last input
data point
Curvature
(or twist angle per
unit length)
y
yyield
0
yyield
2´yyield
0
x
2´yyield
Kinematic hardening
Mixed hardening
Isotropic hardening
axial compression.
In the symmetric case (Figure 2.4-13), enter only the positive
section of the axial force/axial strain, bending moment/curvature or
torsional moment/twist angle curve. The first data point always
corresponds to yielding and the last data point always corresponds
to rupture.
In the non-symmetric case (Figure 2.4-14), enter the entire axial
force/axial strain, bending moment/curvature or torsional
moment/twist angle curve. The first data point always corresponds
to negative rupture and the last data point always corresponds to
positive rupture. One data point - the zero point - must be at the
origin. The data point prior to the zero point corresponds to the
negative yield and the data point after the zero point corresponds to
the positive yield. A different number of data points can be used
for the positive and negative sections of the curve.
Yielding:
First input
data point
x
xyield xrupture
• The same element section can be plastic with respect to the axial
deformation, but still elastic with respect to bending or torsion.
The same remark applies for rupture.
x negative xnegative x
rupture yield x positive
x positive
yield rupture
ynegative
yield
Curve cannot
pass through
this quadrant
ynegative
rupture
y
yu¢
yi ¢
yu¢
>1
yu¢ yi ¢
yi ¢
x
Plastic strain
Figure 2.4-15: Typical material curve with a cyclic rigidity different than
the initial rigidity
computation.
t Moment-t
Moment-s
The results are always given in the element local coordinate system
(r,s,t). Note that a positive value of the Moment-s at a section point
location points in the negative s-direction.
Plastic-multilinear: PLASTIC_FLAG_AXIAL,
PLASTIC_FLAG_TORSION, PLASTIC_FLAG_BENDING-S,
PLASTIC_FLAG_BENDING-T, AXIAL_FORCE,
TORSIONAL_MOMENT, BENDING_MOMENT-S,
BENDING_MOMENT-T, AXIAL_STRAIN, TWIST,
CURVATURE-S, CURVATURE-T,
PLASTIC_AXIAL_STRAIN, PLASTIC_TWIST,
PLASTIC_CURVATURE-S, PLASTIC_CURVATURE-T,
ACCUM_PLASTIC_AXIAL_STRAIN,
ACCUM_PLASTIC_TWIST,
ACCUM_PLASTIC_CURVATURE-S,
ACCUM_PLASTIC_CURVATURE-T, YIELD_AXIAL_FORCE,
YIELD_TORSIONAL_MOMENT,
YIELD_BENDING_MOMENT-S,
YIELD_BENDING_MOMENT-T
• In the case of beam elements with rigid ends, the element stress
resultants are computed at the integration point locations
corresponding to the flexible part of the element, excluding the
rigid ends.
• For the linear elastic beam element, for all cross-sections, if the
boundary conditions of the beam are such that the warping
displacements are not fixed anywhere, then warping can be
modeled just by the proper choice of the St. Venant torsional
constant J. The warping beam then need not be used.
Iω : Warping constant:
Iω = ∫ φ 2 dA
A
2.4.3.1 Kinematics
• The two nodes and the local coordinate system of the beam are
located at the centroid of the beam cross-section. The local
coordinate system of the element may or may not be coincident
with the principal planes of inertia of the cross-section.
u = un − θ t s + θ s t + α φ ( s, t )
v = vn − θ r t
w = wn − θ r s
Construction of φ ( s, t ) :
ref. A. Gjelsvik, The Theory of Thin Walled Bars, John Wiley &
Sons, 1981
Contour branch
dq Contour
p=0 on contour
t dp
dq
Cross-section
Contour, p = 0
t
p
q
ct
Pole P s
C cs s
usˆ = −tˆθ r ,
utˆ = sˆθ r
where usˆ , utˆ are the displacements in the sˆ, tˆ directions. Also
u pˆ = − qˆθ r ,
uqˆ = pˆθ r
u p = − qˆθ r ,
uq = pˆθ r
∂ur ∂uq
γ rq = +
∂q ∂r
∂ur ∂uq ∂θ
=− = − pˆ r
∂q ∂r ∂r
∂ur ∂u p
γ rp = +
∂p ∂r
The choice
∂ur ∂u p ∂θ
=− = qˆ r
∂p ∂r ∂r
∂θ r
ur ( p ,q )
= ur ( 0, q )
+ p qˆ
∂r
and therefore
ur ( p ,q ) ( q
= − ∫ pˆ dq + p qˆ
0 ) ∂∂θr
r
q
φ ( s, t ) = − ∫ pˆ dq + p qˆ
0
We consider here only closed sections with just one closed circuit
in the contour. It is assumed that the shear strain flow γ rq b is equal
∂θ r
to K on the contour, where b is the cross-section thickness
∂r
∂ur ∂uq K ∂θ r
and K is a constant to be determined. Thus + = .
∂q ∂r b ∂r
As in the open cross-section theory, uq = pˆθ r , thus
qK ∂θ
ur = ∫ − pˆ dq r . The constant K is determined by
∂r
(0, q )
0 b
noting that the change in warping displacement resulting from one
complete circuit must be zero, therefore
K=
∫ pdq
ˆ
=
2A
1 1
∫ b dq ∫ b dq
where A is the area enclosed by the contour.
Using γ rp = 0 along lines perpendicular to the contour, the
warping displacement is
q K ∂θ
ur ( p ,q )
= ∫ − pˆ dq + p qˆ r
0b ∂r
q K
φ ( s, t ) = ∫ − pˆ dq + p qˆ
0
b
(a) (b)
2 2
∂φ ∂φ ∂φ ∂φ
∫ ∂r + ∂s − tˆ ∂s + sˆ ∂t dA = 0
but the thin-walled warping function used here has
2 2
∂φ ∂φ ∂φ ∂φ
∫ ∂r + ∂s − tˆ ∂s + sˆ ∂t dA = ∫ 2 p dA
2
dθ r dθ
due to torsion is given by u = φ ( s, t ) , hence α = r . The
dr dr
dimension of α is seen to be [L-1].
θ r = H1θ r1 + H 2α 1 + H 3θ r2 + H 4α 2
dθ r
u= φ ( s, t )
dr
v = θ r tˆ
w = −θ r sˆ
d 2θ r
ε rr = φ ( s, t )
dr 2
∂u ∂v ∂φ ˆ dθ r
γ rs = + = −t
∂s ∂r ∂s dr
∂u ∂w ∂φ dθ
γ rt = + = + sˆ r
∂t ∂r ∂t dr
L
∫τ rr δε rr +τ rsδγ rs + τ rtδγ rt dV = T δθ r 0
d 2δθ r
∫ τ rrδε rr dV = ∫ τ rrφ dr 2
dV
∂φ ∂φ dδθ r
∫τ rs δγ rs + τ rtδγ rt dV = ∫ τ rs − tˆ + τ rt + sˆ
∂s
∂t dr
dV
Define
∂φ ∂φ
M b = ∫ τ rrφ dA , Fb = ∫ τ rs + τ rt dA ,
∂s ∂t
M r = ∫ −τ rs tˆ + τ rt sdA
ˆ
d 2δθ r dδθ r L
∫ Mb dr 2
+ ( Fb + M r )
dr
dr = T δθ r 0
be rewritten as
L
d dM b dM b
− ∫ Fb + M r −
L
δθ r dr = T − Fb + M r − δθ r − M bδα 0
dr dr dr 0
∂φ ˆ ∂φ ∂φ ∂φ
Fb = Gα ∫ −t + + sˆ dA
∂s ∂s ∂t ∂t
2 2
∂φ ∂φ ∂φ ∂φ
= Gα ∫ + − tˆ + sˆ dA
∂s ∂t ∂s ∂t
Fb = Gα ∫ 2 p 2 dA , M r = Gα ∫ 2 p 2 dA
T = Fb + M r = Gα ∫ 4 p 2 dA = GJ α
∫
implying J = 4 p 2 dA . For a cross-section with constant thickness
1 3
a and length b, this formula gives J = a b.
3
column
weld
weld
beam beam
multiple nodes
warping is continuous
warping = 0 warping = 0
q2 q3
q1
qi small
warping displacements
• For the warping beam, the consistent mass matrix of the beam is
implemented using the kinematics given above. Although, in most
practical cases, the effect of rotary inertias is insignificant for
bending deformations, the rotary inertias have a significant effect
on the torsional response of the beam.
• The internal force vector and stiffness matrix of the element are
calculated in closed-form, using the kinematics given above. Notice
that the torsional rigidity is derived directly from the kinematics,
without additional assumptions.
The warping constant Iω and the shear center offsets cs , ct
appear in the force vector and stiffness matrix, so these values must
be specified when using the general cross-section. In addition the
product of inertia I st can be nonzero.
The warping constant Iω , shear center offsets cs , ct and
product of inertia I st are automatically calculated when using the
box, I, U and L cross-sections. The calculations of Iω and cs , ct
are made to be consistent with the calculations done in the plastic
material (in which all integrals are evaluated numerically from the
warping function). Hence the values of Iω , cs , ct used in the elastic
material might not be the same as the handbook values of these
quantities, particularly if the cross-section is thick.
...
312 Ä Top flange NTHICK3
311 Ä Ä ...
321
NTHICK2
NWEB
...
221Ä
ÄÄ
211 212 ...
...
NFLANGE1
...
221Ä
...
NFLANGE
NTHICK1
221Ä
...
NLEG2
...
212 Ä Top flange
211 Ä
Ä
221 ...
NTHICK1
Integration points
evenly spaced
...
321Ä 121Ä
ÄÄ ÄÄ
311 312 ... 111 112 ...
...
NWIDTH
• The options for beam element output are the same as for the
elasto-plastic beam without warping:
Nodal forces: Same as for the elastic beam element. Note that if
end forces are requested in an elastic-plastic analysis, the state of
stress in the element (whether elastic or plastic) is not indicated.
Shear force
Internal hinge released
l l
r r r r
Element 1 Element 2 Element 1 Element 2
ENDRELEASE 1 12
EDATA
ENTRIES EL ENDRELEASE
11
not the mass matrix. If end releases are used in dynamic analysis,
the elements in which end releases are specified should be very
short.
• End releases can be used in all beam analyses, except for large
displacement warping beam analyses.
EDATA
ENTRIES EL RIGID1 RIGID2
1 0.2 0.4
DATAEND
In this example, the length of rigid end 1 is 0.2 units, and the length
of rigid end 2 is 0.4 units.
Physical problem: Finite element model:
Beam element
Rigid end
0.2 0.4
The mass of the rigid ends is taken into account in the mass
matrix calculations.
• When using warping beams with rigid ends, the warping degree
of freedom should be fixed at the rigid end.
Beam element
Rigid links
Hollow square section
Beam
C.G. C.G.
Rigid link
A
A
P P
Figure 2.4-31: Using rigid links to transfer loads to the centroid of the beam
• The plane stress 2-D beam element is identical to the 3-D beam
element but constrained to act only in the YZ plane. Hence, all
motion of the 2-D plane stress beam element must occur in the YZ
plane (see Fig. 2.5-2).
• The difference between the plane strain and the plane stress 2-D
beam elements is that for the plane strain element, it is assumed
that the out-of-plane strain εxx is equal to zero whereas the out-of-
plane stress σxx is equal to zero for the plane stress element.
Axisymmetric shell:
Z
1 radian
l
sxx=0 l exx=0
l
l
l Y
X l Y
X
Figure 2.5-1: 2-D iso-beam, axisymmetric shell, and 3-D iso-beam elements
s
r 2
l
1l
Z Element has constant
rectangular cross-section
1
l
3 s
Axis of l
symmetry r
Z 4
Element has varying l
thickness cross-section
l
Y Element can have 2, 3
or 4 nodes 2
s 2 2-node iso-beam
1 r
3 2
1
r-s plane
3-node iso-beam
Auxiliary node
4 2
3
r-s plane
1
Auxiliary node
4-node iso-beam
Curved beams
ref. KJB • The element matrices and vectors are formulated using the
Sections 5.4.1 isoparametric interpolation, and Gauss or Newton-Cotes numerical
and 6.5.1 integration is used to evaluate these matrices in all analyses. For the
2-D beam and axisymmetric shell elements, numerical integration
is only performed in the r-s plane, hence these elements are
considerably less expensive in terms of computer time than the
general 3-D element. The locations and labeling of the integration
points are illustrated in Fig. 2.5-3. INR is the index giving the
location of the integration point in the r direction, with INR=1
corresponding to the lowest value of r, and INR=(integration point
order) corresponding to the highest value of r. INS and INT are
analogous for the s and t directions.
31 Ä
32 Ä 31Ä
r 21 Ä
r
33 Ä Ä 32
22 Ä
23 Ä Ä 21
s Ä
22
s
12Ä 11
Ä
Ä Ä Ä
13 12 11
label = 10(INR) + INS label = 10(INR) + INS
• For the 3-D beam element, you can choose either 4×4 Gauss or
7×7 Newton-Cotes integration over the beam cross-section (along s
and t).
113
Ä Ä Ä
133
213 t Ä 132
Ä Ä
Ä Ä
Ä
111 121
Ä Ä
Ä Ä
313Ä 323 Ä Ä
Ä Ä s
rÄ
322 Ä Ä
Ä Ä 231
Ä
t 112
Ä Ä 122
312 s
Ä Ä Ä Ä
r 221
Ä Ä
321 label = 100(INR) + 10(INS) + INT
d) Gauss integration (3 2 2)
l l
l
l l
l
l Shell node
l Iso-beam node
a) Stiffened plate
l
l
l l
l l
l l l l
l
l l l
ll l
l l
l l
l l
l
b) Stiffened shell
F = KU
ref. KJB • The element matrices are evaluated using Gauss or Newton-
Section 6.6.3 Cotes numerical integration. In elastic-plastic analysis, the stress-
strain matrix is modified to include the effects of plasticity. This
stress-strain matrix is based on the classical flow theory with the
von Mises yield condition and is derived from the three-
dimensional stress-strain law with the appropriate stresses and
strains set to zero.
ref. KJB • The consistent mass matrix is calculated using the isoparametric
Section 5.4.1 formulation with the displacement interpolations given on p. 408 of
ref. KJB.
• The lumped mass for degree of freedom i is M ⋅ i , where M
L
is the total mass, L is the element length and i is a fraction of the
length associated with node i. The rotational mass is
1 1
⋅ M ⋅ i ( b 2 + d 2 ) where b, d are the width and depth of the
3 L 6
L
cross-section. For a 2-node iso-beam element 1 = 2 = , while
2
L L
for a 4-node iso-beam element 1 = 2 = and 3 = 4 = (see
6 3
Fig. 2.5-6).
variable names.
M M
2 2
l l
a) 2-node element
M M M
4 2 4 Note: M = total mass
l l l of element (2-D or 3-D
beam elements)
b) 3-node element
M M M M
6 3 3 6
l l l l
c) 4-node element
THERMAL_STRAIN(RST), ELEMENT_TEMPERATURE,
EFFECTIVE_CREEP_STRAIN
ref. KJB Nodal forces: Nodal point forces are obtained using the relation
Section 6.3
∫
t +∆t t +∆t t +∆t
F= BT τ dV
t +∆t
V
ref. J.-L. Batoz, K.J. Bathe, and L.W. Ho, "A Study of
Three-Node Triangular Plate Bending Elements", Int. J.
Num. Meth. in Eng., Vol. 15, pp. 1771-1812, 1980.
ref. K.J. Bathe and L.W. Ho, "A Simple and Effective
Element for Analysis of General Shell Structures", J.
Computers and Structures, Vol. 13, pp. 673-681, 1981.
( )
kθ z = min kθ x , kθ y × 10-4 , K θ z = kθ z I
in order to remove the zero stiffness but yet not affect the analysis
results significantly.
• In the linear analysis of flat plates that lie parallel to the global
coordinate planes, the rotational degrees of freedom normal to the
plane of the plates would best be deleted, since the plate/shell
elements do not provide actual stiffness to these degrees of
freedom. Fig. 2.6-2 illustrates these considerations.
z
y
1 qz qy
l
w v
Z l
3
u
Y qx
X l
2
x
z z z
qz
y + w y = wy
v qy u v qy
x u x qx x
qx
KM KB K
mode.
Z Beam
X Y
Note: Torque is taken by the beam only (except for the very
small value resisted by the Kqz stiffness of the plate).
P
Z
Y P
X d2
d1
P
Z
Y
X P d2
d1
P
Z
Y P
d2
X
d1
(c) Cantilever plate model, d1 ¹ d2
Z Z
Y Y
X X
(a) Model of clamped plate. Corner (b) Model of clamped plate. Corner
element shows zero response. region shows non-zero response.
Z
Y
M
• The translational lumped mass assigned to each node is and
3
M t2
the rotational lumped mass is ⋅ , in which t is the thickness of
3 3
the plate element.
The indices XL and YL indicate that the results are output with
respect to the element local coordinate system.
See Section 13.1.1 for the definitions of those variables that are
not self-explanatory.
ref. KJB • The basic equations used in the formulation of the element are
Sections 5.4.2 given in ref. KJB.
and 6.5.2
• The shell element is formulated treating the shell as a three-
dimensional continuum with the following two assumptions used in
the Timoshenko beam theory and the Reissner/Mindlin plate
theory:
For the Timoshenko beam theory, the structure is the beam, and
for the Reissner/Mindlin plate theory, the structure is the plate
under consideration. In shell analysis, these assumptions
correspond to a very general shell theory.
a) Midsurface description
c) Top-bottom description
b) Triangle
e) Multilayered shell
ref. KJB The shell thickness, ak , at the nodal points measured in the
Fig. 5.33
page 437
direction of the director vectors t Vnk (see Fig. 2.7-4 and
Section 2.7.2).
t
For the 9-node shell, the
2 midsurface center node
is local node number 13.
6 9 s
10 14 5
15 1
3 13
12
16
7 r
8
11
26 15 30 21 1
13
3 31 17
29 12
16
19 7
32 28
8
23 11
r
24
27 4
b) Top-bottom nodes 20
q
t q
x = ∑ hk xk + ∑
t t t
ak hk Vnxk
k =1 2 k =1
q
t q
y = ∑ hy yk + ∑
t t t
ak hk Vnyk
k =1 2 k =1
q
t q
z = ∑ hk zk + ∑ ak hk Vnzk
t t t
k =1 2 k =1
where q is the number of element nodes and t Vnxk , tVnyk , tVnzk are
the direction cosines of the shell director vector t Vnk .
l k
element 1 l
element 2
The configuration shown
corresponds to time t=0.
0 k 0 k
Vn Vn
elt.1 elt.2
l
k kl
l
element 1 l
element 2
the rˆ, sˆ, tˆ coordinate system, shown in Fig. 2.7-6(b), with the
condition that the stress in the direction tˆ is zero.
• Note that if the nodal point midsurface director vectors are input
or generated (see Section 2.7.3) to be initially "exactly" normal to
the midsurface of the shell element (the coordinates of this
midsurface are obtained by interpolation from the nodal points on
the midsurface), then assumption 1 is also exactly satisfied in the
finite element solution, but assumption 2 is only approximately
fulfilled in geometric nonlinear analysis, because the directions of
the initially normal vectors are updated.
Program calculates director
vector by taking average of
all normal vectors at node k.
l k
element 1 l
element 2
0 k 0 k
Vn Vn
l
k kl
l
element 1 l
element 2
l k
element 1 l
element 2
0 k 0 k
Vn Vn
l
k kl
l
element 1 l
element 2
l l
k k+1
L1 L2
Angle = 90°
t
l
Vn
l
l
Z Final configuration
Angle ¹ 90°
Y
X
b) Change in direction of director vector due to
displacements and deformations (with shear)
r
r
r= s ´ t s= t ´ r
(t = 0)
s´t 2 t´r 2 r
t t
s
t
s
s
r
r
t= r´s , r=
s ´ t , s=t´r
(t = 0)
r´s s´t r
Y × 0 Vnk
0
V1k =
Y × 0 Vnk
2
0
V2k = 0 Vnk × 0 V1k
For the special case when the 0 Vnk vector is parallel to the Y
axis, the program uses the following conventions:
0
V1k ≡ Z 0
V2k ≡ X when 0
Vnk ≡ + Y
and
0
V1k ≡ −Z 0
V2k ≡ X when 0
Vnk ≡ − Y
No rotational
stiffness for k
Vn
this DOF
wk
l Z
l
l
vk bk
l
Y
X
uk ak
4 3 2 1l
l l l
5 MZ FZ
6
l l
MY FX
9l 8 l
7 l l
Moment causes
z infinite rotation.
y Mz
x t
Vn
Reactions at fixities
are zero.
Zero pivots in
stiffness matrix.
F
d) Structural element/rigid link attached to node,
structural element/rigid link is supported.
l l
l l
l l
l l l
l
l
l
l
l l
l
l l
l
Transition nodes
l
Z wk vk
uk
Y
X
a) Example of a transition element
t DL uniform
over element
DL s
L
r
l l
l
l
l
l l
l l l
l
l l l
l l
l l l l l
l
ref. K.J. Bathe and L.W. Ho, "Some Results in the Analysis
of Thin Shell Structures," in Nonlinear Finite Element
Analysis in Structural Mechanics, W. Wunderlich et al
(eds.), Springer-Verlag, 1981.
ref. KJB • Numerical integration is used for the evaluation of the element
Section 5.5 matrices, and the default integration or a higher order should
always be used; then no spurious zero energy modes are present.
3 3 2 Gauss integration: t s
INR=1,INS=1,INT=2 Ä Ä Ä
Ä ÄÄ ÄÄ Ä
INR=1,INS=1, r
INT=1 Ä ÄÄ ÄÄ Ä
Ä Ä Ä
INR=3,INS=1,INT=1
2 2 3 Newton-Cotes integration:
t s
INR=1,INS=2,INT=3 Ä Ä
Ä Ä
Ä Ä
Ä Ä r
Ä Ä
Ä Ä
INR=2,INS=1,INT=1
s s s
3 2
1
1 1
3 4
2 r r
r
1-point 3-point 4-point
s s
3 3 4
2 6 7 8
5 2 10 12
6
1 7 1 13
11
4 r 5 9 r
7-point 13-point
b. Integration points position on r-s plane for the triangular shell elements
Figure 2.7-13: (continued)
Specify the data of each layer (or ply) in terms of the weight
per unit surface of the fiber Wf, the density of the fiber ρ f and
the fiber volume fraction of the fiber-matrix compound φ f .
The AUI then computes the thickness of each layer
Wf
automatically using the formula h = , in which it is
ρ f φf
assumed that the ply is made of uniaxial fibers, so that the fiber
thickness fraction perpendicular to the plate midsurface is
proportional to the fiber volume fraction. The total thickness of
the multilayered shell is then the sum of the ply thicknesses.
This approach is especially useful when the layers are fiber-
matrix composites.
1 n i
t = −1 + 2 ∑
a i =1
− n
(1 − t )
n
(2.7-1)
with
i
a and are functions of r and s.
e3
l
3
Y t l
e2
e1 l
X l 2
s
r l
a4 4 l l 1
Vn
l
l 1
e2 ´ lV kn 1l V2
l k
V1 =
|| e2 ´ lVnk || 2 l 1 b1
V1 a1
l k
V 2 = lVnk ´ lV 1k
z,w
y,v
x,u
N N
n
xi = ∑ hk xik + ∑ mkn + t n k hk Vnik (2.7-2)
k =1 k =1 2
with
n n
ak
mkn = − +∑ i
k − k
(2.7-3)
2 i =1 2
N N
n t n nk 0 k
ui = ∑ hk u + ∑ mk + ( − V2i α k + V1i β k )
k 0 k
i
k =1 k =1 2
(2.7-4)
with
I xx2
k xz = (2.7-5)
a a
g xz2
∫ ∫
2 2
Gxz dz dz
−a 2 −a 2 Gxz
and the transverse shear stress in terms of the shear strain value
γ xz is
I xx g xz ( z )
σ xz ( z ) = γ xz (2.7-6)
a
g
∫− a 2 Gxzxz dz
2
where
a
I xx = ∫ a Dxx ( z − z0 ) dz
2 2
− 2
Exxn
Dxxn =
1 −ν xyn ν yxn
Exx = Young's modulus
Gxz = shear modulus
The rotational lumped mass for all except explicit dynamic analysis
M 1
is, ⋅ ( tav2 ) where tav is the average shell thickness. The same
n 12
rotational mass matrix is assumed for 5- and 6-degree of freedom
nodes, and is applied to all rotational degrees of freedom.
M 1 2
⋅ ( t + A ) , where tav is the average shell thickness and A is
n 12 av
the cross-sectional area. The rotational masses are scaled up to
ensure that the rotational degrees of freedom do not reduce the
critical time step for shell elements. The same rotational mass
matrix is assumed for 5- and 6-degree of freedom nodes and is
applied to all rotational degrees of freedom.
Xc < τa < Xt
Yc < τb < Yt
Zc < τc < Zt
− Sab < τ ab < Sab
− Sac < τ ac < Sac
− Sbc < τ bc < Sbc
The failure of the material occurs when any one of the following
inequalities is not satisfied anymore:
where (εa, εb, εc, γab, γac, γbc) is the strain vector referred to the
principal material directions.
The above conditions are applicable to 3-D analysis. In plane
stress analysis, however, the parameters Zεt, Zεc, Sεac and Sεbc are not
used.
(G + H )τ a2 + ( F + H )τ b2 + ( F + G )τ c2 − 2 Hτ aτ b − 2Gτ aτ c − 2 Fτ bτ c
+ 2 Lτ bc2 + 2 M τ ac2 + 2 Nτ ab2 < 1
with
1 1 1 1
F = − 2 + 2 + 2
2 X Y Z
1 1 1 1
G= 2
− 2+ 2
2 X Y Z
1 1 1 1
H= 2
+ 2− 2
2 X Y Z
1
L=
2 Sbc2
1
M=
2 Sac2
1
N=
2 S ab2
where
In the case when plane stress conditions are used, the material
behavior with respect to failure is considered as transversely
isotropic (Y=Z), and the failure criterion reduces to
τ a2 τ aτ b τ b2 τ ab2
− + + <1
X2 X2 Y2 Sab2
In this case, the maximum allowable stresses Z, Sbc and Sac are not
used.
τ1 = τ a ; τ 2 = τ b ; τ 3 = τ c
τ 4 = τ ab ; τ 5 = τ ac ; τ 6 = τ bc
and with
1 1 1
F1 = + ; F11 = −
X at X ac X at X ac
1 1 1
F2 = + ; F22 = −
X bt X bc X bt X bc
1 1 1
F3 = + ; F33 = −
X ct X cc X ct X cc
F4 , F5 , F6 = 0
1 1 1
F44 = 2 ; F55 = 2 ; F66 = 2
Sab Sac Sbc
For plane stress conditions, the parameters Xct, Xcc, Sbc, Sac, Fbc
and Fac are not used.
1 1
Fij = − +
X it X ic X jt X jc
with i ≠ j and i, j = a, b, c
τ a2
+
(τ 2
ab + τ ac2 )
<1 (if τ a > 0) (2.7-7)
X t2 Sab2
τa
<1 (if τ a < 0) (2.7-8)
Xc
(τ b + τ c )
2
(τ 2
− τ bτ c ) (τ 2
+ τ ac2 )
Yt 2
+
bc
Str2
+
ab
Sab2
<1 ( if (τ b + τ c ) > 0)
(2.7-9)
Yc
2
(τ b + τ c ) 1 −
2 Str (τ b + τ c ) (τ b − τ bτ c ) (τ ab + τ ac )
2 2 2 2
+ + + <1
Yc 4 Str2 Str2 S ab2
( if (τ b + τ c ) < 0)
(2.7-10)
with
6
Fiτ i + Fijτ iτ j = 1 if ∑α τ
i =1
i i >0
with
τ1 = τ a ; τ 2 = τ b ; τ 3 = τ c
τ 4 = τ bc ; τ 5 = τ ac ; τ 6 = τ ab
The material failure occurs when any one of the failure surfaces is
reached and the corresponding stress condition is satisfied at the
same time.
You can request that ADINA either print and save stresses or
forces.
( )
midsurface rˆ, sˆ, tˆ axes (see Fig. 2.7-6b). For the large
displacement/large strain formulations, both the local and
( )
midsurface axes are the rˆ, sˆ, tˆ axes, so the results in both the
local and midsurface systems are identical. In addition, for
orthotropic material models, the results referred to the material axes
can be requested.
PLASTIC_STRAIN(XYZ), CREEP_STRAIN(XYZ),
THERMAL_STRAIN(XYZ), ELEMENT_TEMPERATURE,
ACCUM_EFF_PLASTIC_STRAIN,
FE_EFFECTIVE_STRESS, YIELD_STRESS,
EFFECTIVE_CREEP_STRAIN
( )
when results are referred to the ( r , s , t ) or rˆ, sˆ, tˆ axes,
STRESS(XYZ) = STRESS-RR, STRESS-SS,
STRESS-TT, STRESS-RS, STRESS-RT,
STRESS-ST
Hashin: FAILURE_FLAG_TENS-FIBER,
FAILURE_FLAG_COMP-FIBER,
FAILURE_FLAG_TENS-MATRIX,
FAILURE_FLAG_COMP-MATRIX,
FAILURE_CRITERION_TENS-FIBER,
FAILURE_CRITERION_COMP-FIBER,
FAILURE_CRITERION_TENS-MATRIX,
FAILURE_CRITERION_COMP-MATRIX
User-supplied: FAILURE_FLAG_SURFACE-1,
FAILURE_FLAG_SURFACE-2,
FAILURE_FLAG_SURFACE-3,
FAILURE_FLAG_SURFACE-4, FAILURE_FLAG,
FAILURE_CRITERION_SURFACE-1,
FAILURE_CRITERION_SURFACE-2,
FAILURE_CRITERION_SURFACE-3,
FAILURE_CRITERION_SURFACE-4
See Section 13.1.1 for the definitions of those variables that are
not self-explanatory.
´
Ä ´
B ´
Ä D
´ Point labels INR INS
A 1 1
B 1 2
r C 2 1
´
Ä
A
´ ´
Ä
C D 2 2
´
h/2
a) Point labeling for section results
(2 2 2 Gauss integration) E r zdz
-h/2
Z 0r =
Layer number h/2
+ _h
5 2 E r dz
4 -h/2
3 Z 0r
2 h/2
1
- h_ Es zdz
2
-h/2
Z 0s =
h/2
b) Definition of the neutral axes Es dz
-h/2
Figure 2.7-16: Conventions and definitions for section results
output
Note that the section results are given with respect to the local Cartesian system of
the shell elements.
t
s
Rss
Rrs
Rrs
r Rrr
t
s
Rst Mrr
Mss Rrt Mrs
r Mrs
ref. KJB Nodal forces: The nodal point force vector which corresponds to
Example 5.11 the element internal stresses can also be requested in ADINA. The
pp. 358-359 procedure for the calculation of the force vector is given in ref.
KJB.
The nodal forces are accessible in the AUI using the variable
names NODAL_FORCE-X, NODAL_FORCE-Y,
NODAL_FORCE-Z, NODAL_MOMENT-X, NODAL_MOMENT-
Y, NODAL_MOMENT-Z.
s s
3 3
6 5
r r
1 2 1 4 2
MITC3 MITC6
s s s
2 2 2
6 9
6 5
10 5
r 1,4 r 1,4 r
1,4
3 3 7 11
3 7
MITC4 collapsed DISP7 collapsed DISP10 collapsed
r
l
l
t t
l s
l
l
l
s
l
l l
l l
l
l l l
l l
l
l l l
l
l
l
l l
r
l r
MITC16 MITC9
(b) 16-node cubic shell element for thick and thin shells.
9-node shell element for thick and thin shells.
ref. KJB • The phenomenon of an element being much too stiff is in the
pp. 403-408 literature referred to as element locking. In essence, the
phenomenon arises because the interpolation functions used for an
element are not able to represent zero (or very small) shearing or
membrane strains. If the element cannot represent zero shearing
strains, but the physical situation corresponds to zero (or very
small) shearing strains, then the element becomes very stiff as its
thickness over length ratio decreases.
The MITC3, MITC4, MITC6, MITC8, MITC9 and MITC16
elements are implemented to overcome the locking problem in the
shell element. When you choose to generate the 4-node shell
element (through the AUI), the MITC4 elements are created for the
quadrilateral elements and MITC3 elements are used for the
triangular elements. When you choose to generate the 8-node shell
element, the MITC8 elements are created for the quadrilateral
elements and the MITC6 elements are created for the triangular
elements. When you choose to generate the 9-node shell element,
the MITC9 elements are created for the quadrilateral elements and
the MITC6 elements are used for the triangular elements. When
you choose to generate the 16-node shell element, the MITC16
elements are created for the quadrilateral elements and DISP10
collapsed triangular elements are used for the triangular elements.
The MITC8 element is not as effective as the other MITC elements,
and its use is not recommended in general.
3
l l
l
l
2 l
2
l
l
l l
l l
l
4
l
3 l
l
l
l l
l
l
l 1 and 4
1
a) Quadrilateral elements b) Triangular elements
Overview
Midsurface
a) Undeformed configuration
Material particles
initially on midsurface
b) Deformed configuration
a) Undeformed configuration
b) Deformed configuration
a) Undeformed configuration
b) Deformed configuration
• The following features available for other shell elements are not
available for the 3D-shell element: 6-, 8-, 9-, 16-node elements,
composite (multilayer) shells, transition elements, anisotropic
failure criteria. When section results are requested, the membrane
strains and curvatures are not available.
Fig. 2.7-25 shows the corner of a 3D-shell element, with its top and
bottom surfaces described by control vectors. Initially the control
vectors are equal and opposite. During deformations, the control
vectors can evolve independently. Thus in the deformed
configuration, the control vectors are in general not equal and
opposite, as shown in Fig 2.7-25(b).
Top surface
Bottom surface
c) Deformed configuration,
warping rotations not used
x, y, z translations
α , β rotations (5 DOF node), or θ x , θ y , θ z rotations (6 DOF
node)
Material models for the usual shell elements are developed using
the assumption of zero stress through the shell thickness. Hence
these models do not directly apply to the 3D-shell element.
When stresses are output in the shell local coordinate system, the
coordinate system is the rˆ, sˆ, tˆ system shown in Figure 2.7-6(b).
The tables at the end of this section show under what conditions the
incompatible modes and u/p formulations are used. Notice that
these conditions are material dependent.
Materials that allow the u/p formulation, but for which the u/p formulation is not the
default: Elastic material
KINEMATICS EGROUP EGROUP Result
INCOMPATIBLE- INCOMPATIBLE- NELPRS /
MODES MODES NELPT
AUTOMATIC DEFAULT AUTOMATIC Incompatible modes
used if implicit, no u/p
or incompatible modes if
explicit
NO DEFAULT AUTOMATIC No u/p or incompatible
modes
YES DEFAULT AUTOMATIC Incompatible modes
used
NO AUTOMATIC No u/p or incompatible
modes
YES AUTOMATIC Incompatible modes
used
AUTOMATIC DEFAULT 0/0 Incompatible modes
used if implicit, no u/p
or incompatible modes if
explicit
NO DEFAULT 0/0 No u/p or incompatible
modes
YES DEFAULT 0/0 Incompatible modes
used
NO 0/0 No u/p or incompatible
modes
YES 0/0 Incompatible modes
used
AUTOMATIC DEFAULT A/B Error
NO DEFAULT A/B u/p used with A*B
pressure dofs
YES DEFAULT A/B Error
NO A/B u/p used with A*B
pressure dofs
YES A/B Error
Materials that allow the u/p formulation, and for which the u/p formulation is the
default: Plastic-cyclic, plastic-orthotropic, Ogden, Mooney-Rivlin, Arruda-Boyce,
Sussman-Bathe materials
KINEMATICS EGROUP EGROUP Result
INCOMPATIBLE- INCOMPATIBLE- NELPRS /
MODES MODES NELPT
AUTOMATIC DEFAULT AUTOMATIC Automatic u/p, print
warning if explicit
NO DEFAULT AUTOMATIC Automatic u/p
YES DEFAULT AUTOMATIC Automatic u/p, print
warning
NO AUTOMATIC Automatic u/p
YES AUTOMATIC Incompatible modes
used
AUTOMATIC DEFAULT 0/0 Incompatible modes
used if implicit, no u/p
or incompatible modes if
explicit
NO DEFAULT 0/0 No u/p or incompatible
modes
YES DEFAULT 0/0 Incompatible modes
used
NO 0/0 No u/p or incompatible
modes
YES 0/0 Incompatible modes
used
AUTOMATIC DEFAULT A/B Error
NO DEFAULT A/B u/p used with A*B
pressure dofs
YES DEFAULT A/B Error
NO A/B u/p used with A*B
pressure dofs
YES A/B Error
tab
Convention for
pipe stresses Pipe global displacement
taa
degrees of freedom
tcc
Integration point
tac
b
3 4 c a
s
t r
1 2
f
q R
Z
Nodes 1 to 4
and KK must
lie in one plane
Y (r-s plane).
X
Auxiliary node KK
s Mt
Mt
(in-plane moment)
Ms
Ms s
(out-of-plane moment)
Nc Nd
wη = ∑ pm cos 2mφ + ∑ qm sin 2mφ
m =1 m =1
ref. K.J. Bathe and C.A. Almeida, "A Simple and Effective
Pipe Elbow Element ─ Linear Analysis," J. Appl. Mech.,
Transactions of the ASME, Vol. 47, No. 1, pp. 93-100,
1980.
ref. K.J. Bathe, C.A. Almeida and L.W. Ho, "A Simple and
Effective Pipe Elbow Element ─ Some Nonlinear
ref. K.J. Bathe and C.A. Almeida, "A Simple and Effective
Pipe Elbow Element ─ Interaction Effects," J. Appl.
Mech., Transactions of The ASME, Vol. 49, pp. 165-171,
1982.
ref. K.J. Bathe and C.A. Almeida, "A Simple and Effective
Pipe Elbow Element ─ Pressure Stiffening Effects," J.
Appl. Mech., Transaction of the ASME, Vol. 49, pp. 914-
916, 1982.
Rigid flange
Pipe centroidal axis
Pipe elements
R1
Junction
R2
∂ wξ
wζ =
∂φ
and
wξ sin φ − wζ cos φ 1 ∂ 2 wζ
e =
ov
− ζ
aa
R − a cos φ ( R − a cos φ )2 ∂θ 2
γ abov = 0 (2.8-1)
1 ∂ wζ
γ acov =
R − a cos φ ∂θ
1 ∂ 2 wξ
eccov = − w
ζ + ζ
a 2 ∂φ 2
cos φ 1 ∂wη
w
eaa = 1 − ζ
R − a cos φ R − a cos φ ∂θ
γ abw = 0
1 ∂wη w sin φ
γ acw = − η (2.8-2)
a ∂φ R − a cos φ
eccw = 0
Here, a is the pipe mean radius, R is the bend radius and θ and
φ are the angles defined in Fig. 2.8-1.
Rδ
λ=
a 2
1 −ν 2
¶Wb
=0
¶a
Wb a Ovalization and warping
equal to zero at node j
for element i.
Node j
Element i
¶Wb
=0
¶a
Node j
ai+1 ai Ovalization and warping
equal to zero at node j
for elements i and i + 1.
Element i + 1 Element i
Wb
(a) Flange conditions
¶ Wb
Symmetry plane, =0
a ¶a
Node j Nonzero ovalization and
zero warping at node j
for element i.
Element i Wb
πai2
dFT = − p s dL
R
and the nodal transversal forces are obtained as
Fk = ∫ hk d FT
L
l
4
3l
(p) 2
dF T dL l
s Fa(p)
2
1l
R l Auxiliary node
(p)
Fa
1
RT
Ra MA Ra = RT = MA = 0
l
A
l
Internal
pressure p
RA RB
2a+d l l l l
A B
d/2 p
In elastic deformations:
RA =RB = paip(ai -2na) a = mean radius
ai = a - d/2 n = Poisson's ratio
= internal radius
where t +∆t
σ cc( p ) is the hoop stress equivalent to the pressure
t +∆t
p (using thin-walled cylinder theory). Then, using that the
normal stress through the pipe skin is equal to zero, the axial
stress t +∆tσ aa is obtained as
t +∆t
σ aa =t +∆t E ( t +∆t eaa − t +∆t eaaIN − t +∆t eTH ) + ν
t +∆t t +∆t
σ cc( p )
1 t +∆t IN
t +∆t
ecc = −t +∆tν ( t +∆t eaa − t +∆t eaa
IN
)− eaa + t +∆t OV
e cc
2
1 −ν 2 ( p )
+ (1 + ν ) t +∆t eTH + σ cc
E
• The element stiffness and mass matrices and force vectors are in
all formulations evaluated using Gauss or Newton-Cotes numerical
integration in the axial and the thickness directions and the
composite trapezoidal rule along the circumference of the pipe
section. The locations and labeling of the integration points are
given in Fig. 2.8-7.
Newton-Cotes formulas:
1 2 3
Order -3
Node 1 Node 2
1 2 3 4 5
Order -5
Node 1 Node 2
1 2 3 4 5 6 7
Order -7
Node 1 Node 2
Gauss formulas:
1
Order 1
Node 1 Node 2
1 2
Order 2
Node 1 Node 2
1 2 3
Order 3
Node 1 Node 2
1 2 3 4
Order 4
Node 1 Node 2
s s
b
ÄÄ Ä b
Ä 23 ÄÄ 5
1 ÄÄ 3 4
12
t t
t t
3
Ä2 4 Ä 2
Ä Ä
3Ä 1 5Ä 1
Ä Ä
s s
Ä Ä
Ä
6 Ä 8
4 7
Order 4 Order 8
t
t
5 4 3 9 8 76
Ä
Ä Ä 10 Ä Ä Ä Ä Ä5 4
6 11 Ä Ä 3
Ä Ä2 Ä 2
1 12 ÄÄ Ä1
7Ä Ä 13 Ä
s Ä
14 Ä Ä 24 s
Ä Ä 12 Ä 23
8 15 Ä Ä
Ä
9 Ä
Ä
11 16 Ä Ä Ä Ä Ä Ä 22
10 17 18 19 20 21
Order 12 Order 24
c) Integration point locations along circumferential direction (label INC)
• The consistent mass matrix associates mass only with the pipe’s
ref. KJB
Section translational and rotational degrees of freedom and is calculated
5.4.1 using the isoparametric formulation with the displacement
interpolations given on p. 408 of ref. KJB. Hence, no masses are
You can request that ADINA either print or save stresses or forces.
EQUIV_INTERNAL_AXIAL_PRESSURE,
EQUIV_INTERNAL_HOOP_PRESSURE,
ELEMENT_TEMPERATURE
with similar definitions for the other abbreviations used above. See
Fig. 2.8-1 for the pipe stress coordinate system convention.
See Section 13.1.1 for the definitions of those variables that are
not self-explanatory.
ref. K.J. Bathe and C.A. Almeida, "A Simple and Effective
Pipe Elbow Element ─ Linear Analysis," J. Appl. Mech.,
Transactions of the ASME, Vol. 47, No. 1, pp. 93-100,
1980.
• General elements are linear elements which can have from one
node to as many nodes as there are in the structure.
Fx1 k1,1 k1,2 k1,3 k1,4 k1,5 k1,6 k1,7 k1,8 k1,9 k1,10 k1,11 k1,12 u1x
1 k2,2 k2,3 k2,4 k2,5 k2,6 k2,7 k2,8 k2,9 k2,10 k2,11 k2,12 u1y
Fy
Fz1 k3,3 k3,4 k3,5 k3,6 k3,7 k3,8 k3,9 k3,10 k3,11 k3,12 u1z
2
Fx k4,4 k4,5 k4,6 k4,7 k4,8 k4,9 k4,10 k4,11 k4,12 u x2
F 2 k5,5 k5,6 k5,7 k5,8 k5,9 k5,10 k5,11 k5,12 u y2
y2
Fz k6,6 k6,7 k6,8 k6,9 k6,10 k6,11 k6,12 u z2
3 = k7,7 k7,8 k7,9 k7,10 k7,11 k7,12 u x3
Fx
Fy3 k8,8 k8,9 k8,10 k8,11 k8,12 u 3y
3
Fz k9,9 k9,10 k9,11 k9,12 u z3
Fx4 sym k10,10 k10,11 k10,12 u x4
4
Fy k11,11 k11,12 u y4
F 4 k12,12 u z4
z
MATRIX STIFFNESS 1 12
1 k1,1 k1,2 k1,3 k1,4 k1,5 k1,6 k1,7 k1,8 k1,9 k1,10 k1,11 k1,12
2 k2,2 k2,3 k2,4 k2,5 k2,6 k2,7 k2,8 k2,9 k2,10 k2,11 k2,12
3 k3,3 k3,4 k3,5 k3,6 k3,7 k3,8 k3,9 k3,10 k3,11 k3,12
4 k4,4 k4,5 k4,6 k4,7 k4,8 k4,9 k4,10 k4,11 k4,12
5 k5,5 k5,6 k5,7 k5,8 k5,9 k5,10 k5,11 k5,12
6 k6,6 k6,7 k6,8 k6,9 k6,10 k6,11 k6,12
7 k7,7 k7,8 k7,9 k7,10 k7,11 k7,12
8 k8,8 k8,9 k8,10 k8,11 k8,12
9 k9,9 k9,10 k9,11 k9,12
10 k10,10 k10,11 k10,12
11 k11,11 k11,12
12 k12,12
If skew systems are used, then either all matrices must be input in
the global coordinate system and proper transformations are made
by ADINA if there are skew systems at the element group nodes, or
F=K U
F=CU
F=MU
σ=SU
u q ü
k
l
l
k m
.
u
c c
l l
Single Single
translational rotational
DOF damper DOF damper
k -k m/2 0
K = Mlumped =
-k k 0 m/2
k = spring constant
m/3 m/6
a) Stiffness matrix Mconsistent =
m/6 m/3
m = total mass
c -c of element
C =
-c c b) Mass matrix
c = damping constant
c) Damping matrix
U2 U2
1 1
U1 U1
FE ME
Nonlinear translational damping: Nonlinear torsional damping:
MD
FD
U = U2 - U 1 2
2
U2 U2
1 1
FD U1 U1
MD
Time
Figure 2.9-4: Nonlinear stiffness, damping and mass
If the nodes are not initially coincident, then the element can be
used to define a translational spring along the direction given by
the vector from node 1 to node 2. Let U1 store only the
translational displacements, then the relative translational
displacement is given by U = (U 2
− U )⋅
1 ( 0
x 2 − 0 x1 )
, i.e., the
( 0
x 2 − 0 x1 )
relative displacement is the difference in displacements in the
direction of the line (spring) element.
(resp. smaller) than the last (resp. first) point on the curve. Note
that this effect is reversible, as the force will be computed again
according to the curve if the displacement value comes back into
the curve displacement range.
• Note that the linear spring element and the nonlinear spring
element act quite differently: the linear spring element connects
specific degrees of freedom, whereas the nonlinear spring element
connects nodes and acts in a specified direction. Likewise the input
for these elements is also different.
Element output
Nodal forces: For linear elements, these are the elastic nodal
forces (in the global Cartesian reference system) equivalent to the
axial elastic stress. Any stresses due to damping are not included.
For nonlinear elements, these are the nodal forces (in the global
Cartesian reference system) equivalent to the axial elastic and
damping stresses.
The forces are accessible in the AUI using the variable names
NODAL_FORCE-X, NODAL_FORCE-Y, NODAL_FORCE-Z,
NODAL_MOMENT-X, NODAL_MOMENT-Y, NODAL_MOMENT-
Z.
Y CID
Z
X
y-element
z-element
x-element
GA
Y CID
Z
X
y-element
z-element x-element
GA GB
Y CID
Z
X
y-element
z-element x-element
GA
GB
z-element
(normal to plane)
y-element
(in plane) G0 or (Ax, Ay, Az)
x-element
GA
GB
Element output
You can request that ADINA either print or save stress resultants or
forces.
The above results are always given in the element local coordinate
system (r,s,t).
Nodal forces: the forces (in the global Cartesian coordinate system)
are accessible in the AUI using the given variable names.
• There are two ways within the AUI to activate the user-supplied
element feature. Note that regardless of the way chosen, the
material definition of the element can only be done with the
MATERIAL USER-SUPPLIED command (see AUI Command
Reference Manual). The material properties can be temperature
dependent or independent. However, this does not imply that a
user-supplied material subroutine is needed. In fact, any material
implementation needed for the user-supplied element must be done
within the user-supplied element subroutine CUSERG.
k ( ∆L ) = k0 + k1∆L (2.9-2)
• These elements are defined within 2-D and 3-D fluid element
groups (in ADINA). Set the formulation of the element group to
either displacement-based without rotation penalty or
displacement-based with rotation penalty (if you are using the AUI
user interfaces, set the formulation with the “interpolation type”
field).
p = − K ev
• You can also request nodal force output. The nodal forces are
accessible in the AUI using the variable names NODAL_FORCE-X,
NODAL_FORCE-Y, NODAL_FORCE-Z (NODAL_FORCE-X is
Bounding surface S
Unknowns on S:
f = fluid potential Fluid region V
u = displacements n
Unknowns in V:
f = fluid potential
ρ p
= 1+ (2.11-3)
ρ0 κ
ρ0 h ρ0 h
ρ = ρ0 exp , p = κ exp − 1 (2.11-4a,b)
κ κ
h = Ω ( x + u ) − φ − 12 v n i v n − 12 v t i v t (2.11-8)
on S .
∂ρ
δ Fφ = ∫ h δφ − ρ∇φ i∇δφ dV + ∫ − ρ u inδφ dS (2.11-9)
V
∂h S
Fluid Structure
n
Pressure p Traction -pn
( )
p = p (h) = p Ω ( x + u ) − φ − 12 v n i v n − 12 v t i v t (2.11-11)
0 0 ∆u CUU CUF ∆u
0 −M ∆φ + C
− ( CFF + ( CFF )S ) ∆φ
+
FF FU
K K FU
UU UF ∆u 0
K FU
(
− K
FF
( ))
+ K
FF S
∆φ = 0 −
F
F ( )
+ F
F S
(2.11-12)
where the increment in the vector of unknown potentials φ is
written ∆φ and the increment in the vector of unknown
displacements u is written ∆u . (We drop the left superscript
t + ∆t here and below for ease of writing.) In the linearization
process, increments in both the nodal displacements and increments
in the nodal potentials (and their time derivatives) are considered.
In (2-11.12),
ρ0 p
ρ + ∇i( ρ∇φ ) ≈ ρ + ρ 0∇ 2φ ≈ + ρ 0∇ 2φ = 0 (2.11-13)
κ
The momentum/equilibrium equation (2.11-2) becomes
p
h≈ ≈ Ω ( x) − φ (2.11-14)
ρ
from which we see
( ) (
p ≈ ρ Ω ( x ) − φ ≈ ρ0 Ω ( x ) − φ ) (2.11-15)
− ρ 0 φ + κ ∇ 2φ = − ρ0 Ω (2.11-16)
− ∫ ρ 0 φ δφ dV − ∫ κ ∇φ ⋅δ ∇φ dV − ∫ κ u ⋅ n δφ dS
V V S
= − ∫ ρ0 Ω δφ dV (2.11-17)
V
∂Ω
δ Fu = ∫ pniδ u dS1 ≈ ∫ ρ0 Ω + ρ0 iu − ρ 0φ niδ u dS1 (2.11-18)
S1 S1
∂x
0 0 U 0 CTFU U ( K UU )S 0 U
0 −M + +
−K FF φ
FF φ CFU 0 φ 0
(2.11-19)
( R ) 0
= UB S +
0 − R FB
where
∆p = ρ c∆v (2.11-20)
p = p∞ + ρ c ( v − v ∞ ) (2-11.21)
At infinity
Direction of p=p¥
wave propagation v=v¥
a) Physical problem
so that the variation in mass flow rate is, in this special case,
ρφ
δ Fφ B = − ∫ δφ dS (2.11-25)
SB
c
p p
ρ v= + (2.11-26)
c iω r
p ∫ p dt
ρ v= + (2.11-27)
c r
− ρφ ρφ
ρ v= − (2.11-28)
c r
ρφ ρφ
δ Fφ B = − ∫ + δφ dS (2.11-29)
SB
c r
ref. L.G. Olson and K.J. Bathe, “An infinite element for
analysis of transient fluid-structure interactions”,
Engineering Computations, Vol 2, pp 319-329, 1985.
U = Ur + Ug (2.11-31)
U g = ∑ u gk d k (2.11-32)
k
( K ) d 0
− UU S k u gk − u gk
0 −CFU d k
h = Ω −φ (2.11-32)
δ Fφ = ∫ ρ δφ dV − ∫ ρ u inδφ dS (2.11-33)
V S
t +∆t
ρ = ( t +∆t ρ − t ρ ) / ∆t , t +∆t
u = ( t +∆t u − t u ) / ∆t (2.11-34a,b)
where we write δφ = δφ / ∆t .
(
K UU
)
S
K UF FU
∆u = 0 −
(2.11-37)
∆φ 0 FF + FF
K
FU ( (
− K FF + K FF ))
S ( )
S
where we use the ~ to emphasize that the vectors and matrices are
different than the corresponding dynamic vectors and matrices,
along with the constraints φ = constant (if there is more than one
fluid region in the problem, then φ = constant in each fluid
( K UU )S CTFU U ( R UB )S
= (2.11-38)
CFU −M FF φ −R FB
K FF φ = 0 (2.11-39)
2 M 0 0 CTFU K + ( K UU )S 0 U ( j) 0
−ω j − ω + =
0 M FF
j
CFU 0 0 K FF F ( j) 0
(2.11-40)
(2.11-43)
n ξ j (j) n
Ur = ∑ U , φ = ∑ −ξ j F (j) (2.11-44a,b)
j =1 ω j j =1
ξ j + ω 2j ξ j = Γ ( j ) (2.11-45)
where
Γ ( j ) = − ( F (j) ) R φ + ω j ( U (j) ) R u
T T
(2.11-46)
R u = R + ( R UB )S − ( K UU )S d k u gk
R φ = CFU d k u gk (2.11-47a,b)
ξ j + 2ω jζ jξ j + ω 2j ξ j = Γ ( j ) (2.11-48)
Then we use
n ξ j (j) n ξ n ξ
Ur = ∑ U , U r = ∑ U (j) , U r = ∑ U (j)
j j
j =1 ω j j =1 ω j j =1 ω j
(2.11-49a,b,c)
n t n n
φ = ∑ ∫ −ξ j dt F (j) , φ = ∑ − ξ j F (j) , φ = ∑ − ξ j F (j)
j =1 0 j =1 j =1
(2.11-50a,b,c)
ξj
xj = (2.11-51)
ωj
n n
U r = ∑ x j U (j) , φ = ∑ − ω j x j F (j) (2.11-52a,b)
j =1 j =1
1
(F )
(j) T
R φ + ( U (j) ) R u (2.11-53)
T
x j + 2ω jζ j x j + ω 2j x j = −
ωj
x j + 2ω jζ j x j + ω 2j x j = ( U (j) ) R
T
(2.11-54)
Γ ( j ) = ( U (j) ) R
T
(2.11-55)
R u = − ( K UU )S d k u gk − Md k u gk ,
R φ = CFU d k u gk (2.11-56a,b)
Hence
1
x j + 2ω jζ j x j + ω 2j x j = − ( F (j) ) CFU d k + ( U (j) ) Md k u gk
T T
ω
j (2.11-57)
− ( U (j) ) ( K UU )S d k u gk
T
( ) (K )
T
We neglect the term U (j) UU S d k u gk (which is probably
small anyway) to obtain
x j + 2ω jζ j x j + ω 2j x j = −Γ ( j )u gk (2.11-58)
1
(F )
(j) T
CFU d k + ( U (j) ) Md k
T
Γ ( j) =
ωj
This factor has the usual physical interpretation, provided that
each fluid region is completely surrounded by interface elements.
The modal participation factor for a φ rigid body mode can be
zero or non-zero. It can be shown that the modal participation
factor for a φ rigid body mode is zero if each fluid region is
completely surrounded by interface elements.
Note: static corrections (residual calculations) are not
implemented for the potential-based fluid element because residual
displacement calculations are based upon non-constant (in space)
body force loading. Such loading is not possible in general for the
potential-based fluid elements.
ωr
waves with <1.
c
Fluid Fluid
n n
1 1
Solid Solid 3
2 2
a) 2-D elements
Fluid Fluid
1 1
n n 8
4 5 4
9
2 7
2
3 6
Solid Solid 3
b) 3-D elements
Structural element
Nodal unknown: u
Fluid-structure
interface element
Nodal unknowns: f, u
Fluid element
Nodal unknown: f Nodes circled by
share degrees of
Assumptions: u is small. freedom
Applied pressure
or no applied pressure
Free surface
interface element
Nodal unknowns: f, u
Fluid element
Nodal unknown: f
Assumptions: u.n is small.
un is small.
Applied tractions
from ADINA-F
ADINA-F
interface element
Nodal unknowns: f, u
Fluid element
Nodal unknown: f
Assumptions: u.n is small.
un is small.
Direction of
wave propagation
indicated by
Fluid-fluid interface
Fluid element of element of element
element group 1 group 1
Nodal unknowns: f(2), un
Nodal unknown: f(1)
Mass-proportional loads
g k = MAGNITUDE × A(k) × TF
3
Ω = ∑ g k ( xk − xk0 ) (2.11-59)
k =1
t t
u gk = − g k , u gk = ∫
tstart
u gk dt , u gk = ∫
tstart
u gk dt (2.11-60a,b,c)
t
gk = sin ωt , t < t p ,
tp
= sin ωt , t ≥ tp
2π
in which t p = is one period of the ground motion.
ω
Centrifugal loads
1) The AUI loops over all fluid element sides on the boundary
of each fluid element region. If the fluid element side already has
an interface element, the side is skipped. If the fluid element side
has a potential-interface, an interface element of the appropriate
type is generated. Otherwise, the side is checked to see if it is
attached to a structure (shares structural degrees of freedom with
structural elements), close to a structure (nodes coincident with
nodes of a structural element) or on a ADINA-F fluid-structure
boundary; and, if any of the above conditions are met, an interface
element of the appropriate type is generated. The intent of this step
is to cover as much of the fluid boundary as possible with interface
elements.
Note that phi model completion relies on determining the nodes
that are “close” together. The tolerances used in this determination
are controlled by command PHI-MODEL-COMPLETION,
parameters CLOSE-TOL, XTOL, YTOL, ZTOL. By default, the
tolerances used during phi model completion are the same
tolerances used in mesh generation coincident node checking .
4) In static analysis, when there are no body force loads, the AUI
loops over all nodes on a free surface, ADINA-F interface, or fluid-
fluid interface. If the node is attached to a structural element, the
node is skipped. Otherwise, constraint equations are defined for all
nodes so that the displacements in the direction of the free normal
are equal.
7) The AUI then loops over all nodes with zero stiffness degrees of
freedom and defines fixities for each zero stiffness degree of
freedom.
Note that the model that results from the phi model completion
process is not stored in the AUI, but is immediately written to the
ADINA data file. Therefore the model that results from phi model
completion cannot be displayed during model definition. Phi
model completion does write some messages to the log file or user
interface, which you may find helpful. However, we recommend
that you also check the model by running the model through
ADINA, for example, for one load step; then use ADINA-PLOT to
display the model.
Potential-interface of type
free surface is defined on
this line
16 3 2 1 10 Potential-based
Fluid nodes are at the same fluid elements
coordinates as structural nodes.
6 5 4 11
The fluid nodes and structural Beam elements
nodes are separated in this figure 15
for clarity. 9 8 7
30o
14 13 12
Interface elements
of type free surface
3 2 1
6 4
9 8 7
Interface elements of
type fluid-structure interface
b z z
y y
c 3 2 1
16 3 1 10
6 4 11
15
9 8 7
14 13 12
Node 1: uy=u10
y
o o
Node 3: uc=-cos30 uy16 - sin 30 uz16
Node 4: uy=u11 11
y , uz=uz Node 6: uy=u15 15
y , uz=uz
Node 7: uy=uz =fixed Node 8: uy=u13 13
y , uz=uz
Node 9: uy=uz =fixed
b z z
y y
c 3 2 1
Node 2: uz=u1z
o
Node 3: ub= uz1 / cos30
Step 4 is only performed in static analysis when
there are no body forces.
This motion affects the total mass of the fluid region, so that
there is no zero pivot in the system matrices.
If there are body forces, then step 4 is not necessary because all
boundary motions are given stiffness by the matrix ( K UU )S .
Step 5 of phi model completion is skipped because there are no
fluid-fluid interfaces.
In step 6 of phi model completion, the zero stiffness direction at
node 2 is fixed (Figure 2.11-12(f)). Vertical motions of the nodes
attached only to free surface interface elements are allowed, but
horizontal motions of these nodes are not allowed (because the
fluid does not provide stiffness, damping or mass to horizontal
motions).
z
y
2
Node 2: uy=fixed
3 2 1
6 5 4
9 8 7
Potential-interface of type
free surface is defined on
this line
11 2 10 Potential-based
3 1 fluid elements
6 5 4 Beam elements
9 8 7
30o
n12 n23
1 2 3
n14 n36
4 5 6
n69
7 8 9
y
10 11 12 x
During step 6 of phi model completion, the AUI fixes the x and
y directions for nodes 5 and 8, because these directions are zero
stiffness directions, and there is no adjacent structure.
15 14 13 2-D potential-based
fluid elements, group 1
3 2 1
Beam elements
20 19 18
15 14 13
3 2 1
6 5 4
17 16
9 8 7
12 11 10
20 19 18
Node 1: uy=u13 13
y , uz=uz Node 2: uy=u14 14
y , uz=uz
Node 3: uy=u15 15
y , uz=uz Node 4: uy=u16
y
Node 6: uy=u17
y Node 7: uy=u16
y
Node 9: uy=u17
y Node 10: uy=u18 18
y , uz=uz
15 14 13
3 2 1
6 5 4
17 16
9 8 7
12 11 10
20 19 18
15 14 13
3 2 1
6 5 4
17 16
9 8 7
12 11 10
20 19 18
region. The initial values of φ in the confined fluid region are that
from the static analysis, so the confined fluid region has the same
initial pressure as in static analysis.
The modeling process is schematically illustrated in Figure
2.11-16.
Potential-interface
of type inlet-outlet
Applied pressure
Confined
fluid region Finite fluid
with nonzero elements
fluid pressure
Auxiliary
Constraint equations fluid element,
automatically generated in same fluid
by the AUI during phi element group
model completion
a) Physical problem b) Modeling in static analysis
Pressure at
beginning of Fix all degrees of freedom
restart analysis attached to the auxiliary fluid
= pressure at element.
end of static
analysis.
Fixed to
ground
uz = constant ground motion
k k
uz is the absolute
Fluid Fluid displacement
No boundary uz = 0
condition
a) Finite element model b) Interpretation of f rigid body mode
Translation alignments
Distance alignments
Rotation alignments
Local
node 2
offset
Local
node 1 Local
offset node 2
Local
node 1
This element can be used to keep the two nodes nearly coincident.
Notice that, although the nodes are nearly coincident, their relative
rotations are not affected by the element, so the nodal rotations can
be very different.
F
Local
node 2
F
Local Misalignment vector
node 1
F
Local
node 2
F
Misalignment vector
F
Prescribed translational alignment
Local
node 1
F
Local
node 2
Misalignment vector
F
Local
node 1 Prescribed translational alignment
(c): Special case, prescribed alignment equals a fraction of the initial misalignment
F F
Prescribed alignment 2b
Misalignment
Offset for
local node 2 Local
1b Actual alignment node 2
Offset for
local node 1
Local
node 1
Figure 2.12-4: Example 3: Translational alignment with
prescribed alignment and nodal offsets
For the distance alignment, the force always acts along the line
connecting points 1b and 2b. The element provides stiffness only in
the direction of the line connecting points 1b and 2b.
Offset for
local node 1
Local
node 1
Figure 2.12-5: Example 4: Distance alignment with prescribed
alignment and nodal offsets
2
a2
1
a2 2
a1
Local q
node 2
Local M M
node 1 1
a1
2
a2
2
d1
1 2
a2 a1
q
Local
node 2
M
Local 1
node 1 g d1
1
M a1
2
b2
2
d1 1
b2 2
b1
q
2b 2
a2
Offset for
M local node 2
1b 1
g d1
1 1
a2 M b1
Local
Offset for node 2
local node 1
2
a1
1
a1
Local
node 1
2.12.2 Triads
In this section the nodal offsets and triads are discussed in more
detail.
d3
d1
b1 c3
c1
b2
c2 xdb d2
a3
a2
b3 xba
Node
a1
A triad
B triad
The origin and coordinate axes of the b triad are defined with
reference to the a triad. Thus the b triad is rigidly attached to the a
triad. In particular, as the node rotates, vectors xba, b1, b2, b3 also
rotate.
C triad
Local node 1 of the alignment element also has a triad that provides
a coordinate system in which the prescribed alignments, total
alignments and misalignments are measured. This triad is called
the "c" triad and the coordinate axes of the c triad are denoted c1,
c2, c3. The c triad is an orthonormal right-handed coordinate
system, with origin at the origin of the b triad. The coordinate axes
of the c triad are defined with reference to the b triad. Thus the c
triad is rigidly attached to both the b triad and the a triad.
D triad
2.12.3 Notations
The notations used in the discussion of the theory are based on the
triad names (a, b, c, d). Although the notations are cumbersome,
the notations clearly show the meanings of the quantities described.
Vectors
Since the above quantities are vectors, they can be expressed using
components in any convenient coordinate system. In order to
denote which coordinate system the components are expressed in,
we use a letter in the subscripts, for example ( xd 1bc11 , xd 1bc12 , xd 1bc13 )
are the components of vector xd1b1 , expressed in the c triad
coordinate system.
Rotation matrices
The first row of the matrix gives the direction b1 for local node 2,
the second row gives the direction b2 for local node 2, and the third
row gives the direction b3 for local node 2. Each of these direction
vectors is expressed using the coordinate system of the b triad of
local node 1.
Clearly if the directions of the b triad axes are the same for both
We also use
c3 1 1 xb2d1 a3
xd b
c2 xb2 a2
Local
xb2b1 node 2
a3
1 1
xb a
a2 a2 a1
c1
Local
node 1 a1
Notice that the directions of the b triad coordinate axes are not used
in the translation calculation.
Prescribed translation
Notice that the directions of the d triad coordinate system are not
used in this calculation.
Translation misalignment
Forces
xb2b1c2
xd1b1c2
Fc1
xb2d1 a2
xd1b1 xb2a2
Local
Fc2 node 2
Fc2 xb2b1
a2 a1
Fc1
xb1 a1
Local xd1b1c1
node 1
a1
xb2b1c1
a3
xb2 a2
Local
a3 xb2b1 node 2
1 1 a2
xb a
a2 a1
Local
node 1 a1
Distance calculation
Also notice that the distance changes if the nodes rotate without
translation, since the offset vectors xb1a1 , xb 2 a 2 also rotate as the
nodes rotate.
Prescribed distance
The user enters a table giving the "prescribed distance" for various
times. The prescribed distance can be specified using the following
options:
Distance misalignment
Forces
a2
xb2 a2
D F Local
node 2
a2 a1
F
DP
1 1
xb a
Local a1
node 1 Figure 2.12-13(a): Distance greater than prescribed distance
a2
D xb2 a2
Local
node 2
a2
P a1
D
1 1
xb a
Local a1
node 1
Figure 2.12-13(b): Distance equal to prescribed distance
F
a2
xb2 a2
D Local
node 2
a2 a1
1 1
xb a
DP
F
Local
a1
node 1
b3
b1
c1 c3
b1 a3
xb2 a2
d1
d2
b2 Local
c2 node 2
a3
b2
a1
b3 xb1 a1 a2
a2 d3
c1
Local
node 1
a1
b11
b32
b21
2
b1 d1
d2
b31
b22 d3
Figure 2.12-14(b): Relative orientation of b and d triads,
corresponding to (a)
Rotation calculation
the form
b1b211 b1b2 21 0
VB 2 B1 = b 2b211 b 2b2 21 0
0 0 1
b21 2
b1
b22
2 b11
b1b21
2
b1b11
b21 2
b1 q
d2 d3d1
P1
gb3
b22 b11
Figure 2.12-15: Rotations in the plane, b3 axes have the same direction
Prescribed rotation
Aligned: VD1B1 = I
(γ P
b11
, γ bP21 , γ bP31 ) = γ P ( nbP11 , nbP21 , nbP31 ) and then calculates the
prescribed rotation matrix using the formula
where S = γ bP31 0 −γ bP11
−γ P 1 γ bP1 0
b2 1
γ
sin γ sin
1 2 S 2 VD1B1
VD B = I +
1 1
S+
γ 2 γ
previous row
2
where S , γ P (n P
c11
, ncP21 , ncP31 ) are defined above. If this option is
used for the first row of the table, then the effect is the same as
if option=Angle.
Rotation misalignment
VB 2 D1 = VB 2 B1 ⋅ (VD1B1 )T
Moments
M c1 = kc1 ⋅ θ c1 , M c 2 = kc 2 ⋅ θ c 2 , M c 3 = kc 3 ⋅ θ c 3 .
Moment
Triadsets
Node-triadset pair
ALIGN-TRANSLATION 1
ENTRIES TIME OPTION XCP YCP ZCP
TBFACTOR
CLEAR
3.0 TBFACTOR 0.0 0.0 0.0 1.0
4.0 COMPONENTS 1.0 2.0 3.0 0.0
5.0 ALIGNED
DATAEND
The lowest time in the table is time 3.0. For solution times before
time 3.0, the translational alignment is inactive (not born yet).
For the first solution time at or after time 3.0, the translational
alignment becomes active. The total alignment for the preceding
solution time is measured and is used to construct xb 2b1 . Since
birth
OPTION=TBFACTOR and TBFACTOR=1.0, the prescribed
translational alignment for time 3.0 is set to xd1b1 = xb 2b1 .
birth
TIMESTEP
CLEAR
1 1.0
DATAEND
*
ALIGN-DISTANCE 1
ENTRIES TIME OPTION
CLEAR
1.0 ALIGNED
DATAEND
*
TIMESTEP
CLEAR
2 1.0
DATAEND
*
ALIGN-DISTANCE 1
ENTRIES TIME OPTION
CLEAR
1.0 TBFACTOR
2.0 ALIGNED
DATAEND
*
EGROUP ALIGNMENT 1 ALIGN-DISTANCE=1
EDATA
ENTRIES N1 N2
CLEAR
1 2
DATAEND
TIMESTEP
CLEAR
3 1.0
DATAEND
*
ALIGN-DISTANCE 1
TIMESTEP
CLEAR
4 1.0
DATAEND
*
ALIGN-DISTANCE 1
ENTRIES TIME OPTION FACTOR
CLEAR
2.0 TBFACTOR
3.0 TBFACTOR 0.1
4.0 ALIGNED
DATAEND
*
EGROUP ALIGNMENT 1 ALIGN-DISTANCE=1
EDATA
ENTRIES N1 N2
CLEAR
1 2
DATAEND
TIMESTEP
CLEAR
5 1.0
DATAEND
*
ALIGN-DISTANCE 1
ENTRIES TIME OPTION FACTOR
CLEAR
2.1 TBFACTOR
3.1 TBFACTOR 0.1
4.1 ALIGNED
DATAEND
*
The nodes are aligned when the b triad axes of the two nodes are
coincident. Since no triadset is defined, the b triad axes correspond
to the r-s-t axes of the attached beam elements.
ALIGN-ROTATION 1
ENTRIES TIME OPTION
CLEAR
3.0 TBFACTOR
4.0 ALIGNED
DATAEND
*
EGROUP ALIGNMENT 1 ALIGN-ROTATION=1
EDATA
ENTRIES N1 N2
CLEAR
1 2
DATAEND
The b triad has offset (1, 2, 3). This offset is measured in the a1, a2,
a3 directions. The b triad directions are b1=(0,1,0), b2=(0,0,1). b3
is computed to create a right-handed orthonormal system. Again,
the components of the b1, b2, b3 axes are expressed in the a triad
coordinate system.
The b triad has offset (-1, -2, -3). The b triad directions are b1=(-
1,0,0), b2=(0,1,0) (again, the components of b1, b2 are expressed in
the a triad coordinate system). b3 is computed to create a right-
handed orthonormal system.
The nodes are aligned when the b triad axes of the two nodes are
coincident. The triad offsets are not used in this example.
TRIADSETS
ENTRIES,
TRIADSET,
AOPTION A1X A1Y A1Z A2X A2Y A2Z,
BOPTION B0A1 B0A2 B0A3,
B1A1 B1A2 B1A3 B2A1 B2A2 B2A3,
CLEAR
1,
VECTORS 0.0 -1.0 0.0 -1.0 0.0 0.0,
VECTORS 1.0 2.0 3.0,
0.0 1.0 0.0 0.0 0.0 1.0
2,
VECTORS 1.0 0.0 0.0 0.0 0.0 1.0,
VECTORS -1.0 -2.0 -3.0,
-1.0 0.0 0.0 0.0 1.0 0.0
DATAEND
SET TRIADSET NODES
CLEAR
1 1
2 2
DATAEND
ALIGN-ROTATION 1
ENTRIES TIME OPTION
CLEAR
3.0 TBFACTOR
4.0 ALIGNED
DATAEND
*
EGROUP ALIGNMENT 1 ALIGN-ROTATION=1
EDATA
ENTRIES N1 N2
CLEAR
1 2
DATAEND
TRIADSETS
ENTRIES,
TRIADSET,
ALIGN-TRANSLATION 1
ENTRIES TIME OPTION
CLEAR
1.0 TBFACTOR
2.0 ALIGNED
DATAEND
*
ALIGN-ROTATION 1
ENTRIES TIME OPTION
CLEAR
3.0 TBFACTOR
4.0 ALIGNED
DATAEND
*
EGROUP ALIGNMENT 1 ALIGN-TRANSLATION=1,
ALIGN-ROTATION=1
EDATA
ENTRIES N1 N2
CLEAR
1 2
DATAEND
The nodes are rotationally aligned when the b triads at node 1 and 2
have the same directions.
ALIGN-TRANSLATION 1
ENTRIES TIME OPTION
CLEAR
5.0 TBFACTOR
6.0 ALIGNED
DATAEND
*
ALIGN-ROTATION 1
ENTRIES TIME OPTION ANGLE AXISC1 AXISC2
AXISC3
CLEAR
1.0 TBFACTOR
2.0 ANGLE 2.1 0.0 0.0 1.0
3.0 ANGLE 4.2 0.0 0.0 1.0
4.0 ALIGNED
DATAEND
*
---
Radius = 3
s
r
Node 8 t
s Radius = 2
r
t
Node 5
(a): At alignment
a2=s
a1=r
Triadset 8 a3=t
b2
b1
b3
b2
b1
b3
a2=s
a1
a1=r b1
Triadset 5
a3=t
a3
30 o
b3
The triadsets are defined so that the b triads are aligned in the
aligned configuration:
a triads: The a triads are the same as the triads from the attached
elements.
b triads, triadset 5: The b triads are offset from the node in the a2
direction:
The b triad directions are the same as the a triad directions (the
default)
b triads, triadset 8: The b triads are offset from the node in the
negative a2 direction:
TRIADSETS
ENTRIES TRIADSET,
AOPTION,
BOPTION B0A1 B0A2 B0A3,
B1A1 B1A2 B1A3 B2A1 B2A2
B2A3,
CLEAR
5,
ELEMENTS,
VECTORS 0.0 2.0 0.0
8,
ELEMENTS,
VECTORS 0.0 -3.0 0.0,
0.86603 0.0 0.5 0.0 1.0
0.0
DATAEND
SET TRIADSET NODES
CLEAR
5 5
8 8
DATAEND
The element can either output nodal point forces or alignment data
(the choice is made in the EGROUP ALIGNMENT command).
When the element outputs alignment data, two levels of output can
be selected:
be selected.
PRINTOUT PRINTDEFAULT=STRAINS
EGROUP ALIGNMENT PRINT=DEFAULT
EDATA ... printi=DEFAULT
or
or
Verbose output includes the usual output described above, and also
the following addtional output:
Translation:
xd1b1 in component form: ( xd 1b1c1 , xd 1b1c 2 , xd 1b1c 3 ) . The
postprocessing variables are ALIGN_TRANS_PRESC-
C{123}.
Distance:
D P . The postprocessing variable is ALIGN_DIST_PRESC.
Rotation:
2.12.15 Recommendations
1 2
2-D Cohesive Element
3 2
4 1
Thickness = 0
4 1
7 6
ref. Turon, A., Camanho, P.P., Costa, J., and Davila, C.G.,
"A Damage Model for the Simulation of Delamination in
Advanced Composites under Variable-Mode Loading,"
Mechanics of Materials, Vol. 38, No. 11, pp. 1072-1089,
2006.
K reloading reloading
unloading
∆
α α
GI GII
+ =1
GIC GIIC
where GI and GII are the mode I and mode II energy release rate;
GIC and GIIC are the mode I and mode II fracture toughness; α is
η
G
GC = GIC + ( GIIC − GIC ) S
GT
2 4
r r
1 2
2-D cohesive element
1 3
3-D cohesive element
COHESIVE_NORMAL_STRESS,
COHESIVE_TANGENTIAL_STRESS,
COHESIVE_NORMAL_STRAIN,
COHESIVE_TANGENTIAL_STRAIN, COHESIVE_ DAMAGE
Output:
For the shell elements, the following material models can be used,
provided that the elements are single layer elements described
using 3, 4, 6, 9 or 16 midsurface nodes:
1 2
− 2
Green-Lagrange strain: ε= 2
0
2 0
Stretch: λ=
0
A
´
A0 D
A0 = initial section area
A = final section area
F = applied force
0 = 0 +D
F
Figure 3.1-1: Rod under uniaxial tension
F
Engineering stress: σ=
A0
F σ A0
Cauchy stress: τ= =
A A
F 0 σ
2nd Piola-Kirchhoff stress: S = = 0
A0
A F σ
Kirchhoff stress: Jτ = τ= =
A0 0 A0 0 0
• When the strains are small, the 2nd Piola-Kirchhoff stresses are
nearly equal to the Cauchy stresses from which the rigid body
rotations have been removed.
X = RU (3.1-1)
U = R L Λ R TL (3.1-2)
y R
(t) RL
(0)
x
E R = ln U = R L ln Λ R TL (3.1-3)
V = R E Λ RTE (3.1-5)
E L = ln V = R E ln Λ RTE (3.1-6)
X = XE XP (3.1-7)
XE = R E UE = V E R E (3.18-a,b)
E ER = ln U E , E EL = ln V E (3.1-9a,b)
τ = ( R E ) Jτ R E
T
(3.1-10)
and when the left Hencky strain measure is used, the stress measure
is the (unrotated) Kirchhoff stress Jτ . J = det X is the volume
change of the material, and, using det X P = 1, J = det X E .
With these choices of stress and strain measures, the stresses and
strains are work-conjugate.
These assumptions are justified because they are used with material
models in which the plastic deformations are incompressible and
the plastic deformations are generally much larger than the elastic
deformations.
For further information, see ref KJB, Section 6.2.2 and also the
following reference:
∂ tu
L = t i = XX −1 (3.1-11)
∂ x j
1
D=
2
( L + LT ) (3.1-12)
1
W=
2
( L − LT ) (3.1-13)
ε J = D +Wε J − ε J W (3.1-14)
t +∆t
ε J = t ε J + ( D∆t ) +( W∆t ) t ε J − t ε J ( W∆t ) (3.1-15)
t TH
e ij = α ( tθ − 0θ ) δ ij (3.1-16)
t TH
ije = tα ( tθ − 0θ ) δ ij (3.1-17)
where
t
α=
(
1
θ − θ)
t 0 (
α ( tθ )( tθ − θ REF ) − α ( 0θ )( 0θ − θ REF ) )
(3.1-18)
and θ REF is the material reference temperature and is input as a
material property, see below for more details.
t TH
e ij = α i ( tθ − 0θ ) δ ij (no summation over i ) (3.1-19)
L
Secant to curve
Length, L
LREF
qREF q
Temperature, q
L − LREF
ε TH =
LREF
Then we define the mean coefficient of thermal expansion for a
given temperature as follows:
ε TH (θ )
α (θ ) =
θ − θ REF
t
ε TH = α ( tθ )( tθ − θ REF ) − α ( 0θ )( 0θ − θ REF )
• This material model is available for the truss, 2-D solid, 3-D
solid, beam, iso-beam, plate, shell and pipe elements.
1l b
Z
l 8
5l b
Y a l4
l
2l l 7
6 l3
Figure 3.2-1: Principal in-plane material axes orientation for the
orthotropic material model for 2-D solid elements
Plane stress:
ea 1 ν ab ν ac σa
− − 0
Ea Eb Ec
eb ν ba 1 ν bc
− − 0 σb
= Ea Eb Ec
ν ca ν cb 1
ec − − 0 σc = 0
Ea Eb Ec
1
γ ab 0 0 0 σ
Gab ab
Plane strain:
ea 1 ν ab ν ac σa
− − 0
Ea Eb Ec
eb ν ba 1 ν bc
− − 0 σb
= Ea Eb Ec
ν ν cb 1
ec = 0 − ca − 0 σc
Ea Eb Ec
1
γ ab 0 0 0 σ
Gab ab
Axisymmetric:
ea 1 ν ab ν ac σa
− − 0
Ea Eb Ec
eb ν ba 1 ν bc
− − 0 σb
= Ea Eb Ec
ν ca ν cb 1
ec − − 0 σc
Ea Eb Ec
1
γ ab 0 0 0 σ
Gab ab
The seven material constants (Ea, Eb, Ec, vab, vac, vbc and Gab)
define the symmetric compliance matrix C−1 .
1 ν ν ac
− ab 0 −
Ea Eb Ec
1 νbc
0 −
Eb Ec
C−1 =
1
0
Gab
1
symmetric
Ec
where the subscript " " in C−1 indicates that the material law is
given in the local system of the material axes. Note that the
determinant of C−1 must not be zero in order to be able to calculate
the inverse C . This imposes the following restrictions on the
constants. The material constants must be defined so that the
stress-strain constitutive matrix is positive-definite; i.e.,:
1
Ei 2
ν ji < , i, j = a, b, c
E
j
Ea E E
ν abν bcν ca < 0.5 1 −ν ab2 −ν bc2 b −ν ca2 c ≤ 0.5
Eb Ec Ea
or
Ea 2 Ea E E
ν abν bcν ac < 0.5 1 −ν ab −ν bc2 b −ν ac2 a ≤ 0.5
Ec Eb Ec Ec
ν ji ν ij
=
Ei Ej
εb
ν ba = −
εa
F eb F
b
a ea
In general,
εi
ν ij = − (i, j = a, b, c)
εj
εj
J
ν ij = − .
εi
ν ji = Jν ij
and the relationship
ν ji ν ij
=
Ei Ej
Ej J
ν ij = ν ij
Ei
C = QT C Q
where
2
1 m12 1 2 0
2 2
m m1m2 0
Q= 2 2
2m1m2 2 1 2 1 m2 + 2 m1 0
0 0 0 1
1/ Ea −ν ab / Eb −ν ac / Ec 0 0 0
1/ Eb −ν bc / Ec 0 0 0
1/ Ec 0 0 0
C−1 =
1/ Gab 0 0
1/ Gac 0
symmetric 1/ Gbc
ea 1/ Ea −ν ab / Eb −ν ac / Ec 0 σa
e −ν / E 1/ Eb −ν bc / Ec 0 σb
b = ba a
ec −ν ca / Ea −ν cb / Eb 1/ Ec 0 σc = 0
γ ab 0 0 0 1/ Gab σab
Note that the transverse normal strain ec is not calculated for the
plate elements.
z,c
b
The a-b and x-y planes 1
l
are coplanar. Nodes
1, 2 and 3 belong to
b y
the x-y plane.
a
Z
l 3
2 l
Y
X
x
C = QT C QT
ea 1/ Ea −ν ab / Eb −ν ac / Ec 0 0 0 σa
e −ν / E 1/ Eb −ν bc / Ec 0 0 0 σb
b ba a
ec −ν ca / Ea −ν cb / Eb 1/ Ec 0 0 0 σc = 0
=
γ ab 0 0 0 1/ Gab 0 0 σ ab
γ ac 0 0 0 0 1/ Gac 0 σac
γ bc 0 0 0 0 0 1/ Gbc σbc
t ,c
b
a
3
r
b
1
s d
r12
2
(a) 4-node element
t, c
l l
b l
l
l
a
rl l
b 3
1
l l
l
r12 s d
l l
2
(b) 16-node element
• Note that the material unloads along the same curve, so that no
permanent inelastic strains are introduced.
σ σs
εe εe
∫ σ dε = ∫ σ
T
dε (3.3-1)
1 1
E0ε 2 = εT C0ε (3.3-2)
2 2
1 T
dε = ε C0 dε (3.3-3)
E0ε
σ
σ= C0 ε (3.3-4)
E0ε
or
σ
σ= σe (3.3-5)
E0ε
t +∆t
σ e = C0 t +∆t ε (3.3-6)
I
1e
r =C (3.3-7)
t +∆t
σe
t +∆t
σ e = E0 t +∆t ε (3.3-8)
t +∆t
σ c = −t +∆t σ (−t +∆t ε ) (3.3-10)
1 + r t +∆t 1 − r t +∆t
t +∆t
σ= σt + σc (3.3-11)
2 2
t +∆t
t +∆t
σ t +∆t
σ= σe (3.3-12)
E0 t +∆t
ε
Stress
s6 l
s5 l
s4 l
e1 e2 e3
s3 e4
l
e5 e6 Strain
l s2
l s1
Stress
Point 3
4.32´104 ´
Point 1 Point 2
´ ´
-1.0 1.0 Strain
Pipe (beam)
n
l
Gap D
Gap element L
m
l
Stress
D
_
L
Strain
Stress Stress
e2
´
e1 e2 e3 e4 e1 Strain
´ ´ ´ ´
Strain e3 e4
´
´
´
Compression Stress
cut-off limit
e6
´
e1 e2
´ ´ ´ ´
e7 e8 Strain
e5
e4 ´
e3 ´ ´ Tension
cut-off limit
• These models can be used with the truss, 2-D solid, 3-D solid,
beam (plastic-bilinear only), iso-beam, shell and pipe elements.
Stress
0s
y
Bilinear stress-strain curve
Strain
Ö 23
0
sy
Ö 23
0
sy
s2 s1 s2
Elastic region
s1
a) Principal stress space b) Deviatoric stress space
Stress Stress
ET ET
0s ts 0s
y y y
E E 2 0sy
Strain E Strain
E ts
y E
E ET
ET
a) Bilinear isotropic hardening b) Bilinear kinematic hardening
Stress Stress
ts
0s y
0s 2 0sy
y y
Strain Strain
ts
y
1 t t 1
t
fy =
2
( s ⋅ s ) − tσ y2 = 0
3
stress at time t.
In the von Mises model with kinematic hardening, the following
yield surface equation is used:
1 t 1
t
fy =
2
( s − t α ) ⋅ ( t s − t α ) − 0σ y2 = 0
3
where t α is the shift of the center of the yield surface (back stress
tensor) and 0 σ y2 is the virgin, or initial, yield stress.
In the von Mises model with mixed hardening, the following
yield surface equation is used:
1 t 1
t
fy =
2
( s − t α ) ⋅ ( t s − t α ) − tσ y2 = 0
3
where
t
σ y = 0σ y + ME p e p
dα = C p (1 − M ) de p
2
Cp = Ep
3
and M is the factor used in general mixed hardening (0 < M < 1)
which can be a variable, expressed as
M = M ∞ + ( M 0 − M ∞ ) exp ( −η e p )
The formulation for the von Mises model with mixed hardening
is given in the following reference:
Note that the convergence might not be good when the mixed
hardening parameter η ≠ 0 . Therefore, η = 0 is preferred.
The yield stress is a function of the effective plastic strain,
which defines the hardening of the material. The effective plastic
strain is defined as
t
P 2 p
t
e =∫ de ⋅ de p
0
3
∑
P P
t
e = ∆e
all solution steps
P 2 p
where ∆ e = ∆e ⋅ ∆e p and ∆e p is the tensor of plastic strain
3
increments in a solution step. Because of the summation over the
P
solution steps, we refer to the calculated value of t e as the
accumulated effective plastic strain.
ε P
σ y = σ y0 1 + b ln 1 +
ε 0
• The Mroz-bilinear model can be used with the 2-D solid and
3-D solid elements.
Stress
C ETB
syB B
Bounding line ET 2 sy
A E
sy
E
D
ET
O Strain
H F
ET
E Bounding line
2 sy
F
G ETB
F¢
s1 s2 s1 G¢ s2
Case I:
The yield surface translates in the deviatoric stress space
during the time step, but does not come into contact with the
bounding surface.
Case II:
- The yield surface translates in the deviatoric stress space
until it reaches the bounding surface, after which both
surfaces translate together along their common contact
point normal.
The implicit Mroz translation rule for the yield surface is based
on a target normal tensor defined as
t +∆t SE − t β
t= E t
S − β
where S E is the trial stress for the time step and t β is the back
stress of the bounding surface at the end of the previous time step.
The target normal tensor is explicit in the sense that no iterations
are required to determine it, but implicit in the sense that it depends
on the trial stress for the time step. This target normal tensor is the
same for both Case I and Case II. In order for the target normal
tensor to be uniquely defined, the following condition on the
choice of the yield stress and the bounding stress must be satisfied:
σ y < σ yB < 2σ y
Case I:
2
t +∆t
m = tβ − tα +
3
( σ yB − σ y ) t +∆t t
with the increment in the back stress of the yield surface given by
t +∆t
α α m
∆α = t +∆t
α − α = ∆γ C p
t
t +∆t
.
m
2 E ET
C αp = .
3 E − ET
The plastic strain increment for the step is given by the associated
flow rule
t +∆t
S − t +∆t α
∆e p = ∆γ α t +∆t n = ∆γ α t +∆t
S − t +∆t α
1 t +∆t 1
fy =
2
( S − t +∆t α ) ⋅ ( t +∆t S − t +∆t α ) − σ y2 = 0 .
3
2mDe S
E
t+Dt
n
t+Dt
t
S S s3
t+Dt
a
t t+Dt
b t
t
a t+Dt
m
t+Dt
m
s1 s2
(a) Case I; Only the yield surface translates during the time step
E
S
2mDe
t+Dt
S
t+Dt
a
t
S s3
t+Dt
b
t t+Dt
b t
t
a
s1 s2
(b) Case II; Both surfaces translate during the time step
Case II:
∆β = t +∆t
β − t β = ∆γ β C pβ t +∆t t
By geometry, the position of the yield surface at the end of the time
step is related to that of the bounding surface through
2
t +∆t
α= t +∆t
β+
3
( σ yB − σ y ) t +∆t t .
2
∆α = t +∆t
α − t α = t β − t α + ∆γ β C pβ +
3
( σ yB − σ y ) t +∆t t
∆α
∆γ α = .
C αp
Finally, the plastic strain increment for the step is given by the
associated flow rule
∆e p = ( ∆γ α + ∆γ β ) t +∆t t .
1 t +∆t 1 2
f yB =
2
( S − t +∆t β ) ⋅ ( t +∆t S − t +∆t β ) − σ yB
3
=0
• This model can be used with the 2-D solid, 3-D solid and shell
elements.
F (σ bb − σ cc ) + G (σ cc − σ aa ) + H (σ aa − σ bb )
2 2 2
1 1 1 1
F= 2+ 2− 2
2Y Z X
1 1 1 1
G= 2 + 2 − 2
2 Z X Y
1 1 1 1
H= 2+ 2− 2
2 X Y Z
1 1 1
L= 2 ; M = 2 ; N = 2
2Yab 2Yac 2Ybc
0
A B sbb
saa
X
f (σ bb − σ cc ) + g (σ cc − σ aa ) + h (σ aa − σ bb )
2 2 2
+ 2 σ ab
2
+ 2mσ ac2 + 2nσ bc2 − σ y2 = 0
where
f = σ y2 F , g = σ y2G , h = σ y2 H , = σ y2 L , m = σ y2 M , n = σ y2 N
and
11
σ y2 = ( X 2 + Y 2 + Z 2 ) + Yab2 + Yac2 + Ybc2
23
σ a2 = f (σ bb − σ cc ) + g (σ cc − σ aa ) + h (σ aa − σ bb )
2 2 2
+ 2 σ ab
2
+ 2mσ ac2 + 2nσ bc2
• There are two options for the input of the universal plastic
modulus and the initial yield stress:
(
1 1 p2 p2 p2 p2 p2
) p2
Ea + Eb + Ec + Eab + Eac + Ebc
23
where
Ei EiT
Eip = i = a ,b ,c
Ei − EiT
and
Eij EijT
E =p
ij ij = ab,ac,bc
Eij − EijT
σ y = C ( e0 + eaP )
n
( r0 + r90 )( 2r45 + 1)
2 = , m = 1.5 , n = 1.5
r90 ( r0 + 1)
t
f y = t QM + t QB + t QMB γ − tσ y2 = 0
t
QM =
1
h2
{( N ) + ( N ) − N ⋅ N + 3( N ) }
t
1
2 t
2
2 t
1
t
2
t
12
2
t
QB =
16
h4
{( M ) + ( M ) − M ⋅ M
t
1
2 t
2
2 t
1
t
2 + 3 ( t M 12 )
2
}
12 1 t 1
3 (
t
QMB = N1 ⋅ t M 1 + t N 2 ⋅ t M 2 ) + t N12 ⋅ t M 12 − ( t N1 ⋅ t M 2 + t N 2 ⋅ t M 1 )
h 3 6
Node 3
tM tN
1 2
tN
12
tN
1
tM
tM 2
_ 12
x3 _
Node 2 x1
Node 1
(a) Element section forces and moments
t
M
2
t
syh
( 4 )
g=0
1.0 1_
g= _
Ö3
0 1.0 t_
N
(tsy h)
• The Gurson model is available for the 2-D solid and 3-D solid
elements.
2
q
σ0
3 q p
2 σ0
2
φ = + 2q1 f * cosh − 2 − 1 + q3 f * = 0 ( )
in which φ is the yield function, q is the von Mises stress, σ 0 is
the equivalent tensile flow stress, p is the pressure (positive in
compression), f * is based on the void volume fraction and q1,q2,
q3 are the Tvergaard constants. Note that this reduces to the von
Mises yield function when f * equals zero. f * is defined as
f for f ≤ f c
f =
*
fu − f c
*
fc + f − f ( f − fc ) for f > f c
F c
df = (1 − f ) d ε kkp + Ad ε p
fN 1 ε p − ε 2
A= exp − N
S N 2π 2
S N
• This material model can be used with the truss, 2-D solid, 3-D
solid, Hermitian beam, shell and 3-D shell elements.
t +∆t
τ =t +∆t s + t +∆tτ m I
t +∆t
e= t +∆t
e′ + t +∆t em I
τ m = 3κ t +∆t em
t +∆t
t +∆t
s = 2G ( t +∆t e′ − t +∆t e P )
where κ and G are the bulk modulus and shear modulus, and
t +∆t P
e is the plastic strain ( ε P in ref LC). Thermal strains are not
included in the above equations, but are included in ADINA when
there are thermal effects.
1 2 1
t +∆t
fy = t +∆t
s − t +∆t α − t +∆t
σ y2 = 0
2 3
where t +∆t
α is the back stress tensor ( X and X′ in ref LC, note
that back stress is always deviatoric) and t +∆t
σ y is the yield stress
( σ Y + R or k in ref LC). The norm of a symmetric tensor a is
defined as a = a : a . The yield condition is always evaluated
at time t + ∆t .
t+Dt
a
2 t+Dts
y
3
s1 s2
Flow rule: The flow rule states that the direction of plastic strain
increments is normal to the yield surface.
directions of ∆e P = directions of ( t + β ∆t
s− t + β∆t
α)
Figure 3.4-10 shows the evolution of the yield surface using the
stress-strain law, yield condition and flow rule.
s3
Ds t+Dt
ts s
t+Dt
Da a
ta
2 t+Dts
y
2 ts 3
3 y
s1 s2
a) Overview
Figure 3.4-10: Incremental update of yield surface and plastic
strains
Yield surface
at time t+Dt
2GDe’
ts
-2GDeP
t+Dt
s
Yield surface
at time t
2GDe’
ts
n -2GDeP
t+Dt
s DeP parallel to n
3 t +∆t t +∆t
q . (In ref LC, ξ is written ζ ). This surface can be
2
written as
2 t +∆t 2
fm = e P − t +∆t ξ − t +∆t q 2 = 0
3
e3P
t+Dt
fm=0
t+Dt
eP
t+Dt
x
3 t+Dtq
2
e1P
e2P
The differential rules used for the evolution of the strain memory
surface are
dq = η (directions of de P ) : (directions of ( t e P − t ξ )) de P
and
directions of dξ = directions of ( t e P − t ξ )
in which the symbol <> means that <u>=0 when u<0, and <u>=u
when u>0. The incremental versions of these rules are
and
n
t+Dt
t
eP De P eP
t+Dt
Dx x
tx
3 t+Dtq
2
3 tq
2 Dx parallel to n
1
η is a material constant that must be between 0 and . Typically
2
1
η= .
2
q
e1P
x1
Time
sy q=q2
q=q1 R
Memory-exponential
Q(q2 ) hardening
Q(q1)
0
sy
Exponential hardening
Q Multilinear hardening
0
sy
(e i , syi)
P
Bilinear hardening
0
sy EP
eP
Bilinear hardening:
t +∆t
σ y = 0σ y + EP t +∆t e P
EET
For bilinear hardening, EP = where ET is the slope of the
E − ET
stress-strain curve during plasticity (this formula assumes no
kinematic hardening).
Multilinear hardening:
Exponential hardening:
σ y = 0σ y + t +∆t R
t +∆t
where t +∆t R is the change in the size of the yield surface. t +∆t
R is
defined using the differential equation
dR = b( tQ − t R)de P
t +∆t
and is calculated using R = t R + ∆R ,
∆R = b( t +∆t Q − t +∆t R)∆e P . t +∆tQ is the asymptotic change in
yield surface size, and is calculated as
t +∆t
Q = Q( t +∆t q ) = QM − (QM − Q0 ) exp(−2µ t +∆t q )
t +∆t
where q is the size of the strain memory surface.
t +∆t
α = ∑ t +∆t α ( m )
2 (m) P
dα ( m ) = h de − ζ ( m ) t α ( m ) de P
3
t +∆t
α ( m ) = A ( m ) − ( A ( m ) − t α ( m ) ) exp(−∆b( m ) )
2 h( m )
where A (m)
= (directions of ∆e P ) , ∆b( m ) = ζ ( m ) ∆e P .
3ζ (m)
∆σ h ∆e P
= tanh ζ +σ y
2 ζ 2
Thus, given several cycles for different cyclic strains, the material
constants can be estimated using the above formula.
s
B
A EP 2sy
Ds
eP
DeP
Figure 3.4-15: Stable uniaxial plastic cycle using one term in
Armstrong-Fredrick nonlinear kinematic hardening
rule
are used to specify the constants for the isotropic part of the plastic-
cyclic model.
is used to specify the constants for the kinematic part of the plastic-
cyclic model.
is used to specify the constants for the rupture part of the plastic-
cyclic model. Rupture is based on the accumulated effective
plastic strain.
Input examples:
is equivalent to
PLCYCL-ISOTROPIC MULTILINEAR 1
0 2E8
1E-3 2.5E8
2E-3 2.7E8
MATERIAL PLASTIC-CYCLIC 1 E=2.07E11 NU=0.3,
DENSITY=7800 PLCYCL-ISOTROPIC=1
DENSITY=7800 PLCYCL-ISOTROPIC=1,
PLCYCL-KINEMATIC=1
Conversion formulas
Plastic-bilinear to plastic-cyclic:
E ET
using the formula EP = .
E - ET
Plastic-multilinear to plastic-cyclic:
PLCYCL-ISOTROPIC MULTILINEAR 1
aeps1 stress-radius1
aeps2 stress-radius2
...
aepsi stress-radiusi
MATERIAL PLASTIC-CYCLIC ... E=E ...,
PLCYCL-ISOTROPIC=1
When there are thermal effects, thermal strains are also output.
When the strain memory surface is used, the strain memory size
t +∆t
q is also output.
• In the data input for the analysis, the nodal point temperatures
must be defined for all time steps (calculated by the ADINA heat
flow solution option or reading these temperatures from a file
created by ADINA-T, reading these temperatures from a mapping
file or defining them by time functions). If the model includes
axisymmetric shell or shell elements, the nodal point temperature
gradients must also be defined for all time steps. See Section 5.6
for more information on prescribing temperatures and temperature
gradients.
• For these models, the elastic moduli, the shear moduli, the
Poisson's ratios and the coefficients of thermal expansion are input
as piecewise linear functions of the temperature, as illustrated in
Fig. 3.5-1. Linear interpolation is used to calculate the material
properties between input points.
E n
l l
tn
t
E l l
l l l l
q1 tq q2 q3 q4 q1 tq q2 q3 q4
Temperature Temperature
a
l
ta
l
l l
q1 tq q2 q3 q4
Temperature
s
l 1 Temperature t q1
l
l
l
l
r
l
Integration points
l
ref. KJB
• The thermo-elasto-plasticity and creep models include the
Section 6.6.3 effects of
t
σij = t Cijrs
E
( t ers − t ersP − t ersC − t ersTH )
where t σij is the stress tensor at time t and t Cijrs
E
is the elasticity
tensor at the temperature corresponding to time t. The tensor t Cijrs
E
Irradiation strains, φ eW
rs
• Note that the constitutive relations for the thermal, plastic and
creep strains are independent of each other; hence the only
interaction between the strains comes from the fact that all strains
affect the stresses. Figure 3.6-1 summarizes the constitutive
description for a one-dimensional stress situation and a bilinear
stress-strain curve.
ref. M.D. Snyder and K.J. Bathe, "A Solution Procedure for
Thermo-Elastic-Plastic and Creep Problems," J. Nuclear
Eng. and Design, Vol. 64, pp. 49-80, 1981.
α J 3 ( t σ ) + β J1 ( t σ ) + (1 − α − β ) J 2 ( t σ ) ≤ σ u
Area A
l l
tP
Temperature tq
Stress
ts
ts (t q)
T( q)
tE t
yv
tE(tq)
Strain
teP (time independent)
teE (time independent)
Creep
strain
teC (time dependent)
t Time
• The plastic strains are calculated using the von Mises plasticity
model (see Section 3.4.1) with temperature-dependent material
parameters (Young's modulus, strain hardening modulus, Poisson's
ratio, yield stress, etc...). Figure 3.6-2 shows the assumptions used
for the dependence of the material parameters on temperature (in
the case of a bilinear stress-strain law). (Note that the plastic
modulus
EP = E · ET /(E- ET)
1t t 1t 2
t
fy = s ⋅ s − σ yv
2 3
1 t 1
t
fy =
2
( s − t α ) ⋅ ( t s − t α ) − tσ y2v
3
dα ij = t CdeijP
2 t E t ET
t
C=
3 t E − t ET
q1
Stress
ET 1
syv1
q2
E1
syv2 ET2 ..
.. .
. E2
syvi qi
ETi
Ei
Strain
(a) Uniaxial relation of stress versus total strain
Stress q1
E1 ET1
syv1
E1 - ET q2
1
..
syv2 E2 ET2 .
.. E2 - ET
. 2 qi
syvi Ei ETi
Ei - ET
i
Plastic strain
Stress 3 4
2
1
qi
e 1i e i2 ei3 e4i Strain
3 4
1 2
qi+1
3
1
e i+1 2
e i+1 e i+1 e 4i+1
Temperature
a) Stress-strain curves input data
Stress 4q
3 i
2
Interpolated
yield curve 1
tq
4 Plastic
3
1 2 qi+1 strain
P2 P3
e i+1 e i+1
Temperature
b) Yield curves
t
e C = a0 tσ a1 t a2
t
e C = F (1 − e − Rt ) + Gt
with
a4
a1 ⋅ tσ tσ t
F = a0 e ; R = a2 ; G = a5 e a6 ⋅ σ
a3
t
e C = S ⋅ T ⋅ e− H
with
a7
S = a0 tσ a1 ; T = t a2 + a3 t a4 + a5 t a6 , H = t
θ
t t
In the above equations, t e C , σ and θ denote the effective
creep strain, effective stress and temperature at time t. The unit
of the temperature can either be K or °C (default). ADINA
automatically applies the conversion to K in the default setting.
d te C 1 GK ( tσ ) 1 t
= exp − t + σ
dt η K ( tσ ) η K ( tσ ) η M ( tσ )
or
−1 GK ( tσ ) t 1 t
t
e =
C
exp − t + + σ
GK ( tσ ) η K ( tσ ) η M ( tσ ) GK ( tσ )
( σ ) =η e m σ , GK ( tσ ) = G*K e k1
t t
σ
where η M t *
M
and
η K ( tσ ) = η * e k
t
σ
K
2
. Also, as usual, t e C is the effective creep
d te C
strain, is the effective creep strain rate and tσ is the
dt
effective creep stress.
In these equations, m, k1, k2, η M* , G *K and η K* are six
material parameters. m, k1, k2 account for the stress effect on
parameters η M* , G *K , η K* . The six material parameters are
identified in ADINA as a0, a1, a2, a3, a4 and a5 respectively.
Unloading and cyclic loading effects are considered in the
LUBBY2 creep law.
Analytical solutions with the LUBBY2 creep law are very
sensitive to the three parameters m, k1, k2, so these parameters
must be very carefully specified.
For more information about the LUBBY2 creep law, see the
following reference:
t
e C = C1 {1 − exp(− r1t )} + C2 {1 − exp(− r2t )} + Gt
with
A0G A1 A G A3
C1 = , C2 = 2
r1 r2
r1 = A4t R− A5 , r2 = A6t R− A7 , G = Ft R− A8
A11
F = A10 exp t
θ + 273.16
A12 A14 A15
B0 = t + A13 , B1 = t , B2 = t
θ + 273.16 θ + 273.16 θ + 273.16
A0 , A1 , A2 ,..., A15 are material constants. The unit of
temperature should be Celsius.
a φσ
φ
e C = 1 − exp ( −a1φ ) 2 + M φσ φ ,
E0
M = a3 exp ( a4θ + a5 )
ref. C.E. Pugh, J.M. Corum, K.C. Liu and W.L. Greenstreet,
"Currently Recommended Constitutive Equations for
Inelastic Design of FFTF Components," Report No. TM-
3602, Oak Ridge National Laboratory, Oak Ridge,
Tennessee, 1972.
1
3 2
t +∆t
σ = t +∆t sij t +∆t sij
2
1
2 2
t
e = ( t eijC − eijorig )( t eijC − eijorig )
C
3
3a) For creep law 1 (and also for creep law 3 with constant
coefficients when the constants correspond to creep law 1),
calculate the effective creep strain and effective creep strain rate
at time t + ∆t using
( eC ) = (teC ) + ( a0 t +∆tσ a1 )
1/ a2 1/ a2 1/ a2
t +∆t
∆t ,
t +∆t
t +∆t e C −t e C
eC =
∆t
3b) For other creep laws (including creep law 3 with variable
coefficients), calculate the pseudo-time t satisfying
where e C ( t +∆t
σ , t +∆tθ , t ) is the generalized uniaxial creep law
d e C ( t +∆tσ , t +∆tθ , t ) . Then
and e C
( t +∆t
σ, t +∆t
θ, t ) =
dt
calculate the effective creep strain and effective creep strain rate
at time t + ∆t using
t +∆t
e C = e C ( t +∆tσ , t +∆tθ , t ) , t +∆t
e C = e C ( t +∆tσ , t +∆tθ , t ) .
4) Calculate t +∆t
γ using
t +∆t
3 eC
t +∆t
γ=
2 t +∆t
σ
∆eijC = ∆t t +∆t
γ t +∆t
sij
TF
in which e C
u ,r is the uniaxial creep strain at failure, t e C is the
effective creep strain and TF is the triaxiality factor.
t
σm
You can choose to use the triaxiality factor TF = (the
t
σ
default) or to use no triaxiality factor (that is, TF = 1 ).
t
σ m = tσ 11 + tσ 22 + tσ 33 is the volumetric stress.
e cu,r(1)1 e cu,r
Effective
stress
s (11) s (12) s (13) s (4)
1
q
q2
You can choose to apply the rupture model only under tensile
conditions ( tσ m > 0 ) or under both tensile and compressive
conditions. The default is to apply the rupture model only under
tensile conditions.
Rupture can be allowed only in tension, based on the sign of
t
σm .
• You can apply a time offset to any of the creep material models
(for example, creep law 1 becomes t e C = a0 tσ a 1 ( t − t0 ) 2 , where
a
t
eC = ∑
all solution steps
∆e C
2 C
where ∆e C = ∆e ⋅ ∆eC and ∆eC is the tensor of creep strain
3
increments in a solution step. Because of the summation over the
solution steps, we refer to the calculated value of t e C as the
accumulated effective creep strain.
ref. KJB • The stresses at the integration points are evaluated using the
Section 6.6.3 effective-stress-function algorithm.
t +∆t
σ (i ) = t +∆t C E ( t +∆t e(i ) − t +∆t e P (i ) − t +∆t eC (i ) − t +∆t eTH ) (3.6-1)
is solved separately for the mean stress and for the deviatoric
stresses. In this equation the index (i) denotes the iteration counter
in the iteration for nodal point equilibrium. For easier writing this
index will be dropped in the discussion to follow. The mean stress
is calculated as
t +∆t
E
t +∆t
σm =
1− 2 t +∆t
ν
( t +∆t
em − t +∆t eTH ) (3.6-2)
1
t +∆t
s= t +∆t e′′ − (1 − α ) ∆t τ γ t s (3.6-3)
t +∆t
aE + α ∆t τ γ + ∆λ
t +∆t
t +∆t E t
where aE = , s = deviatoric stress at the start of the
t +∆t
1+ ν
time step and α is the integration parameter used for stress
evaluation ( 0 ≤ α < 1) The creep and plastic multipliers τ γ and
∆λ are functions of the effective stress t +∆t
σ only, and they
account for creep and plasticity; also
t +∆t
e′′ = t +∆t
e′ − t e P − t eC
The zero of equation (3.6-4) provides the solution for the effective
stress t +∆tσ , where
a = t +∆t aE + α ∆t τ γ + ∆λ
e = t e P + ∆λ t +∆t s
t +∆t P
e = t eC + (1 − α ) t s + α
t +∆t C t +∆t
s ∆t τ γ
n
ε W (θ , φ ) = ∑ ∆ε W (θ i , φi −1 , φi − φi −1 )
i =1
where
∆ε W (θi , φ , ∆φ ) = ε W (θ i , φ + ∆φ ) − ε W (θ i , φ )
φ0 φ1 φ2 φ3
(a)
θ3
• The concrete model can be employed with the 2-D solid and
3-D solid elements.
no temperature effects;
t
σ = E0 t e (3.7-1)
E0 t e
t
σ Es ec
=
σc te te
2
te
3
1+ A + B + C
ec ec ec
~
~
t
~p
e~u
t
e~c
~
et ~
e
~
u
~
c
and hence
te
2
te
3
E0 1 − B − 2C
ec ec
t
E= 2
(3.7-2)
te te
2
te
3
1 + A + B + C
ec ec ec
where
E0 2 E0
E + ( p − 2 p ) E − ( 2 p − 3 p + 1)
3 3 2
A= u s
( p 2 − 2 p + 1) p
E
B = 2 0 − 3 − 2 A
Es
E
C = 2 − 0 + A
Es
σc eu σ
and the parameters E0 , σ c , ec , Es = , σ u , eu , p = , Eu = u
ec ec eu
are obtained from uniaxial tests.
The stress-strain relation in Eq. (3.7-2) assumes monotonic
loading conditions. For unloading conditions and loading back to
the stress state from which unloading occurred, the initial Young's
modulus E0 is used. For strain states beyond eu in compression, it
is assumed that stresses are linearly released to zero, using the
following modulus:
σu −σc
Eu =
eu − ec
t
g < gmax (3.7-4)
where gmax is the maximum value of the loading scalar that has
been reached during the complete solution.
During unloading, the material is assumed to be isotropic and
the initial Young's modulus, E0 , is used to form the incremental
stress-strain matrix, both for stiffness and stress calculations.
In loading conditions, the principal stresses tσ pi (with
t
σ p1 ≥ tσ p 2 ≥ tσ p 3 ) are calculated. For each principal stress
direction, a tangent Young's modulus t E pi corresponding to the
given strain state t e pi is evaluated using Eqs. (3.7-1) and (3.7-2).
In Eq. (3.7-2), the strains t e pi are used together with the
parameters σ c′ , σ u′ , ec′ and eu′ defined in Eq. (3.7-8) which
accounts for the multiaxial stress conditions.
The material is considered as orthotropic with the directions of
orthotropy being defined by the principal stress directions. Once
cracking occurs in any direction i, that direction is fixed from that
point onward in calculating tσ pi .
The stress-strain matrix corresponding to these directions is, for
three-dimensional stress conditions,
1
C= ×
(1 + ν )(1 − 2ν )
(1 − ν ) t E p1 ν t E12 ν t E13 0 0 0
(1 − ν ) t E p 2ν E23 t
0 0 0
(1 − ν ) E p 3
t
0 0 0
0.5 (1 − 2ν ) t E12 0 0
symmetric 0.5 (1 − 2ν ) E13
t
0
0.5 (1 − 2ν ) t E23
(3.7-5)
t
where v is the constant Poisson's ratio and the Eij with i ≠ j are
evaluated using
t
σ pi t E pi + tσ pj t E pj
t
Eij = (3.7-6)
t
σ pi + tσ pj
ˆ
σ = Ce (3.7-7)
where the stress-strain matrix Ĉ is the one defined in Eq. (3.7-5)
but using the Young's moduli τ E pi . These moduli are evaluated
using three point Gaussian integration of E pi between t e pi and
t +∆t t +∆t
e pi , with t e pi and e pi the strain components measured in
the directions of the respective principal stresses tσ pi and t +∆t
σ pi .
σ c′
γ1 =
σc
and
t
p3
~ ,0)
(0,c
t
p2
t
p1
~
~' =
t t t
~ = ~ (1-0.75
p2
t t ~ )
c
c
(1-0.75 tp3 )
=
t t
~c
t
p2
~
c
t
p21
~
c
ADINA
Experimental
t
p3
~ tp2 t
c p3
tp2 t
p3
t tp1
p2
t i
p1
Curve i ~
t = constant
c
Input to ADINA
Curve 1
Experimental
t
p2
~
c
t
p3 t =t p3
~ p2
c t
p2 = p3
t
~
c
t
1 ~ p1
c ~ = constant
c
1 is determined from
failure envelope by
holding tp1 and t p2
constant.
t t
p2 p2
~ = constant
~
c c
~ from given (t , t )
(a) Determination of c p1 p2
~
t
~ /
~ =
t e~
c c 1
u/
~ ~ =
u 1
~e / e~ = C1
12+ C2
1
c c
~e / e~ = C1
12+ C2
1
u u
~
t
u
~
t
c
~
t
u
~
t
c
E0ηn 0 0 0 0 0
1 E p 2 ν E23
t t
0 0 0
(1 − ν 2 ) t
E p 3
0 0 0
C=
E0η s
0 0
2 (1 + ν )
E0η s
symmetric 0
2 (1 + ν )
t
E23
2 (1 + ν )
(3.7-9)
For the tensile stress normal to the tensile failure plane and
the shear stresses in this plane, we use the total strains to
calculate the total stresses with
Ef 0 0
ˆ
C= G12f 0 (3.7-10a)
sym. G13f
~
t
~
tp
t
~
E0
1 Ef Strain normal
1 to tensile
failure plane
l
~
et te e~m (= ~et) ~
e
f
~
~ E0
G ijf = f
1.0 2(1+
)
s Strain normal
to tensile
failure plane
~
et e~m ~
e
h2
~
2E G
~ 20 f with Gf energy released per unit area in
t h2 direction perpendicular to h direction
2
τ
E p 2 ν τ E23 0
1 τ
C= 2
E p3 0 (3.7-10b)
1− ν τ
sym. E23
(1 − ν ) 2 (1 +ν )
2
Stress
1
l
l
2
3l 4
5l
l Strain
Stress
A
12 Strain
11 l l
Point numbers:
2-5,11-12 crack open l
l
8
l
20
σ −σ
E=
t t
τ e p3 + e p3 e p3
e p3
where tep3 and ep3 are the strain component and incremental strain
components at time t measured in the direction of the principal
stress tσp3. To obtain the stress increment, Eq. (3.7-7) is used where
the matrix Ĉ corresponds to isotropic material conditions with
Young's modulus τE. Note that when tep3 becomes equal to or less
than eu′ , the stresses are linearly released to zero.
σ p3
t
ν s = ν when γ 2 = ≤ γa
σ c′
2
γ2 −γa
ν s = ν f − (ν f −ν ) 1 − 1 − γ when γ 2 > γ a
a
• Recall that any stress state is a point in the stress space that can
be represented by the principal stresses (σ 1 , σ 2 , σ 3 ) or,
equivalently, by the stress invariants (I1, J2,, J3). The latter is
composed of
1
J2 = sij sij second deviatoric stress invariant, and
2
where σ ij and sij are the stress and deviatoric stress tensors
expressed in index notation, respectively, and σ m is the mean
stress. Note that these stress quantities are related via
σ ij = sij + σ mδ ij
1 3 3J
θ = arccos 3
2 ( J )2
3
3
2
1 2
σ o = I1 = σ m and τo = J2
3 3
σ o = 3K s ( ε o − ε TH ) − σ id
and
τ o = 2Gsγ o ,
Kc −σ o
for ≤2
−σ fc
1 + A o
fc
Ks =
Kc −σ o
−1
for >2
σo fc
1 + 2 Ab + 2 ( b − 1) + A
b −1 b
fc
and
Gc
Gs = d −1
,
τ
1+ C o
fc
respectively, and
n
k fc τo
σ id = m .
−σ o f c
1+ l
fc
K e = 11000 + 3.2 f c2 ,
4
k= ,
1 + 1.087 ( f c − 15 )
0.23
σ o = 3K e ( ε o − ε TH ) and τ o = 2Geγ o .
• The AUI command that defines the data fitted concrete material
model in ADINA is MATERIAL DF-CONCRETE. In order to
adapt any model to the above constitutive equations, this command
provides the parameter FC, which defines the uniaxial cylinder
compressive strength of the user's concrete, and the parameter
CONFAC, which converts FC from the user's stress unit to N/mm2
such that 16 N/mm2 ≤ FC x CONFAC} ≤ 65 N/mm2.
When temperature effects are desired, the parameter ALPHA,
which defines the mean coefficient of thermal expansion, must be
specified.
0.724
0.944 f c σo
g =τo − TP − − SP f c = 0,
r fc
4 (1 − e 2 ) cos 2 θ + ( 2e − 1)
2
r= ,
2 (1 − e 2 ) cos θ + ( 2e − 1) 4 (1 − e 2 ) cos 2 θ + 5e 2 − 4e
0.133
σ
e = 0.670551 TP − o ,
fo
occur such that they stay valid for the remaining of the simulation
and where the crack planes are not necessarily orthogonal to each
other. The modeling of a crack starts after cracking is detected by
the failure surface function and hinges on the idea that concrete
retains its strength only on the defined crack planes, which
indicates a brittle type of cracking. Consequently, the stress-strain
relations for cracking are understood the best when they are
presented in the local coordinate system (i.e., the coordinate system
of the crack of interest) and always need to be transformed to the
global coordinate system. Following an incremental formulation,
these are
∆σ (i ) = C(i ) ∆ε (i ) ,
STIFAC M 0 0 0 0 0
0 M N 0 0 0
0 N M 0 0 0
C(1) =
0 0 0 SHEFAC G 0 0
0 0 0 0 SHEFAC G 0
0 0 0 0 0 G
where STIFAC and SHEFAC are the normal and shear stiffness
reduction factors, respectively, such that 0.0 < STIFAC < 1.0 and
0.0 < SHEFAC < 1.0, and
4 2
M = K t + Gt , N = K t − Gt and G = Gt ,
3 3
where Kt and Gt are the tangent bulk and shear moduli prior to
failure, respectively. When the state prior to failure is considered
to be a nonlinear compressive state, these moduli are defined as
Ke −σ o
b −1
for ≤2
fc
1 + Ab −σ o
Kt =
fc
Ke −σ o
for >2
1 + 2b −1 Ab fc
and
Ge
Gt = d −1
,
τ
1 + Cd o
fc
principal stress and the second axis is aligned with the intersection
of the two crack planes. After failure, the stresses along the first
and third axes are set to zero (see Fig. 3.7-11) and the
corresponding constitutive matrix is defined as
STIFAC M 0 0 0 0 0
0 M 0 0 0 0
0 0 STIFAC M 0 0 0
C(2) =
0 0 0 SHEFAC G 0 0
0 0 0 0 SHEFAC G 0
0 0 0 0 0 SHEFAC G
C(3) =
STIFAC M 0 0 0 0 0
0 STIFAC M 0 0 0 0
0 0 STIFAC M 0 0 0
0 0 0 SHEFAC G 0 0
0 0 0 0 SHEFAC G 0
0 0 0 0 0 SHEFAC G
• Note that both STIFAC and SHEFAC can be user input and are
parameters of the AUI command MATERIAL DF-CONCRETE.
The values assigned to these factors are constant for the entire
duration of the simulation and must be chosen carefully by the
user.
(STRESS-(XYZ));
temperature (ELEMENT_TEMPERATURE);
OPEN_CRACKS;
CLOSED_CRACKS;
CRUSHED;
CRACKS;
CRACK_STRESS; and
CRACK_STRAIN.
Processing.
• These material models can be employed with the 2-D solid, 3-D
solid and 3D-shell elements.
no temperature dependence
Thermal
Viscoelastic Mullins Orthotropic
strain
effects1 effects1 effects
effects
No
temperature yes yes yes no
dependence
TRS
temperature yes yes yes yes
dependence
Full
temperature yes yes yes yes
dependence
Mooney-Rivlin
Ogden
Arruda-Boyce
hyper-foam
Sussman-Bathe
used. For more information, refer to ref KJB, section 6.6.2. Here
and below, we omit the usual left superscripts and subscripts for
ease of writing. Unless otherwise stated, all quantities are
evaluated at time t and referred to reference time 0 .
Cij = 2ε ij + δ ij (3.8-1)
1 2
I1 = Ckk , I2 =
2
( I1 − Cij Cij ) , I 3 = det C (3.8-2a,b,c)
the stretches λi where the λi ’s are the square roots of the principal
stretches of the Cauchy-Green deformation tensor; and the reduced
stretches
1
λi* = λi ( λ1λ2 λ3 )
−
3 (3.8-4)
Note that
J 3 = λ1λ2 λ3 (3.8-5)
1 ∂W ∂W
Sij = + (3.8-6)
2 ∂εij ∂ε ji
1 ∂Sij ∂Sij
Cijrs = + (3.8-7)
2 ∂ε rs ∂ε sr
WD = C1 ( I1 − 3) + C2 ( I 2 − 3) + C3 ( I1 − 3) + C4 ( I1 − 3)( I 2 − 3) +
2
C5 ( I 2 − 3) + C6 ( I1 − 3) + C7 ( I1 − 3) ( I 2 − 3) +
2 3 2
(
C8 ( I1 − 3)( I 2 − 3) + C9 ( I 2 − 3) + D1 exp ( D2 ( I1 − 3) ) − 1
2 3
)
(3.8-8)
1
WV = κ ( J 3 − 1)
2
(3.8-9)
2
p = −κ ( J 3 − 1) (3.8-10)
G = 2 ( C1 + C2 ) + D1 D2 (3.8-11)
E = 6 ( C1 + C2 ) + D1 D2 (3.8-12)
E
κ= with ν = 0.499 (3.8-13)
3 (1 − 2ν )
2G (1 + ν )
κ= = 500G for ν = 0.499 (3.8-14)
3 (1 − 2ν )
This rule of thumb can be used to estimate the bulk modulus in the
absence of experimental data. However, you can use lower values
of the bulk modulus to model compressible materials.
9
µ
WD = ∑ n λ1α n + λ2α n + λ3α n − 3 (3.8-15)
n =1 α n
1 1
WV = κ ( λ1λ2 λ3 − 1) = κ ( J 3 − 1)
2 2
(3.8-16)
2 2
1 9
G= ∑ µnα n
2 n =1
(3.8-17)
3 9
E = ∑ µnα n (3.8-18)
2 n =1
5
C
WD = µ ∑ 2ii− 2 ( I1i − 3i ) (3.8-19)
i =1 λm
1 1 11 19
in which C1 = , C2 = , C3 = , C4 = ,
2 20 1050 7000
519
C5 = and in which there are two material constants: µ
673750
and λm .
κ ( J 3 − 1)
2
WV = − ln J 3 (3.8-20)
2 2
κ 1
p=− J3 − (3.8-21)
2 J3
N
µn αn 1
W =∑ αn αn
λ1 + λ2 + λ3 − 3 + (
J 3−α n β n − 1 ) (3.8-22)
n =1 α n βn
N
µn α
WD = ∑ λ1 + λ2α + λ3α − 3 J 3α / 3
n n n n
(3.8-23)
n =1 α n
N
µn 1
WV = ∑ ( αn / 3
3 J 3 − 1 + ) (
J 3−α n βn − 1 ) (3.8-24)
n =1 α n βn
1 ∂W ∂WD
SijD = D+ = 0,
εij = 0 2 ∂εij ∂ε ji
εij = 0
1 ∂W ∂W
SijV = V + V =0
εij = 0 2 ∂εij ∂ε ji
εij = 0
1 9
G= ∑ µnα n
2 n =1
(3.8-25)
9
1
κ = ∑ β n + µnα n (3.8-26)
n =1 3
ν
β= (3.8-27)
1 − 2ν
Theoretical background:
∂WD
τi = + p = w′(ei ) + p (3.8-29)
∂ei
where w′(ei ) ≡ dw / de .
(( − )
∞
w′(e) = ∑τ 2)
1 k
e (3.8-31)
k =0
Original
Deformed
L2
F1
L3
L1
0
A1 L2 L3 0 A1
L1
L1
L1 1
e2 e3 e1
e e1 ln
L1
0 2
F1
1 2 3 0
A1
Figure 3.8-1: Uniaxial tension/compression test
τ (e) = ET e, e > 0
= EC e, e < 0
where ET and EC are constants greater than zero, has w′′(e) > 0
only if
1
2 ET < EC < 2 ET
5) Given only simple tension data for τ (e) , there are multiple
w′(e) that exactly correspond to τ (e) , for positive e only. Two
such w′(e) are
w′(e) = 0, e < 0
= τ (e), e > 0
and
1
modulus G is G = E . The results are
3
3 1
E= w′′(0) , G = w′′(0) (3.8-32a, b)
2 2
2) The range of the cubic spline is between the first and last user-
input data points.
3) Outside the range of the cubic spline, the slope of τ (e) is greater
than zero. This ensures that the asymptotic conditions of
τ (−∞) = −∞ , τ (∞) = ∞ are met.
(e)
emin,
emax, e
w’(e)
emin,w’
emax,w’ e
Slope > 0
τ ( e ) − τ ( e)
r = max (3.8-34)
e τ (e )
w′(e) = Ai +1 z + Bi +1 z 3 + Ai (1 − z ) + Bi (1 − z )3 (3.8-35)
z2 z4 1 − (1 − z ) 2 1 − (1 − z ) 4
w(e) = Ci + (ei +1 − ei ) Ai +1 + Bi +1 + Ai + Bi
2 4 2 4
(3.8-36)
The AUI determines the constants Ai, Bi, Ci from uniaxial stress-
strain data, as described above. The constants Ai, Bi, Ci are listed
in the ADINA output file when the model definition is listed.
WV = κ ( J 3 ln J 3 − ( J 3 − 1) ) (3.8-37)
p = −κ ln J 3 = −κ ( e1 + e2 + e3 ) (3.8-38)
2) The data set should contain both tension and compression data
(compression data is possibly converted from equibiaxial tension
data, see below). If the data set contains only tension data, the
program will assume that the true stress / true strain curve in
compression is a straight line, which is most likely not a good
assumption.
3) The stresses and strains in the set of stress-strain data points can
be either
In the latter cases, the program internally converts the stresses and
strains to true stresses and logarithmic strains.
eu = −2eb , λu = λb−2 , 0 eu = (1 + 0 eb ) − 1
−2
(3.8-39)
τ u = −τ b , 0 σ u = − 0 σ b λb3
7) The strain range of the data set should contain the range of
strains anticipated during the analysis.
• You can have the AUI determine the material constants for the
Mooney-Rivlin, Ogden, Arruda-Boyce or hyper-foam materials by
a curve-fit to experimental stress-strain data. Note that the
Sussman-Bathe model does not require curve-fitting.
= (L+)/L e=/L
Y
1 = ; 2 = 3 = -1/2
L
L
0p1 = F/A
X
Deformed area A/
Z Original area A
Y = (L+)/L e=/L
L
0p1
L = F/A
X
Deformed area A/
Z Original area A
1 = ; 2 = 1; 3 = 1/
= (L+)/L e=/L
Y 1 = ; 2 = ; 3 = -2
F Original area A
Deformed area A/
0p1 = F/A
Z L+
L
∂W
σ p1 = = 0σ p1 ( Ci ,λ ) (i = 1,..., M ) (3.8-40)
∂λ
0
Notice that this is the engineering principal stress, that is, the force
divided by the original area of the specimen. Thus, we have an
expression for the principal stress that is linear in terms of the
material constants Ci, and dependent on the principal stretch
determined from the allowed experimental cases. Hence, given a
set of data points ( λk , σ k ) from one or more of the test cases, we
can establish the least squares fit by a curve of the form given by
Eq. (3.8-40).
Consider the sum of discrete error terms at the given data points
of one of the experimental test cases (simple tension, pure shear or
equibiaxial tension)
N
S = ∑ (σ k − σˆ p1 ( λk ) )
2
(3.8-41)
k =1
where σ
ˆ p1 is the approximating function dependent on the
constants Ci (i = 1,...,M) given by Eq. (3.8-40). We can expand
this sum to include data from all three test cases, i.e.:
N 2
j
( )
3
S = ∑ ∑ σ kj − σˆ1j ( λ kj ) (3.8-42)
j =1 k =1
Ogden model: The strain energy density function is, for total
incompressibility, given by Eq. (3.8-15). Thus, for an M-term
Ogden model, there are 2 M constants, i.e., the pairs ( µi , α i ) for i
= 1,...,M.
Assuming the principal stretches for the three test cases shown
in Figure 3.8-4, the following form for the principal stress 0 σ p1
can be derived:
0 σ p1 = ∑ µi ( λ α −1 − λ cα −1 )
i i
(3.8-43)
i =1
with c = -0.5 for simple tension, -1.0 for pure shear, -2.0 for
equibiaxial tension.
Equation (3.8-43) is linear in terms of the coefficients µi , but
nonlinear in terms of the exponents α i . A nonlinear least squares
algorithm could be employed to iterate to a solution for the
Uniaxial tension:
5
iCi
0 σ p1 = 2 µ ( λ − λ )∑ I1i −1 , I1 = λ 2 + 2λ −1
−2
(3.8-44)
i =1 λm2i − 2
Pure shear:
5
iCi
0 σ p1 = 2 µ ( λ − λ ) ∑ I1i −1 , I1 = λ 2 + λ −2 + 1 (3.8-45)
−3
i =1 λm2i − 2
Equibiaxial tension:
5
iCi
0 σ p1 = 2 µ ( λ − λ ) ∑ I1i −1 , I1 = 2λ 2 + λ −4
−5
(3.8-46)
i =1 λm2i − 2
Uniaxial tension:
β
−
λT = λ 1+ 2 β
(3.8-47)
Pure shear:
β
−
λT = λ 1+ β
(3.8-48)
Equibiaxial tension:
2β
−
λT = λ 1+ β
(3.8-49)
∑ µ (λ )
M
1 −α n β
0 σ p1 =
αn
− J3 (3.8-50)
λ
n
n =1
where J 3 = λλT2 for uniaxial tension, J 3 = λλT for pure shear and
J 3 = λ 2 λT for equibiaxial tension. (3.8-50) is then used as the
basis of a nonlinear curve-fitting algorithm to obtain the constants
( µn , α n ) .
2.0
STRESS
-2.0
b)
-2.0
c)
-2.0
Strain e,
stress
E
E1 1
..
E .
Strain Strain
g
σ = σ ∞ + ∑ qα (3.8-51)
α
1
qα + qα = β α σ ∞ (3.8-52)
τα
t − t′
t
qα = ∫ exp − α β α t ′σ ∞ dt ′ (3.8-53)
0 τ
t − t ′ t′ ′
t
σ = ∫ E ∞ 1 + ∑ β α exp − α
e dt (3.8-54)
0 α τ
1 qα
Dα = qα Γα = qα ( e − g α ) = qα e − α ∞ (3.8-55)
β E
Ψ = Ψ ∞ (e) + ∑ Ψ α ( g α ) (3.8-56)
α
1 ∞ 2
where Ψ ∞ (e) = E e is the strain energy of the elastic chain
2
1
and Ψ α ( g α ) = E α ( g α ) is the strain energy in the spring of
2
2
chain α . In terms of β α , Ψ α ( g α ) = β α Ψ ∞ ( g α ) . The 1D
model is recovered by defining
∂Ψ ∂Ψ
σ= , qα = − (3.8-57a,b)
∂e Γα fixed ∂Γα e fixed
Finite strain model: The finite strain model is derived from the
potential-based 1D model as follows. The elastic potential is
defined as
Ψ ∞ (ε ) = W (ε ) (3.8-58)
( )
where W ε ij is the strain energy density from the elastic part of
the material model. The potential of each chain α is defined as
∂Ψ ∂Ψα ∂Ψ ∂Ψ
Sij = , Qijα = =− =
∂ε ij ∂ε ij ∂Γαij ∂Gijα
Γijα fixed Γαij fixed ε ij fixed
(3.8-60a,b)
where Sij are the 2nd Piola-Kirchhoff stresses and Qijα is analogous
to the stress qα .
Following exactly the same arguments as in the 1D case, we
obtain
1
Qijα + Qijα = β α Sij∞ (3.8-61)
τα
t − t′
t
Qij = ∫ exp − α β α t ′ Sij∞ dt ′
α
(3.8-62)
0 τ
and (3.8-62) can be numerically approximated using
∆t
1 − exp − α
∆t τ t +∆t S ∞ − t S ∞ .
t +∆t α
Qij = exp − α tQijα + β α
τ ∆ t
( ij ij )
τα
(3.8-63)
(3.8-63) is exact when Sijα does not change during the time step,
and is more accurate than the formula given by Holzapfel:
∆t ∆t
t +∆t
Qijα = exp − α tQijα + β α exp − α ( t +∆t Sij∞ − t Sij∞ )
τ 2τ
(3.8-64)
∆t
especially when is large.
τα
where
∂ 2 Ψα
α α
Grsα = Qijα (3.8-66)
∂Gij ∂Grs
∂ 2 Ψα ∂ 2W
= β α
= β α Cijrs . (3.8-67)
∂Gijα ∂Grsα ∂ε ij ∂ε rs
where the tensor Cijrs is evaluated at the strain state Gijα . The
dissipation calculation requires the solution of a set of simultaneous
linear equations of at most order 6 (in the 3D case) at each
integration point.
If usage=deviatoric,
∂ 2 Ψα α ∂ WD
2
α α
= β = β α ( CD )ijrs (3.8-68)
∂Gij ∂Grs ∂ε ij ∂ε rs
where the tensor ( CD )ijrs is evaluated at the strain state Gijα . Here,
the dissipation calculation requires a singular value decomposition
of ( CD )ijrs , since ( CD )ijrs has a zero eigenvalue. A similar
situation applies when usage=volumetric, except that the
corresponding material tensor has only one nonzero eigenvalue.
The procedure given in (3.8-65) to (3.8-68) is only approximate,
since the fundamental assumption Gijα = ε ij − Γαij strictly speaking
only holds for small strain analysis.
dζ 1
= (3.8-69)
dt aT ( tθ )
Mooney-Rivlin, hyper-foam
Ogden, Arruda-Boyce,
Sussman-Bathe
Plane Plane strain, Plane Plane strain,
stress1 axisymmetric, stress3 axisymmetric,
3D2 3D
usage=combined yes no yes yes
usage=deviatoric yes yes no yes
usage=volumetric no no no yes
Evidently
t
1
t
ζ =∫ dt ′ (3.8-70)
a ( t ′θ )
0 T
The shift function used here is either the WLF shift function,
C1 ( tθ − θ ref )
log10 aT ( θ ) = −
t
(3.8-71a)
C2 + tθ − θ ref
1 1 t
log10 aT ( tθ ) = C1 t − , θ ≥ θ ref
θ θ
ref
(3.8-71b)
1 1 t
= C2 t −
θ θ , θ < θ ref
ref
dqα 1 α dσ ∞
+ α q = βα (3.8-72)
dζ τ dζ
1
qα + qα = β α σ ∞ (3.8-73)
aT (θ )τ α
d t ′ Sij∞
t
ζ
ζ −ζ ′
Qij = ∫ exp − α β α
α
dζ ′ (3.8-74)
0 τ dζ
∆ζ
1 − exp − α
∆ζ τ t +∆t S ∞ − t S ∞
t +∆t α
Qij = exp − α t Qijα + β α
τ ∆ζ
( ij ij )
τα
(3.8-75)
t +∆t
1
∆ζ = ∫t
aT ( t ′θ )
dt ′ (3.8-76)
A table with one row for each chain. Each row contains β α ,
τ α , the heat generation factor (fraction of dissipation
considered as heat generation, default value is 0.0), and the
usage flag (default value is deviatoric). There is no restriction
on the number of chains permitted. The usage flag can be
different for each chain.
Force
h
f
e
c
b
d g
a
Deflection
Figure 3.8-7: Mullins effect loading-unloading-reloading curves
Note that any permanent set associated with the Mullins effect
is not included in the Ogden-Roxburgh model used here.
The Ogden-Roxburgh model, as implemented in ADINA, uses
the following strain energy density expression:
1 1
η = 1 − erf (Wm − W ) (3.8-78)
r m
x
2
erf( x ) = ∫ exp ( −u ) du
2
(3.8-79)
π 0
dφ (η )
= −W (3.8-80)
dη
1 W
η m = 1 − erf m (3.8-81)
r m
1 W
φ (η m ) = (1 − η m ) Wm = erf m Wm (3.8-82)
r m
Force
Slope kb
Fb
2 1
Slope k b Fb2
Slope ka % rm
Deflection
1 W
Slope 1 erf m k a
# r ! m "$
Wm = strain energy density at b
m Wm W 1 W 2
E
φ = ∫ 0 σ de = erf m − 1 − exp − m
r m m π m
A
(3.8-83)
where
C
Wm = ∫ 0 σ de (3.8-84)
A
=Force/original area
0
B
Dissipation = shaded area
D
A
E e=displacement/
original length
WO =
k1
2k 2
{
exp k2 ( J 4 − 1) + exp k2 ( J 6 − 1) − 2
2
2
} (3.8-85)
the first term represents the response of the n a fibers and the
second term represents the response of the nb fibers.
This material description is described in the following
reference:
We note:
1 nb
2 b
a
b
4
a
3
na
a) 2D element, hoop option not used
z
1
y
2
a
b
4
a
3
na
b) 2D element, hoop option used,
nb = x
2 1
3 4
7 8 c
nb
a
b
a
na
X = X m Xth (3.8-87)
X m = (1 + eth ) −1 X (3.8-89)
Cm = (1 + eth ) −2 C (3.8-90)
1
ε m = (1 + eth ) −2 ε −
2
(1 − (1 + eth )−2 ) I (3.8-91)
θ2 − θ θ − θ1
W= W1 + W2 (3.8-93)
θ 2 − θ1 θ 2 − θ1
3.8.8.1 Introduction
R3, u3
R2, u2
R1, u1
Equilibrium
Wˆ = W + m(e1 + e2 + e3 )
∂W
δ Wˆ = ∑ + m δ ei + (e1 + e2 + e3 )δ m (3.8-95)
i ∂ei
∂W
+ m = Ri λi , no sum on i , (3.8-97a)
∂ei
e1 + e2 + e3 = 0 (3.8-97b)
3.8.8.2 Stability
∂ 2W ∂ 2W ∂ 2W
2 − R1λ1 1
∂e1 ∂e1∂e2 ∂e1∂e3 ∆e
∂ 2W ∂ 2W 1 λ1∆R1
− R2 λ2 1 ∆e2 λ2 ∆R2
∂e2 ∂e3 = (3.8-98)
2
∂e2
∆e3
λ3 ∆ R3
∂ 2W
− R3λ3 1 ∆m 0
∂e3
2
symm 0
1 ∆e1 ∆r1
KEE 1 ∆e2 ∆r2
= (3.8-99)
1 ∆e3 ∆r3
1 1 1 0 ∆m 0
∆e1 ∆y1
∆e = A ∆y (3.8-100)
2 2
∆e3 ∆y3
2 1
0
3 3
1 2 1 1
where A = − . A is orthogonal with
2 3 2 3
− 1 2 −
1 1
2 3 2 3
2 1 2 1 2
− −
3 2 3 2 3 ∆e
∆
1y 1
∆e 1
∆y = AT ∆e = 1
−
1
2 2 0 ∆e2 (3.8-101)
2 2
∆y3 ∆e3 ∆e
1 1 1 3
3 3 3
Note that ∆y1 and ∆y2 are insensitive to the volume change and ∆y3
is proportional to the volume change ∆e1 + ∆e2 + ∆e3 .
In the new basis, the stability matrix equation has the form
0 ∆y1 ∆s1
KYY 0
∆y2 = ∆s2 (3.8-102)
3 ∆y3 ∆s3
0 0 3 0 ∆m 0
∆s1 ∆r1
where KYY = A KEE A and ∆s2 = AT
T ∆r .
2
∆s3 ∆r3
∂W
In uniaxial tension, R2 = R3 = 0 , so, from equilibrium, m = −
∂e2
∂W ∂W
and R1λ1 = − . Also, at equilibrium . However, there are
∂e1 ∂e2
still three admissible strain increments ∆e1 , ∆e2 , ∆e3 , so the 4x4
matrix stability equations given above ((3.8-93) to (3.8-104)) are
applicable.
∂ 2W ∂ 2W
2 − R1λ1 1
∂e1 ∂e1∂e3 ∆e ∆r
∂ 2W 1 1
∆e1 ∆y1
∆e = A ∆y (3.8-106)
3 2
1 1
2 2
where now A = , so that the matrix stability equation
1 1
− 2 2
(3.8-105) becomes
0 ∆y1 ∆s1
KYY
2 ∆y2 = ∆s2 (3.8-107)
0 2 0 ∆m 0
1
KYY11∆y1 = ∆s1 , ∆y2 = 0 , ∆m = ( ∆s2 − KYY12 ∆y1 ) (3.8-108a,b,c)
2
3.8.8.6 Comments
eii
em = ∑3
i
= incremental mean strain
t
gij = t eij − t emδ ij = deviatoric strains
σ ii
σm = ∑i 3
= incremental mean stress
sij = 2 t G gij
σ m = 3 t K em
K LD
l
t
KLD l
l l l
e1v e2v te
v e3v e4v e5v e6v ev
KUN
l Note: Compressive volume strain
Unloading bulk modulus
is assumed positive
l
t
KUN l
l
l l
G LD t
K t
Note: tG UN = t UN GLD
Loading shear modulus
K LD
l l
t l
GLD
l
l
l
t
• The instantaneous bulk and shear moduli, K and tG, are
t
functions of the loading condition, and the volumetric strain ev is
defined as
t
eν = egrav + ( −3 t em )
where egrav is the volumetric strain (three times the mean strain and
t
taken positive in compression) due to the gravity pressure and em
is the mean strain at time t. Defining emin as the minimum mean
strain ever reached during the solution, we have that the material is
loading if t em ≤ emin and the material is unloading if t em > emin
i.e.,
tK when t em ≤ emin
t
K = t LD
KUN when t em > emin
and
tG when t em ≤ emin
t
G = t LD
GUN when t em > emin
• Note that the loading conditions for both the bulk and the shear
t
moduli are determined by the history of em only. The values of
t
KLD, tKUN, and tGLD are obtained using the curves in Fig. 3.9-1,
t KUN
and the modulus t GUN = t GLD t .
K LD
t +∆t
sij = 2 t G ( t +∆t
gij − t g ij ) + t sij
and
t +∆t
σ ij = 3 t K ( t +∆t
eij − t eij ) + tσ ij
t t
To obtain G and K, we check whether the loading or unloading
conditions are active by comparing the current volumetric strains
and previous volumetric strains. To start the procedure at time 0,
a
x2
a
ab
Maximum a =
x1 in-situ gravity pressure
ea
a
x2
a
ab a failure stress =
in-situ gravity pressure
x1
ea
Tension cut-off
Cap hardening
• The Drucker-Prager model can be used with the 2-D solid and
3-D solid elements.
t
f DP = α t J1 + t
J2D − k
the first stress invariant and the second deviatoric stress invariant,
respectively.
The corresponding potential function is given by
t
g DP = β t J1 + t
J2D − u
tan φ 3c
α= k=
9 + 12 tan φ 2
9 + 12 tan 2 φ
The cap yield function depends on the shape of the cap. For a
plane cap:
t
f C = − t J1 + t J 1a
where t
J 1a is a function of the volumetric plastic strain t eVP :
1 e
t P
t
J 1 = − ln 1 − V + 0 J 1a
a
D W
0
where J 1a is the cap initial position and W and D are material
constants.
'
(((
)
(
a) Drucker-Prager yield surface in principal stress space
Plane cap
& J2D
Yield surface
Tension
cut-off H
1
Elliptical cap
k Elastic region
O A C -J1
T
a
- J1 (plane cap)
a
- J 1 (elliptical cap)
f C = ( t J1 + t L ) + R 2 ( t J 2 D − t B 2 )
t 2
t
where B is the vertical semi-axis of the ellipse (AH), R is the cap
t
ratio (AC/AH) and L is equal to OA. The dependence of 0 J 1a on
the volumetric plastic strain is the same as for the plane cap. In the
case of tension cut-off, T is the maximum value that t J 1a can take.
• We note that:
right in Figure 3.9-3 and is not reached in the analysis, and the
position of the tension cut-off T is moved far to the left and is
also not reached in the analysis, then the Drucker-Prager yield
condition approaches the von Mises elastic-perfectly-plastic
yield condition.
e = t e11P + t e22
t P
V
P
+ t e33P
that is, the constraint that exists between these two quantities.
We note that the plane cap movement corresponds to hardening
t a
J1
Material parameters
W and D are negative.
- _1
D
0 a
J1
Possible volumetric te P
v
plastic strain (negative)
W
a
Figure 3.9-4: Relation between plane cap position tJ 1 and volumetric
plastic strain tePv
• The Cam-clay model can be used with the 2-D solid and 3-D
solid elements.
t
q2 t t
t
F= 2
+ p ( p − t p0 ) = 0
M
3
constant, is the slope of the critical state line and t p0 , called the
pre-consolidation pressure, is the diameter of the ellipsoid at time t
along the axis p. All of the variables are shown in Fig. 3.9-5(a).
q
Critical state line
p=tpcs
Elliptical yield
surface
(tp,tq)
M
1
p=t p0 p
a) Cam-clay yield surface
v
N Isotropic consolidation line
Critical state line
1 Overconsolidation line
1
*
1
(tp,tv)
p=1 p=tpcs p=t p0 ln p
b) Compression behavior
Figure 3.9-5: Cam-clay model
t
p0
t
v = N − λ ln p0 + κ ln
t
t
p
Γ = Ν + (κ − λ ) ln p0
t
t vt p
K=
κ
ref. A.M. Britto and M.J. Gunn, Critical state soil mechanics
via finite elements, Ellis Horwood Limited, pp. 161-184,
1987.
• The Mohr-Coulomb model can be used with the 2-D solid and
3-D solid elements.
t
f MC =sin φ t I1
+
1
2
(
3 (1 − sin φ ) sin tθ + 3 ( 3 + sin φ ) cos tθ ) t
J 2 − 3c cos φ
where
1 3 3 t J3
θ = arc cos
t
,
( 2)
3 2 t J 3/ 2
π
stress plane (see Fig. 3.9-8), and 0 ≤ tθ ≤ . In addition, note that
3
φ is the friction angle (a material parameter) such that
π
0 ≤φ < ;
2
t
J 2 is the second deviatoric stress invariant at time t; and
t
J 3 is the third deviatoric stress invariant at time t.
g MC =α MC I1 +
t
t t
J 2 − kMC
where
2sinψ
α MC = .
3 (1 − sinψ ) sin θ MC + 3 ( 3 + sinψ ) cos θ MC
I1 − T = 0,
Drucker-Prager
criterion
1'
Mohr-Coulomb
criterion
3'
2'
Tension
cut-off
Mohr-Coulomb potential surface
(with temperature effects)
I1
Drucker-Prager yield
surface at a tensile Drucker-Prager yield
1' surface at a compressive
meridian match (+ = 0)
meridian match (+ = %/3)
Mohr-Coulomb
yield surface
3'
'2
Drucker-Prager
p
B , ij
yield surface
Mohr-Coulomb
yield surface
, ijp , ijp
A C
+MC
2 3
t
• In some geotechnical analyses, the thermal strains eTH need to
be included. These strains are taken into account by assuming that
the Mohr-Coulomb material is an isotropic temperature
independent material (see Section 3.1.4) affecting the material
behavior only hydrostatically. The thermal strains are calculated
from the temperature conditions and α the mean coefficient of
thermal expansion, where α ≥ 0 , as given in Eq. (3.1-16).
( ) 3c
2
( 2aMC sinψ )
2
t
g MC = + t r M MC t
J2 + 2sinψ t I1 − ,
tan φ
where
3cε 3c
tan φ if T ≥ tan φ with 0 < ε < 1,
aMC =
3c − T if T < 3c ,
tan φ tan φ
M MC = 3 ( 3 − sin φ )
and
4 (1 − e 2 ) cos 2 tθ + ( 2e − 1)
2
t
r= ,
2 (1 − e ) cos θ + ( 2e − 1) 4 (1 − e ) cos θ + 5e − 4e
2 t 2 2 t 2
π
and is defined for 0 ≤ tθ ≤ . In order for such an elliptic
3
function to be convex and smooth, it is required that 0.5 < e ≤ 1 ,
where e is known as the eccentricity (in the deviatoric plane) and it
will be assumed to be
3 − sin φ
e= ,
3 + sin φ
π
φ is the friction angle such that 0 < φ < ;
2
aMC is the distance along the hydrostatic axis from the apex
of the Mohr-Coulomb yield surface to the surface described
by the potential function;
• Fabric materials are typically very thin and flexible. They can
support only in-plane tensile stresses, and will wrinkle under
compressive stresses.
t
σ ′p1 = A′ t ε p1
α A 0 0
0 ηA 0
0 0 η A
dG (τ )
t
sij (t ) = 2G (0)eij (t ) + 2 ∫ eij (t − τ ) dτ (3.11-1)
0+
dτ
dK (τ )
t
σ kk (t ) = 3K (0)ε kk (t ) + 3 ∫ ε kk (t − τ ) dτ (3.11-2)
0 + dτ
1
where t is the time, sij = σ ij − δ ijσ kk is the deviatoric stress, δ ij
3
1
is the Kronecker delta, σ ij is the stress, eij = ε ij − δ ij ε kk is the
3
deviatoric strain, ε ij is the strain, G(t) is the shear modulus and
K(t) is the bulk modulus.
In the presence of a temperature variation θ ( t ) , the stresses for
an isotropic and thermorheologically linear viscoelastic material
may be written as
ξ
dG (ζ )
sij (t ) = 2G (0)eij (t ) + 2 ∫ eij (ξ − ζ ) dζ (3.11-3)
0 + dζ
ξ ξ
dK (ζ ) dK (ζ )
σ kk (t ) = 3K (0) ( ε kk (t ) − ε (t ) ) + 3 ∫ [ε kk (ξ − ζ )]
th
dζ − 3ε kkth (t ) ∫ dζ
dζ dζ
kk
0+ 0+
(3.11-4)
where
τ t
ζ = ∫ ψ θ (η ) dη , ξ = ∫ ψ θ (τ ) dτ (3.11-5)
+ +
0 0
dψ
ψ (θTALPHA ) = 1, ψ (θ ) > 0, >0 (3.11-7)
dθ
ηK
K (t ) = K ∞ + ∑ K i e −γ it (3.11-9)
i =1
θ (t ) ≠ 0 (3.11-10)
α = α (θ (t ) ) (3.11-11)
C1 (θ − θ ref )
log10 ψ (θ ) = − (3.11-12)
C2 + (θ − θ ref )
or the Arrhenius shift function
1 1
log10 ψ (θ ) = C1 − , θ ≥ θ ref
θ θ
ref
(3.11-13)
1 1
= C2 −
θ θ , θ < θ ref
ref
• You can either enter the Prony series constants in the AUI, or
you can enter the measured data for the shear and bulk modulus
and let the AUI construct the Prony series. In the latter case, the
AUI constructs the Prony series using a least-squares technique, see
the reference by Frutiger and Woo given below.
• The types of problems for which the porous media model can be
employed are:
σ ij , j + f i = 0 (3.12-1)
σ ij = C e ε kle + Pf δ ij ,
ijkl
i, j , k , l = 1, 2,3 (3.12-2)
∂ε ii n ∂ Pf
∇ ⋅ ( k ⋅∇Pf ) = + (3.12-3)
∂t κ f ∂t
Transient statics:
t +∆t R u
= t +∆t (3.12-4)
R pf
Transient dynamics:
(3.12-5)
K uu = ∫ BTu DB u dV
V
K u pf = ∫ BTu I H pf dV
V
1
p p =
K (1) ∫nH
T
H pf dV
κf
f f pf
V
p p = ∫ B p k B p dV
K (2) T
f f f f
V
∫ ( H ) f dS
T
R u = ∫ HTu f dV +
S
u f
V Sf
∫( ) q dS
S
T
Rp = Hp q
f f
Sq
model:
t +∆t K uu t +∆t
K u pf t +∆t U
t +∆t T t +∆t t +∆t
K u pf K (1)
p f pf − ∆t t +∆t K (2) Pf
pf pf
t +∆t
Ru
= t +∆t t +∆t t +∆t
−∆t R pf + K u pf U + K pf pf Pf
T t (1) t
(3.12-6)
t +∆t
R ud
= t +∆t t +∆t t +∆t
−∆t R pf + U+
T t (1) t
K u pf K pf p f Pf
(3.12-7)
in which
t +∆t
R ud =t +∆t R u + t +∆t
M ( a0 t U + a2 t U + a3 t U ) +
t +∆t
C ( a1 t U + a4 t U + a5 t U )
(3.12-8)
and where a0, a1, ... , a5 are the integration constants for the
Newmark method (see Ref. KJB, Section 9.2.4).
t +∆t t +∆t
components of p f p f − ∆t
K (1) K (2)
pf pf in equations (3.12-6)
Gasket in-plane
directions
• The gasket model can be used with 2-D solid and 3-D solid
elements. It can also be used with small displacement/small
strain, large displacement/small strain formulations.
• In the full gasket, elastic material properties are used for the in-
plane deformations. Therefore the full gasket element has stiffness
for all translational degrees of freedom.
User input
The user input for the gasket material is now described in terms of
the command-line interface.
lcurvei
Unloading curves
Leakage
pressure
Closure strain
Etensile
The point number of the first yield point on the yield curve is given
by parameter NPOINTS. NPOINTS must be greater than or equal
to 3.
For each point on the main loading curve higher than the first yield
point, there is an associated unloading curve, as shown. Notice that
each unloading curve has the same number of points, and this
number of points is equal to NPOINTS.
5 4
3
3
4
2
2
3
1 2 1 1
Closure strain
Etensile
Interpolated unloading
curve based on highest M
closure strain I
Initial yield
Gasket pressure
point
G
E L
D
H
J
B
Leakage F
pressure A
N
C K Closure strain
Modeling issues
• If the user does not explicitly set the material axes directions in
gasket elements, the AUI attempts to automatically determine the
gasket material axes directions. This automatic determination is
made based on the element layout. Fig 3.13-5 shows element
layouts in which the AUI can and cannot automatically determine
the gasket material directions. If the AUI cannot automatically
determine the gasket material directions, and if the material axes
for the gasket are not explicitly defined by the user, then the AUI
cancels dat file generation.
(b) At least one element is surrounded by neighboring elements on opposite sides only.
Gasket direction cannot be automatically determined.
MISC-OPTIONS GASKET-AUTOMATIC=NO
material axes directions. And, if simple gaskets are used, the user
must explicitly eliminate the degrees of freedom corresponding to
in-plane deformations.
See Section 13.1.1 for the definitions of these variables. For the 2D
gasket type, the components represented by the * are AA, BC and
CC, and for 3D gasket type, the components represented by the *
are BB, CC, AB, AC, BC.
Detwinned
martensite CM CM CA CA
Austenite
Twinned
martensite
Mf Ms As Af
Temperature, +
1.0
0.8
0.6
0.4
0.2
0.0
Ms Ms Af Af
Temperature, +
Figure 3.14-2 : Volume fraction of martensite vs. temperature
+ > Af + < As
(a) , (b) ,
ε = ε e +ε t +εθ
where
ε e= elastic strain
ε θ = thermal strain
ε t = transformation strain; to be evaluated
The one-dimensional macro-scale model,
ξ = ξ s + ξt ; 0 ≤ ξ ≤1
ξ + ξΑ = 1
ε t = ε tmax ξs
∆ε ijt = ∆ξ sε max
t
nijt
3 sij
nijt = ; for the martensitic transformation
2σ
3 ε ij
t
n =
t
; for the reverse transformation
2 ε t
ij
where
3
σ= sij sij is the effective von Mises stress
2
3 t t
εt = ε ij ε ij is the effective transformation strain
2
t +∆t
E (ξ )
t +∆t
sij =
1+ t +∆t
ν (ξ )
( t +∆t
ε ij′′ − ∆ε ijt )
where
t +∆t
ε ij′′ = t +∆t ε ij′ − t ε ijt
f M S = 3J 2 − CM (θ − M S )
f M f = 3 J 2 − CM (θ − M f )
f AS = 3J 2 − C A (θ − AS )
f Af = 3J 2 − C A (θ − Af )
Rξ
∆ξ = ∆f f
ff
t +∆t
3 J2 − t J2
∆f = − c( t +∆tθ − tθ )
2 σ
t +∆t
f f = - f M f , c = CM and Rξ = 1 - ξ
f f = f A f , c = C A and Rξ = ξ
f R = 3 J 2 − CRθ − σ R
where
ξs = ξ
ξs
ξs = ξ
ξ
ξt
ξt = ξ
ξ
E ( tξ ) t +∆t
t +∆t TR
s = ( ε ij′′ )
1 +ν ( tξ )
ij
3 t +∆t t +∆t
g 2 (∆ξ s ) = sij sij − CR t +∆tθ − σ R = 0
2
(3.14-1)
E (∆ξ ) t +∆t "
t +∆t
t +∆t
sij = ( ε ij − ∆ξ s ⋅ ε max
t t +∆t t
nij )
1 + ν (∆ξ )
t +∆t
t +∆t
Rξ 3 t +∆t
J 2 (∆ξ ) − t J 2
g (∆ξ ) = ∆ξ − − c( t +∆tθ − tθ ) = 0
t +∆t
ff 2 σ
t +∆t
(3.14-2)
where
Rξ = 1 − ξ , f f = − f M f and c = CM
Rξ = ξ , f f = f Af and c = CA
• The material model can be used with the 2-D plane strain, 2-D
axisymmetric and 3-D solid elements.
1/ m
Q σ
ε VP
= A1 exp − sinh ξ (3.15-1)
Rθ s
A2
s s
s = h0 1 − * sgn 1 − ε VP (3.15-2)
s s*
n
ε VP Q
with s* = sˆ exp (3.15-3)
A1 Rθ
material constants:
E = 20 ⋅103 MPa
υ = 0.31
A1 = 2.3 ⋅107 s −1
Q / R = 11262 K
ξ = 11
m = 0.303
sˆ = 80.79MPa
n = 0.0212
h0 = 4121.31MPa
A2 = 1.38
s0 = 42.32MPa
Stress (MPa)
120
Deformation resistance, s
100
80
40
20
0
0 0.2 0.4 0.6 0.8 1.0
Strain
This call is for the calculation of the element stresses at time t + ∆t.
It is made during the solution of the governing equilibrium
equations, and in the stress calculation phase after convergence is
reached. In addition to the stress calculation, the integer and real
history arrays IARRAY and ARRAY must also be updated when
KEY = 2 (if they are used).
The stress integration can be performed in a single step using
the total increment of strains and displacements or in subincrements
of time. In this case, each subincrement spans a time of ∆t/n where
n is the number of subincrements. This subincrementation may be
useful for incremental rate type formulations such as
viscoplasticity.
The actual stress integration adopted by the user within the
increment or subincrement can be implicit or explicit. Details of the
stress integration procedure are provided in Section 3.16.3.
If temperature-dependent material properties are present, the
variables ALFA, ALFAA, CTD(*), CTDD(*) containing the
properties at times t and t + ∆t , are available for the user-supplied
subroutine.
Note that if the ULH formulation is employed or if mixed-
interpolated displacement/pressure elements are used, then the
STRESS array will be internally modified by ADINA immediately
after this stage to reflect the correct stresses at the element level.
KEY=4: Stress printout call. Called for stress printout to the .out
file.
This call is for the printing of the element response in the stress
calculation phase. No stress calculations or updates should be
performed when KEY=4.
Note that if the TL formulation is employed, ADINA
automatically transforms, just for printout, the stresses used in the
constitutive relations (see Table 3.16-1) to the Cauchy stresses in
the global coordinate system before calling the user-supplied
subroutine with KEY = 4.
In all cases of 1-D elements, the cross-sectional area of the
element is assumed to remain unchanged, and the strain is equal to
the longitudinal displacement divided by the original length.
If the ULH formulation is employed, total stretches PRST(3) are
calculated and passed to the user-coded material subroutine.
• The contents of the real and integer working array are written to
the porthole file for post-processing at every integration point (the
first LGTH3 entries for the real array, and the first LGTH4 entries
for the integer array) . They can be accessed via the
USER_VARIABLE_I and INT_USER_VARIABLE_I variables in
ADINA-PLOT.
Note that the PHIST, PHIST1, RN, and PRST arrays are not
applicable to 1-D elements.
Note also that the ARRAY and IARRAY arrays are generated
per integration point. Therefore, the size of these arrays should
not be overestimated. Temporary local arrays should be used
instead for temporary calculations if needed.
analysis.
σ = f (e)
the user must decide between single and multi-step calculation and
also between implicit or explicit stress calculation. An example of
an explicit stress calculation is
t +∆t
σ = t σ + t C E ( ∆e − ∆eth − ∆e IN ) (3.16-1)
t +∆t
σ= t +∆t
C ( t +∆t e − t +∆t eth − t +∆t e IN ) (3.16-2)
where CE is the elastic constitutive matrix and ∆eth and ∆eIN are the
increments of thermal and inelastic strains in the subincrement. For
the 1-D elements, CE is the elastic modulus and ∆eth and ∆eIN are
the increments of axial thermal strain and axial inelastic strains i
the subincrement.
The calculation of Eq. (3.16-2) can be done iteratively or a
predictor-corrector algorithm can be used (such as the elastic
predictor followed by a radial vector corrector).
t +∆t
σ*E = t +∆t C E ( t +∆t e − t +∆t eth − t e IN ) (3.16-3)
t +∆t
∆σ = − C E ∆e IN
If you require the stress invariants, your coding can call subroutine
XJ1232 or XJ123 described in Section 3.16.4.
3.16.3.1 Subincrementation
Strain
Time Time
or t + (k-1) or t + k
e(k)
ij
EPS(*) t t
STRAIN(*) t+t t t
THSTR1(*)
THSTR2(*)
TMP1 No change
TMP2 in value
ALFA
ALFAA
CTD(*)
CTDD(*)
ARRAY(*)
IARRAY(*)
STRESS(4)
where CE is the elastic constitutive matrix and ∆ keth and ∆ keIN are
the increments of thermal and inelastic strains at time t + k∆τ.
For the 1-D element case, equations (3.16-4) and (3.16-5)
Variable descriptions:
Variable descriptions:
Variable descriptions:
3.16-4.
zz STRESS(1)=xx
yz
STRESS(2)=yy
xz
STRESS(3)=zz
yy
STRESS(4)=xy
Z xx
xy STRESS(5)=xz
Y STRESS(6)=yz
X
- Viscoplastic
- Elasto-plastic
- Viscoelastic
- Nonlinear elastic
t
3 sij
t VP
e =γ <φ > (3.16-6)
2 tσ
ij
where
t
sij = component of deviatoric stress
3t t
t
σ= sij sij = effective stress
2
γ = fluidity parameter
In the above,
σ
t
φ= −1 (3.16-7)
σy
• The index i has a range of 1 to 4 for 2-D and 1 to 6 for 3-D, and
follows the conventions defined in Figures 3.16-3 and 3.16-4.
• The rest of this section assumes a 2-D material but all the
expressions can be generalized to the 3-D case.
where
∆e(jk ) =
( t +∆t
ej − tej )
(3.16-9)
NSUBD
3 τσ τ sj
∆e VP ( k )
= ( ∆τ )( γ ) − 1 τ (3.16-10)
j
2 σ σ
y
ν ν
1 (1 − ν ) 0
(1 − ν )
ν
CE =
E (1 − ν ) 1 0
(1 − ν )
(1 + ν )(1 − 2ν )
SYMM
(1 − 2ν ) 0
2 (1 − ν )
1
(3.16-11)
3 tσ
t VP
e =γ <φ > (3.16-6a)
2 tσ
∆e (k )
=
( t +∆t
e − te ) (3.16-9a)
NSUBD
∆eth ( k ) = α ( τ +∆τ θ − τ θ )
3 τσ τσ
∆eVP ( k ) = ( ∆τ )(γ ) − 1 τ (3.16-9a)
2 σy σ
t +∆t
σ iE* = CijE ( t +∆t e j − t +∆t ethj − t eVP
j ) (3.16-12)
∆σ i( k ) = −CijE ∆eVP
j
(k )
(3.16-13)
σ (k ) < σ y (3.16-14)
t +∆t
σ = − x ∆σ ( k ) + σ ( k ) (3.16-15)
σ y − σ (k ) t +∆t
where x = and the σ are the final stresses at time
∆σ ( k )
t+∆t. Similarly, the increments of the viscoplastic strains are
corrected using
t +∆t VP
e = − x ∆eVP ( k ) + eVP ( k ) (3.16-16)
t +∆t VP
where e are the final viscoplastic strains at time t+∆t.
t +∆t
σ *E = C E ( t +∆t e − t +∆t eth − t eVP ) (3.16-12a)
∆σ ( k ) = −C E ∆eVP ( k ) (3.16-13a)
t +∆t
σ = − x ∆σ ( k ) + σ ( k ) (3.16-15a)
t +∆t VP
e = − x ∆eVP ( k ) + eVP ( k ) (3.16-16a)
1
∆si =
aE
( ∆ei′ − ∆eiVP ) (3.16-18)
where
1
∆em = ( ∆e1 + ∆e2 + ∆e4 )
3
(3.16-19)
1
∆e3′ = ∆e3
2
E
cm =
1 − 2ν
(3.16-20)
1+ ν
aE =
E
∂σ m 1
= cm j = 1, 2, 4
∂e j 3
(3.16-21)
∂ si 1 ∂ e′ ∂ ( ∆eiVP )
= i −
∂ e j aE ∂e j ∂e j
∂ ei′
The derivatives can be written in matrix form as
∂e j
2 −1 −1
∂ ei′ 1
= −1 2 −1 i, j = 1, 2, 4 (3.16-22a)
∂e j 3
−1 −1 2
∂ e3′ 1
= (3.16-22b)
∂ e3 2
∂ ei′
=0 for i = 3 or j = 3 and i ≠ j (3.16-22c)
∂e j
In the case of viscoplastic flow in the time step, an approximate
expression for ∂
( ∆e )
VP
i
can be obtained from Eq. (3.16-10) as
∂e j
∂ ( ∆eiVP ) ∂ si
= cVP (3.16-23)
∂e j ∂e j
where
3 ∆τ t +∆tσ
cVP = γ − 1
(3.16-24)
2 t +∆t
σ σy
where
1
c′ = (3.16-26)
aE + cVP
VP
The constitutive matrix C can be obtained using the relation
σ i = si + σ m i = 1, 2, 4
(3.16-27)
σ 3 = s3
∂ si ∂σ m
CijVP = + i = 1, 2, 4 (3.16-28a)
∂e j ∂e j
∂ s3
C3VPj = (3.16-28b)
∂e j
1 1 1
3 ( cm + 2c′ ) 3 ( cm − c′ ) 0
3
( cm − c′)
1 1
( cm + 2c′) 0 ( cm − c′)
3 3
CVP =
SYMM 1
c′ 0
2
1
( cm + 2c′ )
3
1
C VP =
aE + cVP
case.
d ε ijC = γ sij
where ε ijC are the creep strain components, sij are the deviatoric
stress components and γ is the creep multiplier. Note that the
effect of the hydrostatic pressure on the creep deformations is
considered to be negligible.
The creep law is the eight-parameter creep law, i.e.:
ε C = f1 (σ ) f 2 (θ ) f3 ( t )
where ε C is the effective creep strain, σ is the effective stress, θ
is the temperature and t is the time, and
The same creep law also is used in the standard ADINA creep
material model (see Section 3.6).
For multiaxial creep deformation, the calculation of the stresses
is based on the use of an effective stress versus effective (creep)
strain curve. The curve is assumed to be the same as the uniaxial
stress-strain curve of the material.
In case of cyclic loading, the origin of the creep strain tensor
depends on the stress tensor reversal and is determined using the
ORNL rule (subroutines XCRCY1, XCRCY2 and XCRCY3).
The effective-stress-function algorithm is used for stress
integration (subroutines XEFSF1, XEFSF2 and XEFSF3). Either
d ε ijC = γ sij
where ε ijC are the creep strain components, sij are the deviatoric
stress components and γ is the creep multiplier. Consequently,
ε = f1 (σ ) f 2 (θ ) f3 ( t ) f 4 ( x ) + f5 (σ ) f 6 (θ ) f 7 ( t ) f8 ( x )
t0
t - t0
Time
• The second entry in the SCP array (SCP(2)) is stores the flag for
choosing between the strain hardening (value of 1) or the time
hardening (value of 0) procedure in the effective-stress-function
algorithm.
(
E ( t ) = a1 exp a2 1 − ( a3 / t )
a4
)
Elastic
modulus Asymptotic limit
Time
f1 (σ ) = σ
f 2 (θ ) = a5φ a6 + a7φ a8 with φ =exp ( a9θ + a10 )
a25
1 a11 + t a12
a23
a15 + t0a16
a24
a19 + ( t − t0 )a20
f 3 ( t , t0 ) = a18
E ( t ) a13 + t a14 a17 + t0 a21 + ( t − t0 ) 22
a
f4 ( x ) = 1
and
f 5 (σ ) = σ
f 6 (θ ) = a26 (θ + a27 )
a28
1
f7 ( t ) =
E (t )
f8 ( x ) = 1
Note that the term of f3 which depends only on (t - t0) could also be
programmed as the member function f4 (t - t0).
t +∆t ∂ t +∆tσ
case is C EC =
∂ t +∆t ε
Constant, 1: ηG
Constant, 2: G∞
Constant, 3: G1
Constant, 4: β1
...
Constant, 2 (ηG + 1) + 1 : η K
Constant, 2 (ηG + 1) + 2 : K ∞
Constant, 2 (ηG + 1) + 4 : γ 1
...
Constant, 2 (ηG + η K + 2 ) + 1 : T0
Constant, 2 (ηG + η K + 2 ) + 2 : α 0
Constant, 2 (ηG + η K + 2 ) + 4 : C2
In the 1-D element case, only one series of modulus and decay
constants is needed for the property description, i.e.,
Constant: η E
Constant: E∞
Constant: E1
Constant: γ 1
...
Constant, 2 (η E + 1) + 3 : T0
Constant, 2 (η E + 1) + 4 : α0
Constant, 2 (η E + 1) + 5 : C1
Constant, 2 (η E + 1) + 6 : C2
σ = 0σ y + C y ( e p )
t n
de p = de pi + de pk = Mde p + (1 − M ) de p
p p
where de i and de k are, respectively, the isotropic and kinematic
contributions, and the yield function of the mixed hardening model
is
1 t 1
t
fy =
2
( s − t α ) ⋅ ( t s − t α ) − tσ y2 = 0
3
dα = C p (1 − M ) de p
2 E p − ME p
M
Cp =
3 1− M
M = M ∞ + ( M 0 − M ∞ ) exp ( −η e p )
Constant 1: ν
Constant 2: ε1
Constant 3: σ 1
Constant 4: ε 2
Constant 5: σ 2
Constant (2*n): ε n
Constant (2*n+1): σ n
(σ i,ε i ) (σ n ,ε n )
ε
(σ 1,ε 1 )
(σ 2 ,ε 2 )
• In Solution Control table, the first row is a flag for using tangent
or secant stress-strain matrix when the stress-strain curve enters
into a softening region. The secant stress-strain matrix is employed
for softening when this flag set to 1. Otherwise, the tangent stress-
strain matrix will always be used.
4. Contact conditions
• Contact conditions can be specified in ADINA to model contact
involving solid elements (2-D and 3-D solids) and/or structural
elements (truss, beam, iso-beam or axisymmetric shell, plate, shell
and pipe elements).
ref. KJB ref. Bathe, K.J. and Chaudhary, A., "A Solution Method for
Section 6.7 Planar and Axisymmetric Contact Problems," Int. J.
Num. Meth. in Eng., Vol. 21, pp. 65-88, 1985.
4.1 Overview
• Contact groups (and their contact surfaces) in ADINA can be
either 2-D or 3-D. The contact surfaces should be defined as
regions that are initially in contact or that are anticipated to come
into contact during the solution.
Contact surface 3
Body 2 l
l l
l
D l l
l Body 3 l
Contact surface 2 l
l l
l l l
l
l l l l l
l
l l l
A
l
l l
l l
l l l
l
C l
Bl Contact surface 1
Body 1
Contact surface 1
(surface of cylinder)
Body 1
Contact surface 2
(top surface of
body 2)
Body 2
No penetration Target
surface
Contactor
surface
Contactor
No penetration surface
Target
surface
Figure 4.1-3: Contactor and target selection
pair surface A can be the contactor and surface B the target, and in
another contact pair surface B is the contactor and surface A is the
target.
A non-zero contact surface compliance should always be used
with symmetric contact pairs. Without compliance, the system of
contact equations can become overconstrained, that is, the same
contact equation can be present more than once in the system of
contact equations. When this happens, the global system of
equations becomes singular and the solution may diverge.
• Basic concepts
τ ≤1
and τ < 1 implies u = 0 (4.1-3)
while τ = 1 implies sign ( u ) = sign (τ )
Sliding: the gap between the contactor node and the target
segment is closed; a compression force is acting onto the
contactor node and the node kinematically slides along the
target segments (either due to frictionless contact or due to a
frictional restrictive force less than the limit Coulomb force.
Continuum
n
l
l k+2
k-1 nr
l l
Contactor
k k+1 surface
Z
Contactor segment
n =normal vector, acting into the continuum
Y nr=tangent vector
N4
l
N2
l
r
n
s
l
N1
• Single-sided contact
For single-sided contact (see Fig. 4.1-6), one side of the contact
surface is assumed to be internal and the other side to be external.
Any contactor node within the internal side of a target surface is
assumed to be penetrating and will be moved back to the surface.
This single-sided option is ideal for contact surfaces on the faces of
solid elements since in that case it is clear that one side is internal
to the solid while the other is external. In this case, the external
side can usually be predicted from the geometry. This option is
also useful for shells when it is known that contact will definitely
occur from one direction. In this case, however, the program
cannot intuitively predict the internal side of the contact surface.
Single-sided contact is selected using CGROUP ...
PENETRATION-ALGORITHM=ONE (the default).
Internal side
(no contactor nodes allowed)
Target
surface
External side
• Double-sided contact
In double-sided contact (see Fig. 4.1-7), there are no internal or
external sides. The contactor surface nodes in this case are
prevented from crossing from one side of the target contact surface
to the other during solution. This option is more common for shell-
based contact surfaces. If a contactor node is on one side of the
target surface at time t, it will remain on the same side at time t +
∆t.
Double-sided contact is selected using CGROUP ...
PENETRATION-ALGORITHM=TWO. Double-sided contact is
only available for 3D contact.
• Tied contact
When the tied contact feature is selected for a contact group,
ADINA performs an initial contact check at the start of the
analysis. All contactor nodes that are found to be in contact or
Gap between
contact surfaces
(exaggerated)
Target surface
x*1
N x*1
x1 x1
Contactor surface
x*1
x1 x*1
x1
• Node-node contact
When node-node contact is used, each node of the contactor
surface is paired with a node on the target contact surface (a node-
node contact pair). A node on the contactor surface can come into
contact with the infinite plane passing through the paired node on
the target surface; the infinite plane is perpendicular to the normal
direction indicating the direction within the target body (Fig. 4.1-
10). This normal direction, and hence the infinite plane, is constant
during the response. The normal directions can be based on the
geometry of the target segments or on the vector connecting each
contactor and target nodes.
Node-node contact is selected using CGROUP ...
NODETONODE=YES.
Contactor node
in node-node
contact pair
Constraint-function method,
algorithms.
2
g +λ g −λ
w ( g, λ ) = − + εN
2 2
g
w(g,)
results in a smooth transition from stick to slip and vice versa, and
it also results in a differentiable friction law that is less likely to
cause convergence difficulties.
Two friction regularization algorithms are available in ADINA.
Both constraint functions take the form v ( u ,τ ) = 0 .
The newer default algorithm involves a more accurate
linearization of the frictional constraints and, in general, converges
much faster than its predecessor. The v function is defined
implicitly as a multilinear function as shown in Fig. 4.2-2. Here εT
is a small parameter (EPST parameter in the CGROUP commands)
which has the physical meaning of the "sticking velocity", that is,
the maximum velocity corresponding to sticking conditions.
.
-10 -2 -1 1 2 _
10 u
,T
-1
2 u−v
τ + v − arctan =0
π εT
_
2
%
1
.
-10 -2 -1 1 2 _
10 u
,T
-1
• Note that the target surface can be rigid in all three contact
algorithms. The presence of a rigid target does not mean that the
rigid target algorithm must be used.
δN
FCN = M C aCN * = M C N
∆t 2
contactor nodes are touching it. So, the FCN force above is
projected to the target segment nodes:
FTNi = N i FCN
M Ti = N i M C
and this mass is added to that of the target node itself. Then the
acceleration of the target node is determined as
aTN ( M T + ∑ M Ti ) = ∑ FTNi
F = A P = A N ( K N δ N + K Dδ N )
x
FT = min A KT xT , µ FN T
xT
Offset
t Actual contact
surface
(a)
Internal side Defined contact
surface
Offset
Actual contact
t surfaces
t
(b) Offset
Defined contact surface
Figure 4.4-1: Contact surface offsets for: (a) single-sided contact, and
(b) double-sided contact
Nodal normals
• Initial penetration
The treatment of initial penetrations in ADINA is governed by the
INITIAL-PENETRATION parameter of the CGROUP commands.
By default, if there is initial overlap (penetration) between a
contact node and a target segment in the first solution step, ADINA
attempts to eliminate the overlap. ADINA can eliminate the overlap
at the first step or over a user-specified time using the TIME-
PENETRATION parameter of the CGROUP commands. This
feature is useful if the initial penetrations are too large to be
eliminated in a single step.
The program can also calculate initial penetrations at the start of
solution and ignore them in future steps. In this case the program
does not detect penetration for a contactor node if the amount of
penetration is less than or equal to the recorded amount. Fig. 4.4-4
shows some of the possibilities. See Section 4.8.2 for modeling tips
related to this feature.
This feature can be interpreted as a redefinition of the contact
gap, as follows:
Target surface
Continuum
Eliminate penetration
over time period Ignore penetration
Overlap
distance
• Gap override
In the gap override feature, the gaps and penetrations calculated
from the finite element mesh are replaced by a fixed user-specified
value. A positive value represents an initial gap, zero means that
the contact is touching the target, and a negative value represents
an initial penetration (which can be removed either immediately or
Target
surface e = L x contact surface extension
L
e
Extended target
surface
where
,P /A
• Contact birth/death
The contact birth feature activates a contact group at a specific
time, while the contact death feature disables a contact group at a
specific time. They are set via the TBIRTH and TDEATH
parameters of the CGROUP commands. A 0.0 birth time means
that the contact group starts active at the beginning of the analysis,
and a death time less than or equal to the birth time means that the
contact group does not die.
• Friction delay*
When the friction delay feature is activated, frictional conditions
are applied to a contactor node one time step after contact is
established. This feature can be useful in many problems, since it
delays the non-linearity associated with friction until contact is
established.
Note that the relative sliding velocity cannot be uniquely
determined when a node was not in contact at time t, and is in
contact at time t+∆t. That velocity depends on the exact time at
which contact started, which is somewhere between times t and
t+∆t (see Figure 4.4-8). Delaying friction is equivalent to assuming
that contact was established close to time t+∆t, and hence the
sliding velocity is zero and so is the frictional force.
The friction delay feature is activated using the FRIC-DELAY
parameter of the CGROUP commands. Friction delay is off by
default.
• Contact slip*
Contact slip allows a tangential or slip velocity to be added
component between contact surfaces. With this feature, you can
model out-of-plane slip in 2-D axisymmetric analysis. This feature
is considered to be a load added to the model. See Section 5.11 for
further details.
Time = t
t+2t t+t
x1t
t+2t t+t
FT FT .t
x1 R= Relative sliding velocity
t+2t t+t t+t
FN FN Friction delay FT = 0
. .
x1t+2Rt unique x1t+Rt non-unique
It should be noted that if symmetric contact pairs are used, then the
tractions output are one-half of the expected values. This is
because the tractions are computed solely from the contactor
forces, and when symmetric contact pairs are used, the contactor
forces are one-half of the total forces.
Tractions:
NODAL_CONTACT_GAP
NODAL_CONTACT_STATUS (see note below)
NODAL_SLIP_VELOCITY (magnitude of slip
velocity)
NODAL_SLIP_VELOCITY-{XYZ}
NORMAL_TRACTION
NORMAL_TRACTION-{XYZ}
TANGENTIAL_TRACTION
TANGENTIAL_TRACTION-{XYZ}
See Section 13.1.1 for the definitions of those variables that are
not self-explanatory.
Forces:
Force output is available for all contact nodes under all contact
conditions. The AUI variables are: CONTACT_FORCE-X (3-D
contact only), CONTACT_FORCE-Y, CONTACT_FORCE-Z,
CONTACT_FORCE-A (3-D contact only),CONTACT_FORCE-B,
CONTACT_FORCE-C. These forces are sometimes called
"consistent contact forces".
4.5 Friction
ADINA has a general Coulomb type friction model, where the
coefficient of friction µ can be a constant or calculated based on
several user-defined friction laws. Explicit analysis however, only
supports standard Coulomb friction.
( )
Model 1: Tt = A1 1 − exp ( − A2Tn ) , where A1, A2 are specified
constants. This implies that the instantaneous coefficient of friction
1 − exp ( − A2Tn )
is µ = . Model 1 is graphically shown in Fig.
Tn / A1
4.5-1.
Tt
A1
A1 A2
Tn
Tt
Slope A1
Slope A2
Tn
A1
A2
A2 can be less than, equal to,
or greater than A1.
Tn
µ = A1
A1 if u ≤ A3
µ=
A2 if u > A3
u
A1 + ( A3 − A1 ) if u < A2
µ= A2
A3 if u > A2
( A v ) 2 + ( A v ) 2 + ( A v ) 2 if u > A5
µ= 1 (1) 2 (2) 3 (3)
A4 if u ≤ A5
where v(1), v(2) and v(3) are the x, y and z components of the sliding
direction.
µ = A1 + A2 Fn , 0 ≤ µ ≤ 1
t
A1 + ( A3 − A1 ) if t < A2
µ= A2
A3 if t > A2
A1 + A3 x (2) + A4 x (3) in 2D
µ= , 0 ≤ µ ≤ A2
A1 + A3 x (1) + A4 x (2) + A5 x (3) in 3D
A2 if θ < A1
A −θ θ − A1
µ = A2 3 + A4 if θ < A3
A3 − A1 A3 − A1
A4 if θ ≥ A3
1 − exp(− A2Tn )
µ=
Tn A1
1 − exp(− A2 Fn )
µ=
Fn A1
µ = A2 + ( A2 − A1 ) exp(− A3Tn )
µ = A2 + ( A2 − A1 ) exp(− A3 Fn )
contact once the outer ones are removed. This option can also be
used to model 3D contact problems involving beams (since the
beams cannot define a 3D contact surface).
Contact bodies are defined using the CONTACTBODY
command.
Contact bodies cannot be used in tied contact.
Segment on
target surface
Contact body
(a collection of nodes)
No segments on
contactor surface
• The contact algorithm itself for a certain contact group can also
change in a restart. For this purpose, the contact algorithms can be
divided into two categories. The first catergory includes the
constraint function (implicit), Lagrange multiplier segment
(implicit), kinematic constraint (explicit), and penalty (explicit).
Restarts are possible between different algorithms in this catergory.
The second category includes the implicit and explicit Rigid Target
algorithms. Restarts are possible between these algorithms.
However, restarts are not allowed between the two categories.
• The contact damping feature allows the user to add normal and
tangential grounded viscous dampers to all contactor and target
nodes in the model. Using the same value for both normal and
tangential direction results in isotropic viscous damping. The
damping force on each node is
FDamp = CN u N + CT uT
Body 1 (contactor)
Body 2 (target)
Contactor nodes
Target segments
Contactor
surface
Target
surface
Outer cylinder
Inner cylinder
Contactor surface,
marked with
Target surface,
marked with
Geometry before
discretization
Overlap to ignore
Contact surface
with sharp corners
Contact surface 1
Contact
Contact surface 3
surface 1 Contact surface 1
(with fine
corner mesh)
Contact surface 2
The inaccuracy in this case results from the way the contact is
enforced. The kinematic constraint method first predicts
displacements without contact then applies a contact correction.
The contact conditions are satisfied more accurately when the
penetrations in the predicted configuration are small which is
usually the case due to the small time step size of explicit analysis.
However, some cases such as that mentioned above lead to large
projected penetrations which results in incorrect contact conditions
and tensile contact forces.
One node
in contact
Tensile forces
Wrong contact
region
Projected configuration
(before contact correction) Final configuration
• Three non-contact related norms are given: first for the energy
convergence criterion, the displacement and rotation convergence
criterion (boxes d and e), and the force and moment convergence
criterion (boxes b and c). Each box has 3 lines of output with the
top one giving the norm of the quantity, the second one giving the
node corresponding to the maximum value, and the third line
giving the maximum value. See Chapter 7 for definitions and more
details on these norms.
equilibrium iterations
F
Applied Blankholder
Contact pair 1 load pressure
Symmetry
Fixed die
Contact Contact
pair 2 pair 3
The damping constants have units of force per unit velocity. Hence,
their proper value is problem dependent. If initial contact damping
is used to stabilize a problem involving two contact bodies at least
one of which is unsupported, and with a gap between them, then a
good estimate of the damping constants CN and CT is one in which
the gap is nearly closed in the first iteration. Starting with the
dynamic equations of motion (see Equation 7.3-1) and canceling
out the inertial term (static analysis), and the stiffness term (since
one or both bodies are initially unsupported), we obtain
C U=R
R ∆t
CN = CT =
Ng
where R is the sum of the applied loads at the first time step.
Note that this is only an estimate, but is frequently an acceptable
one.
4.9.1 Introduction
The rigid target contact algorithm is intended for use in
applications in which the target surfaces are considered to be rigid.
Prescribed punch
displacement
Punch
Blank holder
Blank
Die
Drawbead
a) Physical problem
Contact Drawbead
surface Contact
4 (target) surface 2 Drawbead
(target) projection
Contact
Contact surface 3
surface 1 (target)
(contactor),
offsets used
to model blank
thickness
Figure 4.9-1: Sample metal forming analysis using the rigid target
contact algorithm
Models that were set up using the V83 rigid target contact
algorithm may need to be revised when using the current rigid
target contact algorithm, see the conversion hints in Section 4.9.6.
We suggest that new models not be set up using the V83 rigid
target contact algorithm.
When there are offsets specified, the offsets can either be described
using spheres centered around the contactor nodes (Fig. 4.9.2(b)),
or using the contactor normals (Fig. 4.9.2(c)). In either case, the
offset magnitude is either constant or taken from the current
thickness of attached shell elements, as described below.
Contact Contact
segment 1 segment 2
Segment normals
a) Contactor segments without offsets. Segment normals are not used.
When the offsets are described using the contactor normals, offset
vectors are constructed using the averaged contactor normals and
the offset magnitude. The upper and lower contactor points are
constructed from the contactor nodes and the offset vectors.
Radius of sphere
Upper surface = contactor offset
Lower surface
b) Contactor segments with spherical offsets. Segment normals are not used.
Offset
vector
Contactor
node normal
Lower contactor point
Contactor surface
Targ nt 2
et seg gme
men et se
t1 Targ
Normals offset
Targ t2
et seg gmen
men et se
t1 Targ
Target edge 1
Contactor node
Target segment
Closest point on
target segment
Target segment normal
n
Contactor node
Target edge
n Closest point on
target edge
Fig. 4.9-5 shows two target segments with a common target edge.
The shaded volumes indicate which of the target entities any
contactor node is closest to. Notice that the shaded volume in
which the contactor node is closest to the target edge depends upon
the angles between the target segments attached to the edge.
g = d − GAPBIAS
where GAPBIAS can be chosen to, for example, not model contact
even if there is geometric overlap. (The default for GAPBIAS is
0.)
Edge 1
Segment 1
Segment 2
g = d − OFFSET − GAPBIAS
Contactor node
OFFSET
d
d - OFFSET
Target segment
Closest point on
target segment
Slope -kn
Fn
-DEPTH g
Tensile contact: The basic ideas for tensile contact are illustrated
in Fig. 4.9-8 to 4.9-10.
Fig. 4.9-8 shows the iteration history when tensile contact is not
used. For iteration ite-2, the contactor node and target segment
overlap. Hence contact is assumed between the contactor node and
target segment. For iteration ite-1, because of the relative motion
of the contactor node and target segment, the contactor node and
target segment do not overlap. For this iteration, no contact is
assumed between the contactor node and target segment. For
iteration ite, there is a large overlap because the contactor spring
unloads, since there are no forces acting on the contactor spring,
and the target does not provide any stiffness to the contactor node.
This large overlap causes large contact forces, which can cause
trouble in convergence in the successive iterations.
Spring is compressed
Contactor node Spring is uncompressed
Target segment
Fig. 4.9-9 shows the iteration history when tensile contact is used.
Now, in iteration ite-1, tensile contact is assumed between the
contactor node and target segment. In tensile contact; the target
surface still provides stiffness to the contactor node. Hence the
overlap in iteration ite is small.
Fig. 4.9-10 shows the iteration history when tensile contact is used,
and the gap is large. In iteration ite-1, tensile contact is assumed,
and in iteration ite, no contact is assumed.
Contactor node
Target segment
to converge, unless the tensile forces are all less than the value of a
user-input parameter (see Section 4.9.3). Hence the use of the
tensile contact feature does not affect the converged solution.
Contactor node
Target segment
Free overlap: The basic ideas for free overlap are illustrated in Fig.
4.9-11 to 4.9-13.
Fig. 4.9-11 shows the iteration history when free overlap is not
used. For iteration ite-3, there is a gap between the contactor node
and target segment. Hence no contact is assumed between the
contactor node and target segment. For iteration ite-2, because of
the relative motion of the contactor node and target segment, the
contactor node and target segment overlap. For this iteration,
contact is assumed between the contactor node and target segment.
For iteration ite-1, the contactor spring is stretched because there is
a large stiffness between the contactor and target due to the contact.
There is a small gap between the contactor and target, hence no
contact is assumed. For iteration ite, the contactor spring is
unloaded, and there is a gap between the contactor and target.
Fig. 4.9-12 shows the iteration history when free overlap is used.
Now, in iteration ite-2, free overlap is assumed between the
contactor node and target segment. Hence there is no contact
assumed for this iteration. In iteration ite-1, the contactor spring is
unaffected by the motion of the target.
Fig. 4.9-13 shows the iteration history when free overlap is used,
and the overlap remains between target and contactor. In iteration
ite-2, free overlap is assumed, and in iteration ite-1, contact is
assumed.
u f = ( u c − ut ) − ( ( u c − ut ) ⋅ n ) n
uf
Ff = µ Fn , u f < u f min
u f min
= µF, u f ≥ u f min
.Fn
Sliding
. . .
uf uf u
.Fn
Sliding Sticking
4.9.2.5 Drawbeads
Blank holder
Blank
Blank motion
Die
a) Drawbead diagram
Drawbead node
Projection of drawbead node
Target 1
Contactor
Contactor motion
Target 2
c) Drawbead tractions
Drawbead node
Drawbead segment
Projection of
drawbead onto
target surface 1
Projection of
drawbead onto
contactor surface
Projection of
drawbead onto
target surface 2
Distance between
Contactor target surfaces
thickness
u DB
TR = TRMAX , u DB < u DB min
u DB min
= TRMAX , uDB ≥ uDB min
TR
TRMAX
Sliding
.
uDBmin
. .
uDBmin uDB
TRMAX
Sliding Sticking
TR
TU = TUMAX
TRMAX
Restraining traction
acting onto
contactor surface
CONTACT-CONTROL … RT-ALGORITHM=V83
Rigid links
and
When normals are used for the offset description, small steps
should be used in static and implicit dynamic analysis. This is
because the offset vectors are assumed to remain constant during
the equilibrium iterations. In particular, at convergence, the offset
vectors corresponding to the previous converged solution are used.
In both Fig. 4.9-22 and Fig. 4.9-23, the large overlap is unintended.
In static and implicit dynamics, the equilibrium iterations would
most likely not converge. In explicit analysis, the large forces
between contactor and target would overdistort the elements
attached to the contactor node.
One way to avoid the large overlaps is to use the DEPTH feature so
that contact is not detected between the contactor node and the
incorrect target segment. Another way to avoid this issue is to
create additional target segments as shown. Then the contactor
node is closest to one of the additional target segments.
We recommend that you use the smallest value of kn such that the
maximum overlap is still acceptably small. For example, if the
target surface is curved, there will be a geometric error associated
with using a coarse contractor surface (Fig. 4.9-24). There is no
advantage if the maximum overlap is less than the geometric error.
So, if the mesh is coarse, a large maximum overlap can be used.
Lower target
Contactor node
cannot be in contact
with lower target Contactor node is closest
Upper target to target edge
Upper target
Closest point
on target surface Closest point
that interacts on target surface
with contactor node Target surface that interacts
with contactor node
Contactor
node
Contactor
node
Closest point
on target surface
that interacts
with contactor node
DEPTH
This issue also arises when lower-order elements are used, but
when lower-order elements are used, the variation in the consistent
nodal point forces is much less, so kn can be smaller for the same
accuracy in the element stresses.
Explicit dynamics time step size: In explicit dynamics, the time step
should be smaller than
m
∆t = 2
kn
Geometric error
Young’s
modulus E
L Contact area A
ttu.
Contactor node at t
Time step selection in frictional contact: The time step size will
affect the frictional velocities and hence the results. This is
because, in static analysis, the nodal velocities used in the friction
calculations are calculated as the incremental displacements
divided by the time step.
2mu f min
∆t =
µ Fn
Figure 4.9-27: Reverse sliding due to too large time step in explicit
dynamics
2mu f min
Note that when the time step is greater than ∆t = , the
µ Fn
solution is still stable.
2mDB u DB min
∆t =
TRMAX
(similar to the time step for friction). Here mDB is the contactor
mass/unit length.
mi
∆t = min
i kn
where the minimum is taken over all contactor nodes. Notice that
friction is not considered in the automatic time step selection; this
is because the model remains stable even if the time step is larger
than the friction time step discussed above.
CGROUP … TENS-CONTACT={YES/NO}
This controls the use of the tensile contact feature discussed above.
The default is YES. Tensile contact is not used in explicit
dynamics.
CGROUP … FREE-OVERLAP={YES/NO}
This controls the use of the free overlap feature discussed above.
The default is NO. Free overlap is not used in explicit dynamics.
CGROUP … GAP-PUSH
CGROUP … TENSILE-FORCE
The maximum tensile force for a node in tensile contact for which
convergence is allowed (default value 0.001). All nodes in tensile
contact must have a tensile force less than this value for the
solution to converge.
CGROUP … OSCILLATION-CHECKING
4.9.4 Rigid target contact reports for static and implicit dynamics
Meaning: Self-explanatory
The following messages are output at the end of each solution that
did not converge.
Meaning: The contactor node for which the contact force had
the largest change is output.
The following items are output for each time step in which results
are printed or saved:
Prescribed displacement
Target surface 2
Target surface 2
• The contact search algorithm may take a relatively long time for
the first iteration of the first time step. Similarly, the contact search
algorithm may take a relatively long time for the first iteration of
any time step in which a contact group or contact pair is born.
a) Poor modeling
b) Good modeling
c) Good modeling
The following hints may be useful when running models that were
set up using the rigid target algorithm of ADINA 8.3 and earlier:
• The results from the V83 and current rigid target contact
algorithms will usually be quite different:
tR tR
200 l 200 l
1 2 t 2 4 t
Run 1: t = 1.0 Run 2: t = 2.0
(
• A time function is defined as a series of points t , f i ( t ) in )
which t is time and fi ( t ) is the value of time function i at that
time. Between two successive times, ADINA uses linear
interpolation to determine the value of the time function.
Constant: f ( t ) = f *( t )
µ0
f ( t ) = f *( t ) ⋅ ⋅ 2 ⋅ I ⋅ ( sin (ωt − φ + α ) + exp ( −t / τ ) ⋅ sin (φ − α ) )
4π
Time functions:
f(t) Input time function Time function shifted
by arrival time
4
2
Time at solution
1 start=0.0
t
2t 4t 6t 8t 10t
-1 Arrival
time Load active
-2
Solution period
-3
Time
step 1 2 3 4 5 6 7 8
n tR 0 0 30 40 40 40 0 -10
Time functions:
f(t) Input time function Time function shifted
by arrival time
4
2
Time at solution
1 start = arrival time
t
Time at 2t 4t 6t 8t 10t
-1 solution
start Load active
-2 Arrival
time Solution period
-3
Time
step 1 2 3 4 5 6 7 8
B
Y
i j
X
A
X Vm
1
Vk1 A
2-D elements, the area over which a pressure load acts includes the
element thickness.
Auxiliary node
Rigid link
Node of load
application
k
Thin cantilever
Follower force
at t = 0
a) Configuration at time t=0
Auxiliary node
t+
k
Rigid link
Follower force
at time t
b) Configuration at time t
Z
Y
X
X Y
a) Two-dimensional element b) Three-dimensional element
l
l
l
l
l
l
l Z
Y
X
a) Beam element b) Shell element
P7
P5 P6
P1 P2 P3 P4
P2 P2 P2 P2 P2
P1 P 1 P1 P 1 P1 P1 P1 P1 P1 P1
P2 P2 P2 P2 P2 P = P1 + P2
1
P= (P + P2)
2 1
q2
q1
1l l2
L
L L
F1 = (7q 1 + 3q 2) F2 = (7q 2 + 3q 1)
20 20
l l
L
L2 (3q + 2q ) L2 (3q
M1 = 1 2 M2 = 2 + 2q 1)
60 60
A B
Bending
moment
MA
MB
A B
l l l l
Bending
moment
MA
MB
b) Numerical results using 3 beam elements
Shear Bending
force (V) moment (Mb)
0 0
x x
L/2 L L/2 L
V Mb
Zero
0 0
x x
L/2 L L/2 L
l l l
L/2 L/2
V Mb
0 0
x x
L/2 L L/2 L
l l l l l
V Mb
0 0
x x
At the beginning of the first time step, the prescribed load factor is
1.0.
At the beginning of the second time step, the prescribed load factor
is 1.0 if the contact pressure at the node is less than the critical
pressure, and the prescribed load factor is 0.0 if the contact
pressure at the node is greater than the critical pressure.
• When the element birth and death option is active (see Section
11.4), the consistent load vector only contains the contributions
from the elements currently alive.
• Note that the arrival time option (see the beginning of this
chapter) is not applicable for centrifugal, rotational, and mass
proportional loading.
t
R = − ∫ ρ[ ( α × r)
t t
+ t ω × ( t ω × t r ) ] dV (5.4-1)
angular acceleration centripetal acceleration
(
acceleration, t ω = OMEGA FACTOR ⋅ f ( t ) n is the angular )
velocity, t r is the radial vector from the axis of rotation to the
node, ALPHA, OMEGA and FACTOR are constants specified in
the load definition, ρ is the density, f(t) is the time function and n
is the unit vector parallel to the axis of rotation.
t
t
0 = angular velocity t
0
t
= angular acceleration
B
Points A and B are not Tangential forces
necessarily nodal points.
Z A
Centrifugal forces
Y
X Structure
M t +∆t U (i ) + t +∆t K (
i −1)
∆U (i ) = t +∆t R ( i ) − t +∆t F (
t +∆t i −1)
(5.4-2)
t +∆t
( (
R (i ) = − ∫ ρ ω × ω × ( 0r + t +∆t U (i −1) + ∆U ( i ) ) dV ))
(5.4-3)
• The position and direction of ω rel and α rel remains fixed with
respect to the reference frame. The total angular velocity of the
structure is ωtot = Ω frame + ω rel , and the total angular acceleration
is α tot = α frame + α rel .
B r rel
r frame
A
Z
structure
reference frame Y
t
R = ∫ −ρ ( t
α frame × t
r frame + t
α rel × t
r rel )
Angular Acceleration
+ ( t
Ω frame × ( t Ω frame × t
r frame ) + t
ω rel × ( t ω rel × t
r rel )) (5.4- 4)
Centripetal Acceleration
+ 2 t
Ω frame × ( t ω rel × t
r rel ) dV
Coriolis Acceleration
t
ai = MAGNITUDE ⋅ f ( t ) ⋅ Ai , i = x, y , z
the direction vector for the nodes attached to the ground, with “1”
in the portions of the translational degrees of freedom acting into
the direction i and “0” in the other portions.
The matrix equation of motion becomes
M11 M12 u1 d1i xgi K11 K12 u1 d1i xgi R1
MT + + T + =
12 M 22 0 d 2i xgi K12 K 22 0 d 2i xgi R 2
(5.4-7)
K K d 0 d
Now 11 T
12 1i 1i
d = 0 since the vector di = d corresponds
K
12 K 22 2 i 2i
to a rigid body motion. The matrix equation of motion becomes
M11 M12 u1 K11 K12 u1 R1 M11 M12 d1i xgi
MT + T = −
12 M 22 0 K12 K 22 0 R 2 M12
T
M 22 d 2i xgi
(5.4-8)
Please note:
t
N t NSTEPt
Figure 5.4-3: Use of gravity loading in nonlinear static
analysis with other types of loading
ui = f (t)
where f is a general time function (see Section 5.1) which
prescribes the nodal point displacement ui. Prescribed
displacement conditions can be used in static and dynamic analysis.
- Newmark
- Bathe method
and
N
1
t
θi =
N
∑ f (t ) F
j =1
j i
j
Temperature Temperature
+ +
t * t
-0+
+
0 t = ARTM Time
(c) Combined temperature variation (cases (a) and (b) are added)
input.
+1
+0 l l l l l l l
x
El. 1 El. 2 El. 3 El. 4 El. 5 El. 6
Quadratic elements:
+1
+1+ 0
+0 l l l l l l l
x
El. 1 El. 2 El. 3
Cable 1 i1(t)
dF21
dr1
r21
Cable 2
dr2
i2(t)
Total
velocity slip
0 0/r Relative velocity
between contact
surfaces
r
Contact surface
Node of
contact surface
Axis of contact
slip rotation
3-D surface
T T
T
T
Solid
domain T
2-D line
DO 100 M=1,NUMNP
DO 200 N=1,NUMNP
.
.
.
CALL IUSER(M,N,NUMNP,MFLAG,NFLAG)
IF (MFLAG.EQ.0) GO TO 100
IF (NFLAG.EQ.0) GO TO 200
.
.
.
CALL USERSL(...)
.
.
.
200 CONTINUE
100 CONTINUE
Fc = 0.5 D ρ CD ( U f − U c ) U f − U c
∂U f ∂U f
+0.25π D 2 ρ (1 + C A ) + ( U f − U c )i (5.14-1)
∂t ∂x
−0.25π D 2 ρ C A U c
Tow speed
Water surface
X
Hydrodynamic drag
Towing cable and inertia forces
Towed sphere
(1) (2)
(1) Initial geometry
(2) Structure submitted to gravity and hydrodynamic forces
Fc = −0.5 D ρ CD U c U c − 0.25π D 2 ρ C A U c
2 3
D 4 D
Fs = −0.5 ρ CD U s U s − π ρ C A U s
2 3 2
U k = ∑ β jU j (5.15-1)
j
∑β U
j
j j =0 (5.15-2)
Note that each constraint of the first type reduces the number of
independent degrees of freedom by one, while each general
constraint of the second type increases the number of degrees of
freedom by one (by adding a Lagrange Multiplier). Hence, the first
type should be used whenever possible. In some cases, however,
one cannot easily express a constraint in the form of Eq. (5.15-1)
where a slave degree of freedom is only related to master degrees
of freedom. See the example below for an illustration.
• Note that in the above example, the problem with the second
constraint could be resolved by manually applying the first
constraint to obtain
U 3( z ) = 2U1( z ) + 3θ1( x ) − 4θ 2( x )
Zero stiffness
for director vector
rotational DOF
Rigid links
Nonzero stiffness
for director vector
rotational DOF
Shell elements
KINEMATICS … RIGIDLINK-6DOF=ALL
With this option, the AUI does not fix the rotational DOFs of nodes
connected to continuum elements and rigid links, and the model in
Figure 5.15-2 will behave as shown in Fig 5.15-2(c).
S S M = Master node
S = Slave node
M M
M S
Rigid link
a) Two element groups connected with rigid links
∫ Γ
λ ⋅ (u1 − u 2 )d Γ = 0 (5.16-1)
∫ ΓS
λ S ⋅ (u M − u S )d Γ S = 0 (5.16-2)
• Only 2-D solid elements can be used in the gluing of 2-D lines.
The glued element side can have linear or quadratic displacement
interpolation. Truss, beam and pipe elements are not supported.
• If one glue surface is smaller than the other, the smaller surface
Master surface
Slave surface
Slave surface
Master surface
Glue surface 2
Glue surface 1
One glue-mesh
Four glue-meshes
(one for each side)
• If the two surfaces have different mesh densities, either one can
be used as slave (unlike contact where the finer one should be the
contactor). Using the finer meshed surface as a slave will produce
more equations, since the Lagrange Multipliers degrees of freedom
are on the nodes of the slave surface.
6. Eigenvalue problems
ref. KJB The two types of analyses in ADINA for which the solution is
Section 6.8.2 obtained by solving the corresponding generalized eigenvalue
and Chapter 10 problem are:
t1
Kφi = γ i t0 Kφi ; γi > 0
or (classical formulation)
t1
Kφi = γ i ( t1 K − t1 K NL ) φi , γi > 0
t1 t0
where K and
K are the stiffness matrices of the structural
systems at times t1 and t0, respectively, t1 K NL is the geometrically
t1
nonlinear part of K , t0 is the time at solution start, t1 equals
t0 + ∆t where ∆t is the time step increment, φi is the buckling
mode shape for mode i and the critical (buckling) load factor λi for
mode i is related to the eigenvalue γi using
1
λi =
1− γ i
1 < λ1 ≤ λ2 ≤ ... ≤ λM
and
λM +1 ≤ λM + 2 ≤ ... ≤ λN < 0
t
where t1 R and 0 R are the externally applied load vectors at t0
and t1, respectively.
1
Error
( λi( k ) )
2 2
bound
for = 1 − ( k ) T ( k )
mode i ( qi ) qi
γ i( k +1) − γ i( k )
≤ SSTOL; i = 1, 2,..., N
γ i( k +1)
(a) a larger load level for the first time step; and
(b) a larger number of iteration vectors.
K φi = ωi2 M φi
φiT M φi = 1
1
Error
( λi( k ) )
2 2
bound
for = 1 − ( k ) T ( k )
mode i ( qi ) qi
Physical
error Kφi − ωi M φi
2
bound =
for mode i Kφi
frequencies.
In order to obtain a frequency solution when there are zero
frequencies, you must activate the "allow rigid body modes"
option, otherwise ADINA will stop. This option is also useful if
any part of the model would be insufficiently supported (i.e., have
rigid body modes) if all contact, mesh glueing and generalized
constraints are removed.
When the rigid body mode shift option is used, you can enter
the rigid body mode shift value (which must be negative). If you
instead use the default shift value of 0.0, ADINA calculates it using
the formula
k
RBMSH = f min ii
i =1,..., N m
ii
-3
where f = -10 for the determinant search method or if contact is
-7
present, f = -10 otherwise and N is the number of mass degrees
of freedom in the finite element model.
ref. KJB
• The use of the determinant search method is usually effective
Section 11.4 when
ref. KJB • The subspace iteration method used in ADINA contains all the
Section 11.6 capabilities of the original basic subspace iteration method
described in
• If q iteration vectors are used and there are only q mass degrees
ref. KJB of freedom, the subspace iteration method corresponds to a static
Sections 10.3.1 condensation (Guyan reduction) and subsequent solution of the
10.3.2 and
11.6.3 eigenvalues.
This analysis feature may be useful when you assign zero mass
ref. KJB • The Lanczos method can be used to calculate the smallest
Section 11.5 frequencies and corresponding mode shapes. The method is very
effective for large problems.
γ i( k −1) − γ i( k )
≤ ε; i = 1,..., N
γ i( k −1)
σ m
=
(σ (U ) − σ )
i
m
i
t
αi
i
( )
where σ i U im are the stresses computed from the perturbed
displacement vector U im = t U + α i φi , t σ is the stress vector at
time t, and α i is a perturbation factor, controlled by parameter
DUSIZE as follows.
DUSIZE
If DUSIZE is not equal to 0, then α i = ; this choice
φi
implies U − U = DUSIZE . Here the symbol
m
i
t
is the norm of
the enclosed vector only including translations. DUSIZE is seen to
be the size of the displacement perturbation.
If DUSIZE is equal to 0, then ADINA computes
t
U
α i = 0.001 t
where U is the total displacement vector at time t
φi
(last step of analysis before frequency analysis). This choice
implies U im − t U = 0.001 t U , which is usually quite
reasonable. In the special case in which the total displacement
vector is zero (for example, if the modal stress calculation is
requested at the start of the analysis), then ADINA sets αi to 0.001.
The choice α i = 0.001 can lead to very large or very small
0.001
displacement perturbations because then U im − t U = and,
φi
φT K φ
ρ (φ) = T
φ φ
• The ADINA sparse matrix solver is very reliable and robust and
is generally to be used for most problems in ADINA. The ADINA
sparse matrix solver is the default solver.
2) In the AUI, you should provide a memory limit for the sparse
solver using the following formula
r(k )
r (k )
≤ EPSIA and 2
≤ EPSIB
2 b 2
or
r(k ) ≤ EPSII
2
Note that EPSII should be less than or equal to EPSIA. The
b ≤ ε0 or r i ≤ ε0 or x i − x i−1 ≤ ε0
0.001σe i <3
or equation residual: r i rscale ≤
σe i≥3
0.001σv i <3
or solution residual: x i − x i−1 xscale ≤
σv i ≥3
x = 1 ∑ 1 xi
n
The solution norm is
n
In the above, ε0 ≡ 10−16 , the equation scale rscale = b , and the
variable scale
{
max ε , 1 x1 + x 2 + x 3
0 3( ) } for linear problems
xscale =
{
max ε , 1 x1 + x 2 + x 3 , ∆x
0 3( ) } for nonlinear problems
The equation residual (EQ) and the solution residual (VAR) are
available in the .out file.
t+∆t
where R is the vector of externally applied nodal loads at time
t+∆t
(load) step t+∆t, and F is the force vector equivalent (in the
virtual work sense) to the element stresses at time t+∆t.
τ
KU = t +∆t R − t F
t +∆t
U =tU + U
τ
where K is the tangent stiffness matrix at time τ, where τ ≤ t;
t+∆t
R is the externally applied load vector at time t+∆t; tF is the
consistent nodal force vector corresponding to the element stresses
t
due to the displacement vector U, and U is the incremental
displacement vector.
τ
Note that K is the stiffness matrix corresponding to the last
stiffness reformation. Stiffness reformations are performed only at
the solution steps that you specify.
Since equilibrium iteration is not performed, the solution in
general does not satisfy accurately nodal point equilibrium. Hence,
in practice this method should only be used with a sufficiently
small time step increment so that divergence from the equilibrium
solution is small. Whether nodal point equilibrium is sufficiently
satisfied can be approximately checked by performing equilibrium
iterations at the final solution step and noting how many iterations
are required.
τ
K ∆U (i ) = t +∆t R − t +∆t F (i −1)
t +∆t
U (i ) = t +∆t U (i −1) + ∆U (i )
(7.2-1a,b)
τ
K ∆U (i ) = t +∆t R − t +∆t F (i −1)
(7.2-1c,d)
t +∆t
U (i ) = t +∆t U (i −1) + β (i ) ∆U (i )
τ t+∆t
where K is the tangent stiffness matrix at time τ, where τ≤t; R
t+∆t (i-1)
is the externally applied load vector at time t + ∆t; F is the
consistent nodal force vector corresponding to the element stresses
+∆t (i-1) (i)
due to the displacement vector t U ; ∆U is the incremental
(i)
displacement vector in iteration (i) and β is an acceleration factor
obtained from line search.
τ
Note that K is the stiffness matrix corresponding to the last
stiffness reformation. Stiffness reformations are performed only at
the solution steps that you specify.
ref. KJB BFGS method: In the BFGS matrix update method, the following
Section 8.4.2 algorithm is used:
t +∆t
K *(i −1) ∆U ( i ) = t +∆t R − t +∆t F (i −1)
t +∆t
U (i ) = t +∆t U ( i −1) + β (i ) ∆U (i )
t+∆t *(i-1)
where K is an updated stiffness matrix (based on the
iteration history). The definitions of the other variables are as for
Eq. (7.2-1).
t+∆t *(i-1)
Note that K is not explicitly formed, but instead the
inverse of the stiffness matrix is updated using vector products to
provide a secant approximation to the stiffness matrix in successive
iterations.
BFGS matrix updates are performed only at the solution steps
t +∆t
K (i −1) ∆U ( i ) = t +∆t R − t +∆t F (i −1)
t +∆t
U (i ) = t +∆t U ( i −1) + ∆U ( i )
t +∆t
K (i −1) ∆U ( i ) = t +∆t R − t +∆t F (i −1)
t +∆t
U (i ) = t +∆t U ( i −1) + β (i ) ∆U (i )
t +∆t
U (i ) = t +∆t U (i −1) + β (i ) ∆U (i )
T
∆U (i ) t +∆t R − t +∆t F (i )
T
≤ STOL (7.2-2)
∆U (i ) t +∆t R − t +∆t F (i −1)
• Note that line search increases the computational time for each
iteration. Most of the extra time goes towards the evaluation of
t+∆t (i)
F in Equation (7.2-2). However, for the types of problems
where M is the mass matrix and δ is a mass scaling factor that can
vary from 0.0 to 1.0 (default 1.0), to partially account for the
dynamic inertia effect . The C matrix is evaluated using
C=βK
C =α M+ β K
• Note that mass effects may not be needed in the low speed
dynamic analysis. In this case, set the δ parameter of Eq. (7.2-4) to
zero, or, alternatively, set the material densities to zero. That way,
only structural damping effects will be present in the otherwise
static analysis.
• The user sets the division factor that ADINA uses to subdivide
the time step when there is no convergence. Successive
subdivisions can be performed, if necessary, provided that the time
step size is not smaller than a minimum value. This minimum value
is set as the original step size divided by a scaling factor provided
by the user.
• There are three options for controlling the time step size once
convergence is reached after ATS subdivisions. You can select any
of these options, or request that ADINA does the selection.
1. Use the time increment that gave convergence
In this case, the program continues the analysis using the last
converged time step. Once the end of the user-specified time
step is reached, the program may increase the time step based on
the iteration history. This option is the default in analyses
without contact. Corresponding command input: AUTOMATIC
TIME-STEPPING RESTORE=NO
2. Return to original time step size
In this case, the program continues the analysis using the
original user-provided time increment. This option is the default
when contact is present. Corresponding command input:
AUTOMATIC TIME-STEPPING RESTORE=YES
3. Proceed through user-defined time points
In this case, the program sets the time step size such that the
final time is that initially provided by the user. Hence, the
analysis always proceeds through all user-specified time steps.
Corresponding command input: AUTOMATIC TIME-
STEPPING RESTORE=ORIGINAL
The program automatically switches to option 3 if, when option 2
is specified, one of the following features is used: Bathe’s
composite method, substructuring, FSI, TMC, or mapping.
• The ATS method can also increase the time step beyond the
user-specified value if the iteration history shows that such an
increase is useful. This is only possible in a static analysis without
low-speed dynamics. The increase in time step cannot be larger
than a user-specified factor. Due to this increase, the analysis may
be completed in fewer time steps than requested. This time step
increase is only possible when the ATS subdivision is set to
“Return to original step size”.
·
0 0.5 1.0 1.5 t
obtained, the program will provide solutions for times t = 0.2 and t
= 0.4. If the time step of 0.2 is used again, then the maximum time
function value at time t = 0.5 will be missed. Therefore, the
program uses two substeps of value 0.1 to get to time t = 0.6.
∆t ∆t
t+ t+
2
U im − 2
U ih ≤ DISTOL
max
t+
∆t
2
U im =
(U t
i + t +∆tU i )
2
∆t
∆t t ∆t 2 t
( U i + t +∆tU i )
t+
2
U ih = U i +
t
Ui +
2 16
∆t
t+ ∆t
and 2
U ih is the displacement evaluated at time t + .
2
• Note that all static and dynamic nonlinear analyses, except those
only possible with the load-displacement-control (LDC) method,
can be carried out with the ATS method.
7.2.5 Total Load Application (TLA) method and Stabilized TLA (TLA-S)
method
Most of the TLA settings listed above can be modified by the user
(via the AUTOMATIC TOTAL-LOAD-APPLICATION
command).
- external forces
I EF = R i∆U
- damping forces
I DF = R D i∆U
- bolt forces
- inertia forces
I I = R I i∆U
I CD = R CD i∆U
- stiffness stabilization
(
I SST = K i∆U i∆U)
where K is the extracted stabilization part from the stiffness
matrix.
I DR = R DR i∆U
t +∆t
K (i −1) ∆U (i ) = ( t +∆t λ ( i −1) + ∆λ ( i ) ) R + R P − t +∆t F (i −1)
t +∆t
U ( i ) = t +∆t U (i −1) + ∆U (i ) (7.2-5)
f ( ∆λ (i ) , ∆U ( i ) ) = 0
where
t +∆t
K ( i −1) = tangent stiffness matrix at time t+∆t, end of
iteration (i-1)
R = constant reference load vector
R P = load vector from previous solution run
λ
t +∆t ( i −1)
= load scaling factor (used on R) at the end of
iteration (i-1) at time t+∆t
∆λ = increment in the load scaling factor in iteration
(i )
(i)
The quantities t +∆t F ( i −1) and ∆U ( i ) are as defined for Eq. (7.2-1).
Note that in Eq. (7.2-5), the equation f = 0 is used to constrain
the length of the load step. The constant spherical arc length
constraint method is usually used and the constant increment of
external work method is used if the arc length method has
difficulties to converge.
The reference load vector R is evaluated from all the
mechanical loads (except user-supplied and prescribed
displacement loadings).
• To start the LDC method, the load multiplier for the first step
(∆tλ), (used to obtain the corresponding load vector ∆tλ R) is
calculated using a user-specified prescribed displacement. This
prescribed displacement (see Fig. 7.2-1) acts in the specified
direction at the specified node. The direction of the displacement is
given by its sign.
Note that the input for the initial prescribed displacement (in
particular whether it is positive or negative) is critical in
establishing successive equilibrium positions using the LDC
method.
As an example, two entirely different solution paths will be
obtained for the same model shown in Fig. 7.2-1 if initial
displacements of different signs are prescribed for the first solution
step.
R
0
Y
a) Model considered
t
0 specified negative
b) Equilibrium paths
U 2 ≤ 100α ∆t
U
2
energy only,
translation/rotation only.
T
∆U (i ) t +∆t R − t +∆t F (i −1)
T
≤ ETOL (7.2-6)
∆U (1) t +∆t R − t F
t +∆t
R − t +∆t F (i −1)
2
≤ RTOL
RMNORM
∆U (i )
2
≤ DTOL
DNORM
∆U (i )
2
≤ DTOL
DMNORM
(
max R c(i −1) − R (ci − 2) , λ (i ) − λ (i −1)
2 2
) ≤ RCTOL
(
max R c(i − 2) , RCONSM
2
)
where R (ci −1) is the contact force vector at the end of iteration
( i − 1) , λ (i ) is the Lagrange multiplier vector at the end of iteration
i., RCONSM is a reference contact force level used to prevent
possible division by zero and RCTOL is a user-specified contact
force convergence tolerance.
T
∆U (i ) t +∆t λ (i −1) R − t +∆t F (i −1)
T
≤ ETOL
∆U (1) ∆λ (1) R
t +∆t
λ (i −1) R − t +∆t F ( i −1) 2
≤ RTOL
RNORM
t +∆t
λ (i −1) R − t +∆t F ( i −1) 2
≤ RTOL
RMNORM
• If you are not at all familiar with the nonlinear response of the
structural model considered, it may be advisable to first perform a
nonlinear analysis without equilibrium iterations, but using a
sufficiently large number of load steps. The calculated response is
then not necessarily very accurate, but at least you now have a
good idea of the nonlinear response of the model. The next
objective must be to ensure that the analysis results are close
enough to the accurate solution of the model. This can be achieved
by performing the analysis again with more load increments and by
the use of equilibrium iterations.
P
P
P
P
I N T E R M E D I A T E P R I N T O U T D U R I N G E Q U I L I B R I U M I T E R A T I O N S
ITE= 0 2.24E+03 2.32E+06 3.27E-03 4.79E-03 7.45E-01 4.74E+03 1.00E+00 2.32E+05 6.92E-02 1.15E+00
740-Z 319-Y 159-Z 239-Y 4.11E+03 3.27E-04 8.63E-01
-2.32E+06 -6.22E-04 -1.14E-03 1.64E-01
ITE= 1 3.95E+00 1.96E+04 9.93E+00 6.16E-04 2.03E-01 4.74E+03 1.77E-03 1.96E+03 8.90E-03 9.01E-01 9.99E-01 -5.95E-08
239-X 39-Y 239-X 279-Y 5.26E+03 9.93E-01 2.35E-01 ( 3)
4.14E+03 -2.37E+00 1.52E-04 6.71E-02
ITE= 2 2.67E-01 5.89E+03 4.45E+00 1.29E-04 5.98E-02 4.54E+02 1.19E-04 5.89E+02 1.86E-03 9.06E-02 4.84E-01 2.85E-06
740-Z 199-Y 42-X 279-Y 5.01E+03 4.45E-01 6.92E-02 ( 5)
-1.35E+03 -1.16E+00 -1.67E-05 -1.95E-02
ITE= 3 4.28E-03 1.00E+03 5.20E-01 2.45E-05 7.30E-03 4.46E+02 1.91E-06 1.00E+02 3.53E-04 9.67E-02 9.81E-01 -3.45E-05
39-Z 479-Y 39-X 79-Y 4.61E+03 5.20E-02 8.45E-03 ( 3)
-2.80E+02 1.72E-01 3.53E-06 -1.86E-03
ITE= 4 2.48E-06 1.77E+01 1.35E-02 3.47E-07 1.71E-04 3.12E+01 1.11E-09 1.77E+00 5.01E-06 6.80E-03 1.00E+00 8.78E-03
39-Z 239-Y 159-Z 119-Y 4.58E+03 1.35E-03 1.98E-04 ( 1)
-5.76E+00 -5.12E-03 -1.16E-07 -5.42E-05
ITE= 6 1.87E-12 5.22E-03 9.02E-06 4.84E-10 2.62E-07 2.64E-02 8.35E-16 5.22E-04 6.99E-09 5.76E-06 1.00E+00 -2.39E-02
119-Z 119-Y 159-Z 119-Y 4.58E+03 9.02E-07 3.03E-07 ( 1)
2.05E-03 -4.21E-06 -1.80E-10 -1.30E-07
843
7.2 Nonlinear static analysis
Chapter 7: Static and implicit dynamic analysis
I T E R A T I O N T I M E L O G
Row ITE=0: This row shows the result of the initial iteration
called iteration 0. For this iteration, the program performs the
following steps:
t +∆t
Compute U (0) = t U .
t +∆t t +∆t t +∆t
Compute F (0) and K (0) using U (0) .
Compute ∆U (0) using t +∆t K (0) ∆U (0) =t +∆t R − t +∆t F (0) . Only
considering translational degrees of freedom, the norm of the
CFORCE
= 1.15×100
max(CFNORM,RCONSM)
Note that the displacement and rotation norms are also substituted
into the displacement convergence criterion, which results in
∆U (0)
2
= 6.92×10-2 and
DNORM
∆U (0)
2
= 8.63×10-1.
DMNORM
Since DNORM and DMNORM are not provided by the user, they
are automatically estimated by the program. The above
displacement convergence values however are not used in
determining convergence.
Row ITE=1: This row shows the results of the first equilibrium
iteration. In this iteration, the program performs the following
steps:
t +∆t t +∆t
Compute U (1) = U (0) + ∆U (0) , t +∆t F (1) and the line
T
∆U (0) t +∆t R − t +∆t F (1)
search ratio T
. This ratio turns out to
∆U (0) t +∆t R − t +∆t F (0)
be greater than STOL=0.01, so line search is performed for 3
steps and ends up with a line search energy ratio of 5.95×10-8
which is less than STOL, corresponding to β (1) = 0.999 .
Compute t +∆t
U (1) = t +∆t
U (0) + β (1) ∆U (0) .
t +∆t t +∆t t +∆t
Compute F (1) using U (1) , and compute K (1) .
Note that the stiffness matrix is updated since the standard
Newton method is used.
Row ITE=2: This row shows the results from the second
equilibrium iteration. This row is interpreted exactly as is row
ITE=1. The line search factor in this case is 4.84×10-1 obtained in
5 line search iterations.
Row ITE=3: This row shows the results from the third equilibrium
iteration. This row is interpreted exactly as is row ITE=2.
Row ITE=4: This row shows the results from the fourth
equilibrium iteration. In this case, the previous increment of
displacement from the solver ∆U (3) satisfies the line search energy
tolerance STOL so no line search is performed.
t +∆t
∆U (4) = t +∆t ∆U (3) + ∆U (3)
Two of these four criteria (force and contact) are not satisfied, so
convergence is not satisfied.
Row ITE=5: The row shows the results for the fifth equilibrium
iteration. Contact convergence criterion is now also satisfied
leaving only force convergence unsatisfied. The solution
continues.
Row ITE=6: The row shows the results from the sixth equilibrium
iteration. In this case, all convergence criteria are satisfied and
convergence is reached.
t
U, t +∆t U = vectors of nodal point accelerations at time t, t+∆t
U = vector of nodal point displacement increments from
time t to time t+∆t, i.e., U = t+∆tU - tU.
t +∆t
U = t U + (1 − δ ) t U + δ t +∆t
U ∆t (7.3-2)
1
t +∆t
U = t U + t U ∆t + − α t U + α t +∆t
U ∆t 2 (7.3-3)
2
ˆ t +∆t U = t +∆t R
K ˆ (7.3-4)
where
ˆ =K+a M+aC
K (7.3-5)
0 1
ˆ =t +∆t R + M ( a t U + a t U + a t U ) + C ( a t U + a t U + a t U )
R 0 2 3 1 4 5
(7.3-6)
and where a0, a1, ... , a5 are integration constants for the Newmark
method (see Ref. KJB, Section 9.2.4).
A similar procedure can be followed for the Bathe time
integration scheme.
s
U = ∑ φi xi
i=r
xi + 2ξiωi xi + ωi2 xi = ri
where ξi is the critical damping ratio corresponding to the
frequency ωi , and ri = φiT R . The Newmark method (trapezoidal
rule) is used for the time integration.
p
∆R = R − ∑ ri ( M iΦi )
i =1
where
We then solve
K i∆U ( t ) = ∆R ( t )
C Rayleigh = α M + β K
• See Ref. KJB, Section 9.3.3, for information about selecting the
Rayleigh damping constants α, β. In the modal basis, the damping
ratio can be written as
α βωi
ξi = +
2ωi 2
Tp
periods smaller than Tp can be suppressed by choosing β = .
π
This may be of interest when using damping to suppress numerical
oscillations.
The above comments apply only when the stiffness matrix does
not change significantly during the analysis, however.
The above formula and comments do not apply when potential
based elements are used.
Objective
∆Ei
hi = (7.3-7)
Ei
where ∆Ei is the energy dissipated in the i-th mode, and Ei is the
total strain energy for that mode. In axisymmetric analysis, this
energy damping ratio provides the energy dissipated per unit
radian.
Ei = 12 φ Ti Kφ i = 12 ωi2 ; φ Ti Mφ i = 1 (7.3-8)
If the strain energy damping ratios are not uniform throughout the
structure, ADINA will allow each sub-structure or component
(represented as element group in ADINA) to have its own local
energy damping ratios
∆Ei | j
hi | j = (7.3-9)
Ei | j
where ∆Ei|j and Ei|j are the dissipated and total strain energy for
Ei | j = 12 φ Ti K j φ i
ADINA allows each element group to have its own input of local
energy damping ratios hi|j and then calculates the global energy
damping ratios using
∑E
j =1
i|j hi | j
hi = N
∑Ej =1
i|j
∑φ T
i K j φ i hi | j
1 N
∑φ
j =1
hi = = 2 T
i K j φ i hi | j (7.3-10)
N
ωi
∑φj =1
T
i K jφi j =1
Once the global strain energy damping ratios hi are available, the
modal damping ratio for each mode is set by ADINA as
hi
ξi = (7.3-11)
2
Notes
In the case of local strain energy damping, the AUI will provide
means for the user to apply the damping ratios to multiple element
groups simultaneously, instead of each element group separately.
δ
ξ= (7.3-12)
δ 2 + 4π 2
x0
δ = ln( )
x1
E1 E (1 − h) 2 π
δ = − 12 ln( ) = − 12 ln( 0 ) = −π ln(1 − h) (7.3-13)
E0 E0
AUI Input
( t +∆t
K (i −1) + c3c3M + c3C ) ∆U (i )
t + γ∆t t + γ∆t
= t +∆t R − t +∆t F (i −1) − M (c1 U + c2
t t
U + c3c1 U + c3c2 U
t + γ∆t
U (i −1) ) − C(c1 U + c2 U (i −1) )
t +∆t t t +∆t
c3c3 U + c3
where
c1 =
(1 − γ )
∆tγ
−1
c2 =
(1 − γ ) γ∆t
c3 =
(2 − γ )
(1 − γ ) ∆t
Evaluating velocities in terms of displacements and accelerations in
terms of velocities, we have
t +∆t t t + γ∆t t +∆t
U = c1 U + c2 U + c3 U
t +∆t t t + γ∆t t +∆t
U = c U + c2
1 U + c3 U
T
∆U (i ) t +∆t R − M t +∆t U (i −1) − C t +∆t U (i −1) − t +∆t F (i −1)
T
≤ ETOL
∆U (1) t +∆t R − M t +∆t U (0) − C t +∆t U (0) − t F
t +∆t
R − M t +∆t U (i −1) − C t +∆t U ( i −1) − t +∆t F (i −1)
2
≤ RTOL
RNORM
t +∆t
R − M t +∆t U (i −1) − C t +∆t U ( i −1) − t +∆t F (i −1)
2
≤ RTOL
RMNORM
• The implicit method can use much larger time steps since it is
ref. KJB
Section 9.2 unconditionally stable. However, it involves the assembly and
solution of a system of equations, and it is iterative. Therefore, the
computational time per load step is relatively high. The explicit
method uses much smaller time steps since it is conditionally
stable, meaning that the time step for the solution has to be less
than a certain critical time step, which depends on the smallest
element size and the material properties. However, it involves no
matrix solution and is non-iterative. Therefore, the computational
time per load step is relatively low.
• Note that the explicit solution provided in ADINA does not use
reduced integration with hour-glassing. This technique reduces the
computational time per load step. However, it can have detrimental
effect on the accuracy and reliability of the solution.
• Since the explicit time step size depends on the length of the
smallest element, one excessively small element will reduce the
stable time step for the whole model. Mass-scaling can be applied
to these small elements to increase their stable time step. The
implicit method is not sensitive to such small elements.
The .rto file can contain an arbitrary number of lines. Each line in
the .rto file is a string of up to 80 characters. A line beginning with
a * or # is a comment line. Any text following // is a comment.
Each line of the .rto file sets a parameter. Parameters are set using
syntax such as
DTOL=0.234
ETOL=1E-5
• ADINA reads the .rto file at the start of program execution, and
also before each equilibrium iteration. It is allowed to modify the
.rto file during program execution, for example to reset the
convergence tolerances during equilibrium iterations, or to save the
current solution, then stop. Use your favorite text editor to modify
the .rto file.
Rubber-like ***
Drucker-Prag.
Thermo-plast.
Nonlin.-elast.
Gurson-plast.
Thermo-orth.
Plastic-cyclic
Plastic-bilin.
Plastic-mult.
Elastic-orth.
Thermo-iso.
Plastic-orth.
Element type Nodes
Gasket**
Concrete
Ilyushin
Elastic
Fluid
Truss 2 9 9 9 9 9 9 9
Beam 2 9 9 9
Isobeam 2 9 9 9 9 9
Plate 3 9 9 9
3D-Shell 3/4 9 9 9
4/5/6/8
Solid (3D, displ.) 9 9 9 9 9 9 9 9 9 9 9 9 9
/10/27
4/5/6/8
Solid (3D, u/p) 9 9 9 9 9 9 9 9
/10/27
Disp.fluid 2-D 4
9
Disp.fluid 3-D 8
*) 3-node triangular 2-D and 4-node 3-D tetrahedral elements do not use incompatible modes.
**) No critical time step checking for gasket materials
***) The rubber-like materials supported in explicit analysis are the Mooney-Rivlin, Ogden and
Sussman-Bathe materials.
Table 8-3 lists the ADINA options not available in explicit analysis.
cyclic symmetry
substructuring
mapping
zooming
Thermo-mechanical coupling (TMC)
Fluid-structure interaction (FSI)
consistent mass matrix
consistent Rayleigh damping
temperatures provided via tape
load penetration (for ruptured elements)
8.1 Formulation
The central difference method (CDM) is used for time integration
in explicit analysis (see ref. KJB, Section 9.2.1). In this case, it is
assumed that
1 t −∆t
t
U=
∆t 2
( U − 2 t U + t +∆t U ) (8.1-1)
1
t
U=
2∆t
( −t −∆t U + t +∆t U ) (8.1-2)
M tU + C t U = tR − tF (8.1-3)
1 1 t +∆t 2 1 1 t −∆t
2 M+ C U = tR − tF + 2 M tU − 2 M + C U
∆t 2∆t ∆t ∆t 2 ∆t
(8.1-4)
t +∆t
from which we can solve for U.
ref. KJB • The central difference method has the following characteristics:
Sections 9.2.1,
9.4 and 9.5.1 It is an explicit integration method, meaning that equilibrium
of the finite element system is considered at time t to obtain the
solution at time t+∆t.
TNmin
∆t ≤ ∆tCR =
π
where ∆tCR is the critical time step size, and TNmin is the
smallest period in the finite element mesh.
8.1.2 Damping
C Rayleigh = α M
where M is the total lumped mass matrix. See Ref. KJB, Section
9.3.3, for information about selecting the Rayleigh damping
constant α.
8.2 Stability
ref. KJB • The stable time step for a single degree of freedom with central
Section 9.4.2 difference time integration is
TN 2
∆tCR = =
π ωN
TN min 2 2
∆t ≤ ∆tCR = = ≤
π ω N max ωE max
2
∆t = K × ∆t E min = K × (8.2-1)
ωE max
• For most element types the critical time step can be expressed in
terms of a characteristic length and a material wave speed
L
∆t E = (8.2-2)
c
where the definition of the length L and the wave speed c depend
on the element and material type. For all elastic-plastic materials
the elastic wave is used. This condition is used in the program
instead of actually evaluating the natural frequency in Eq. (8.2-1).
• Note that the critical time step calculated for all elements is only
an estimate. For some elements and material combinations it is
exact, and for others it is slightly conservative. However, it may
not be small enough for excessively distorted elements (3-D solid
and shells), and it will therefore need scaling using K factor in Eq
(8.2-1).
• The time step also changes with deformation, due to the change
in the geometry of the elements and the change in the wave speed
through the element (resulting from a change in the material
properties).
L
∆t E =
c
E
c=
ρ
L 12 I
∆t E = / 1+ 2
c AL
Shell elements
The critical time step for shell elements is
L
∆t E =
c
E
c=
ρ (1 − ν 2 )
The critical time step estimated here is only approximate, and may
be too large for excessively distorted shell elements.
L
∆t E =
c
E (1 − ν )
c=
ρ (1 + ν )(1 − 2ν )
The critical time step estimated here is only approximate, and may
be too large for excessively distorted 3-D solid elements.
Spring elements
The critical time step for a spring element is
2 M 1M 2
∆t E = =2
ωN K (M1 + M 2 )
where M1 and M2 are the masses of the two spring nodes and K is
its stiffness. Massless springs are not taken into account in the
calculation of the stable time step.
Rigid links
These elements are perfectly rigid and therefore do not affect the
stability of explicit analysis.
Note that the element having the smallest critical time step size is
always provided in the output file.
• The evaluation of the critical time step for each element takes
some computational time. Therefore, it does not need to be
performed at every time step. The parameter NCRSTEP in the
ANALYSIS DYNAMIC command determines how frequently the
critical time step is reevaluated.
• The time step size for explicit analysis can be unduly small for a
realistic solution time. Three features are provided to deal with this
problem.
.out file.
Mu + Cu + Ku = −Md k ak (t ) (9.1-1)
t
1 −ξiωi ( t −τ ) ω 1 − ξ 2 ( t − τ ) a (τ ) dτ
2 ∫
xik (t ) = −Γ ik e sin
i i
k
ωi 1 − ξi 0
(9.1-3)
or
xik (t ) = −Γ ik Dik (9.1-4)
u = − ∑ Γ ik Dik φi + u r (9.1-5)
ωi ≤ωu
σ=− ∑ΓDσ
ωi ≤ωu
k
i i
k
i + σr (9.1-6)
Γ ik = φT M d k
i
(9.1-7)
1) ∑
all modes
Γ ik φ = d k
i
(9.1-8)
∑ (Γ ) = (d ) k 2 T
2) i k M dk (9.1-9)
all modes
This formula states that the sum of the squares of the modal
participation factors for direction k equals the structural mass for
direction k, excluding mass associated with fixed degrees of
freedom of the structure. Hence the “modal mass” for mode i,
( )
2
direction k, is Γ ik .
( ui )max = Γ ik Sa(ωi , ξi ) φ i
(9.1-10)
1
( ui )max = Γ ik Sa (ωi , ξi ) φ (9.1-11)
ω i
2 i
1
( σi )max = Γ ik Sa (ωi , ξi ) σ (9.1-12)
ωi2 i
The maximum element forces and section results for mode i are
1
( Fi )max = Γ ik Sa (ωi , ξi ) Fi (9.1-13)
ω i
2
1
( R i )max = Γ ik Sa (ωi , ξi ) R i (9.1-14)
ω i
2
1
2
2
M SUM = ∑ (Mi ) (9.1-15)
ωi ≤ωu
M SUM = ∑
ωi ≤ωu
Mi (9.1-16)
1
2
= ∑ (Mi ) + 2 ∑ MiM j
2
M SUM (9.1-17)
ωi ≤ωu ωi ≤ωu
ω j ≤ωu
ω j − ωi
≤ 0.1
ωi
where
−1
ω ′ − ω ′ 2
ε rs = 1 + r s
ξ r′ωr + ξ s′ωs
and
1
ωr′ = ωr (1 − ξ r2 ) 2
2
ξ r′ = xr +
t d ωr
M SUM = ∑M
ωi ≤ωu
i (9.1-19)
• Note: The results obtained with all methods except the algebraic
sum method and DSC method depend upon the coordinate system
used in modeling the problem. This is because formulas (9.1-15) to
(9.1-18) and (9.1-20) are applied component by component and
because the sign of each component is lost (either the component is
squared or the absolute value of the component is taken).
For example, to evaluate the SRSS displacement at a node, (9.1-
15) is applied to the maximum x displacement, then to the
maximum y displacement and finally to the maximum z
displacement. Figure 9.1-1 shows the results for a two-dimensional
problem. As shown, the expected result occurs only if one
coordinate system direction is aligned with the response direction.
Although the figure illustrates only displacements, the same
phenomenon occurs for all results; accelerations, velocities,
displacements, reactions, stresses and element forces.
This observation also applies to accelerations, velocities and
displacements when skew systems are used at nodes.
Residual accelerations:
u r = d k − ∑ Γ ik φi S ZPA (9.1-21)
ωi ≤ωu
l (u1)MAX = (u1)SRSS
u
(u1)MAX =
0
u2 u
(u1)SRSS = =
02 0
Residual displacements:
u r = K −1M d k − ∑ Γ ik φi S ZPA (9.1-22)
ωi ≤ωu
M = ( M SUM ) + ( M r ) 2
2 2
(9.1-23)
M = ( M SUM ) + M r
where
M r if M SUM ≥ 0
SIGN ( M r , M SUM ) =
− M r if M SUM < 0
M SPATIAL = ∑ α k M k (9.1-26)
k
where
M SPATIAL if M ST ≥ 0
SIGN ( M SPATIAL , M ST ) = (9.1-28)
− M SPATIAL if M ST < 0
N
u kR = d k − ∑ Γ ik φi (9.1-29)
i =1
u kR = t0 K −1M u kR (9.1-30)
The definitions for the variables on the right side of the equation
t
are given above and 0 K is the stiffness matrix at the start time
of the solution.
FRk =M u kR (9.1-31)
s kR = t0 C t0 B u kR (9.1-32)
t t
and 0 C, 0 B are the element stress-strain matrix and strain-
displacement matrix, respectively at the start time of the
solution.
If the nodal force output option is used in an element, the
residual forces corresponding to a displacement vector equal to
u kR are instead calculated, printed and stored.
In the AUI, all of the information needed for the calculation for
one ground motion direction is entered into a response of type
response-spectrum.
To perform spatial combination calculations for more than one
ground motion direction, define a response of type response-
combination and include the response spectrum responses for the
individual ground motion directions in the definition of the
response-combination response.
∞
f ( t ) = a0 + ∑ ( an cos nω1t + bn sin nω1t )
n =1
where
2π
ω1 = , t p = t1 − t0
tp
t1
1
a0 =
tp ∫ f ( t ) dt
t0
t1 t1
2 2
an = ∫ f ( t ) cos nω1t dt , bn = ∫ f ( t ) sin nω t dt
1
tp t0
tp t0
Since
bn
cn = an2 + bn2 , φn = tan −1
an
t1
1 1 ∞ 2
∫ f ( t ) dt = c02 + ∑ cn
2
tp t0
2 n =1
nω1
• The power spectral density of f(t) for frequency fn = may
2π
be defined as
c02
S f ( f0 ) =
∆f
cn2
S f ( fn ) = , n = 1, 2,...
2∆f
ω1
where ∆f = is the interval between two successive frequencies
2π
t1 ∞
1
∫f
2
dt =∑ S f ( f n ) ∆f
tp t0 n =1
It is seen that the mean square value of f(t) is related to the area
under the power spectral density curve.
( )
Hence, N = int t p f max gives the upper limit of the Fourier series
summation
N
f ( t ) = c0 + ∑ cn cos ( nω1t − φn )
n =1
N
Mu + Cu + Ku = ∑ Fk ( t ) (9.3-1)
k =1
N
xi + 2ξiωi xi + ωi2 xi = ∑ Γ ik bk (t ) (9.3-2)
k =1
Fk ( t ) = Fk 0 bk ( t ) (9.3-3)
Fk ( t ) = −Md k bk ( t ) (9.3-4)
Harmonic vibration analysis with all loads acting with the same
loading frequency
N
xi ( t ) = ∑ ( Aki sin (ωt − α k ) − Bki cos (ωt − α k ) ) (9.3-7)
k =1
where
2
ω
1−
b Γ ωi
k
Aki = k0 2 i (9.3-8)
ωi 2 2 2
ω ω
1 − + 2 ξi
ωi ωi
ω
2 ξ
bk0 Γ k
ωi i
Bki = i
(9.3-9)
ωi2 ω ω
2 2 2
1 − + 2 ξi
ωi ωi
n N
u ( t ) = ∑ φ( ) ∑ xki ( t )
i
(9.3-10)
i =1 k =1
n N
uL( t ) = ∑ φL ∑ x (t )
(i )
ki
i =1 k =1
n (i) N n (i ) N
= ∑ φL ∑ Cki sin ωt − ∑ φL ∑ Dki cos ωt
i =1 k =1 i =1 k =1
(9.3-11)
where
2 2
n N
n N
uL0 = ∑ φL(i ) ∑ Cki + ∑ φL( i ) ∑ Dki (9.3-13)
i =1 k =1 i =1 k =1
n (i ) N
∑ φL ∑ Dki
θ L = tan i =n1
−1 k =1
N
(9.3-14)
φ(i ) C
∑
i =1
L ∑
k =1
ki
π
uL( t ) = ω uL0 sin ωt − θ L − (9.3-15)
2
and
u L( t ) = ω 2 u L0 sin (ωt − (θ L − π ) ) (9.3-16)
from which the maximum amplitude and phase are easily seen.
• For more insight into the above equations, we consider only one
load case (N=1) and draw (9.3-6) and (9.3-12) in the complex
plane (Fig. 9.3-1). Notice that the actual response is the imaginary
component of the vectors.
Im
b
uL b0 sin0t - )
0t - uL0 sin0t - +)
0t - + Re
Length of b is b0 .
Length of u L is uL0 .
Im b
uL
0t - 34
uL0 sin0t - +)
0t - + Re
Im
0t - 4
b Re
0t - +) uL0 sin0t - +)
uL
The reason why we use 90o to obtain the real part of the
response is because equations (9.3-6) and (9.3-12) are written in
terms of the sine function, not the cosine function.
Then you can obtain either the entire structural response for a
single forcing frequency or the response at a point for a range of
forcing frequencies.
T
1
∫ p (t )
2
p 2 = lim dt (9.4-1)
T →∞ 2T
−T
1 p (ω )
2
T
T →∞ 2π ∫
p = lim
2
dω (9.4-2)
0
T
in which
∞
p (ω ) = ∫ p ( t ) exp ( −iω ) dt (9.4-3)
−∞
p (ω ) = p (ω ) p*(ω )
2
(9.4-4)
y (ω )
2
yˆ (ω ) = (9.4-5)
T
as follows:
∞ ∞
1
y (t ) = yˆ (ω ) d ω = ∫ yˆ ( f ) df
2π ∫0
2
(9.4-6)
0
N
xi (ω ) = ∑ Γ ik H i (ω ) bk (ω ) (9.4-7)
k =1
where
1
H i (ω ) = (9.4-8)
(ω − ω ) + i ( 2ξi ωi ω )
i
2 2
n N
uL(ω ) = ∑ φL( ) ∑ Γ ik H i (ω ) bk (ω )
i
(9.4-9)
i =1 k =1
and therefore
uL(ω ) =
2
n n N N
∑∑∑∑ φ( ) φ( ) Γ
i =1 j =1 k =1 l =1
L
i
L
j
i
k
Γ lj H i (ω ) H *j (ω ) bk (ω ) bl*(ω )
(9.4-10)
min (ωi , ω j )
H i (ω ) H *j (ω ) = 0 for < CUTOFF
max (ωi , ω j )
(9.4-11)
and
2
bk (ω ) bl*(ω ) ≈ bk (ω ) δ kl (9.4-12)
( ) ( ) ) ( Γ )( Γ ) H (ω ) H (ω ) b (ω )
N n n 2
u L(ω ) = ∑∑∑ φL( ) φL(
2 i j k k *
i j i j k
k =1 i =1 j =1
(9.4-13)
Using the definition of the power-spectral-density, we write
( )( )
N n n
uˆ L ( ω) = ∑∑∑ φ(L ) φ(L ) ( Γ ik )( Γ kj ) H i ( ω) H *j ( ω) bˆk ( ω)
i j
k =1 i =1 j =1
(9.4-14)
and therefore
( )( )
N n n
uL2 ( ω) = ∑∑∑ φ(L ) φ(L ) ( Γik )( Γ kj ) ∫ H i ( f ) H *j ( f ) bˆk ( f ) df
i j
k =1 i =1 j =1 0
(9.4-15)
∞
bk ( t ) = ∫ bˆk (ω ) dω ,
2
0
∞
bk ( t ) = ∫ bˆ ( f ) df ,
2
k
−∞
∞
bk ( t ) = ∫ bˆ (ω ) dω
2
k
−∞
iω
Velocity : H i (ω ) =
(ω − ω ) + i ( 2ξiωiω )
i
2 2
−ω 2
Acceleration : H i (ω ) = 2
(ωi − ω 2 ) + i ( 2ξiωiω )
Interpretation of mean-square responses: The mean-square
value of a response quantity can be used to predict the probability
that the actual response will exceed any given value. This is done
by assuming that the response has a Gaussian distribution and has
zero mean value. Then it may be shown that the standard deviation
is well approximated by the root-mean-square (RMS) value of the
response quantity:
σ = uL2 (9.4-17)
Hence we can write
λ
Pr ( uL > λσ ) = erfc (9.4-18)
2
Pr ( uL > σ ) = 0.3173
Pr ( uL > 2σ ) = 0.0455
Pr ( uL > 3σ ) = 0.0027
Then you can obtain either the RMS response or the power-
spectral-density of the response at a point.
m x(t)
c
uc(t) k uc(t)
..Ground acceleration
xg(t) ..
xg(t) input to ADINA as a mass-
proportional load
uc(t) calculated by ADINA
..
x(t)|max calculated by the AUI
Frequency
Figure 9.5-2: Floor response spectrum analysis from response
spectrum data
Maximum
response Damping = 2%
Damping = 5%
ref. 1992 ASME Boiler and Pressure Vessel Code, Rules for
Construction of Nuclear Power Plant Components,
Division 1, Appendix N.
Response
Broadened The floor response spectrum
curve commands construct a roof
Slope of roof for each peak on the
is maximum slope unbroadened curve. Set the
of unbroadened delta value in the frequency
curve near peak. table.
The floor response spectrum
Unbroadened commands raise points on
curve unbroadened curve to roof
for the peak.
ωi , φi , Γ ik , ξi
ωi → ωi* , φi → φi* ,
Γ ik → ( Γ ik ) , ξi → ξi*
*
(x )
k
0 max = modal-combination of ( x0ki )
max
, i = 0,..., N (9.5-2)
where
1
N
2 2
(x ) = ∑ ( Γ ik ) φi*0 S a (ωi* , ξi* )
k *
(9.5-4)
0 max
i =0
( ) ,φ
*
We see from eq. (9.5-3) that we need Γ ik *
i0 , i = 1,..., N and
(Γ ) φ
k *
0
*
00 to apply the response spectrum method to the structure
plus SDOF system mathematical model. From the last reference
above, if the SDOF system is very light compared with the
structure,
φ
φi ≈
* i
(9.5-5a)
*
φi 0
Also,
1
φi*0 ≈ 2
φia (9.5-6)
ω
1− i
ω0
n
( Γ 0k ) φ00* = δ kl − ∑ ( Γ ik ) φi*0
* *
(9.5-7)
i =1
N
(Γ ) φ
k *
≈ δ kl − ∑ ( Γ ik ) φi*0
* *
0 00 (9.5-8)
i =1
by neglecting
n n
1
∑ (Γ ) φ i
k * *
i0 = ∑Γ i
k
2
φia (9.5-9)
i = N +1 i = N +1 ω
1− i
ω0
(x )
k
0 max = modal-combination of ( x0ki )
max
and ( x0k j )
max
,
i = 1,..., N , i ≠ j
(9.5-10)
where
(Γ ) φ
k * *
S a (ω *j , ξ *j )
(x )
k
=
j ja
(9.5-11)
2 ξ 0 (ξ 0 + ξ j )
0 j max
2
ω
1 + 2ξ eff i
ω0
φi*0 ≈ ± 2
φia (9.5-12)
ω 2 ω
2
1 − i + 2ξ eff i
ω0 ω0
ξ eff = ξ 0 (9.5-12a)
ξ eff = ξ 0 (ξ 0 + ξ j ) (9.5-12b)
• Once the responses are known for a single load case, the
response for a combination of load cases can be computed as
follows:
For example, when the SRSS method is used to combine three load
cases, each load case corresponding to ground motions in one of
the three coordinate directions,
1
3 2 2
( x0 )max = ∑ ( x0k ) (9.5-14)
k =1
• Note on the use of the above theory with the CQC method of
modal combinations. When the above theory is used, along with
equation (9.5-12), and along with the CQC method, the resonances
tend to be underestimated. This is because, in the CQC method,
the magnitude of the predicted resonance depends on the
cancellation of the quantities M r , where M r is the maximum
modal response component for mode r. To be specific, as two
modes r and s approach each other, the contribution of these modes
to the CQC sum approaches ( M r + M s ) . When equation (9.5-12)
2
ref. H.G. deLorenzi, 'On the energy release rate and the J-
integral for 3-D crack configurations', Int. J. Fract., Vol.
19, pp. 183-193, 1982.
Crack front
z
y
x
Plane of self-similar
Crack front crack advance
dΠ
G=− (10.3-1)
dA
Π =U − ∑ Ru (10.3-2)
dΠ
• The operation is interpreted as follows:
dA
dΠ Π −Π A
= lim A+∆A (10.3-3)
dA ∆A→0 ∆A
that is, the difference of the total potential energy for two slightly
different crack areas divided by the change in crack area.
∂u
J = ∫ Wdx2 − τ ij n j i ds (10.4-1)
Γ ∂x1
ds
K I2 K II2 K III2
J= + + (10.4-2)
E′ E′ G
E
with E ′ = E in plane stress and E ′ = in plane strain. In
1 −ν 2
this equation, G is the shear modulus, not the energy release rate.
Isothermal analysis
∂ui 0
J= ∫ W dx2 − 0T ji 0 n j
0
0 ds (10.4-3)
0
Γ
∂ 0 x1
∫
0
T
0 ji 0 n j 0ds = 0 (10.4-5)
Γ
1 ∂u j ∂∆X k
G1 =
∆Ac ∫
V
τ ij
∂xk
− W δ ik
∂x j
dV (10.5-1)
III III
II II
I I
x1 x1 x1
Crack front
X’
t X
ds
1 ∂u
Ghoop =
∆AC ∫V
∆X 2 τ xx x − W dV
∂x
(10.5-2)
1 ∂eijth ∂θ
Gth =
∆AC ∫V
τ ij
∂θ ∂xk
∆X k dV (10.5-3)
in which eijth are the components of the thermal strain tensor and θ
is the temperature. Note that, for those materials in which the
material constants are temperature-dependent, the effect of material
constants changing with temperature is not included. (However,
the effect of thermal coefficient changing with temperature is
included.)
1 ∂ui
G pr = −
∆Ac ∫tS
i
∂x j
∆X j dS (10.5-4)
1 ∂ui
Gdyn =
∆Ac ∫V
ρ ui
∂x j
∆X j dV (10.5-5)
∆X k is sufficiently small.
No body forces.
Crack front
z
y
Crack front
Crack front
or
CRACK-PROPAGATION POINT
pointi
• In the above cases, only one data input line of the CRACK-
PROPAGATION command is used.
or
CRACK-PROPAGATION LINE
linei front-pointi
• In the above cases, only one data input line of the CRACK-
PROPAGATION command is used.
Plane of self-similar
crack advance
Crack front
a) Free meshing
Generator line 2
or
CRACK-PROPAGATION POINT
pointi
Here there is one data input line per node or point on the crack
front. It is evident that the CRACK-PROPAGATION POINT
command is not very useful for 3-D meshes since there is most
likely no geometry point at each node on the crack front.
Here there is one data input line per generator line, with the first
data input line corresponding to the first generator line, the second
data input line corresponding to the second generator line, etc.
node-1i is the node on the crack front for generator line i,
node-2i is the next corner node along the generator line, etc.
Generator lines for mid-side and mid-face nodes should not be
specified, that is, only corner nodes define generator lines.
CRACK-PROPAGATION SURFACE
surfacei front-linei
Here there is one data input line per geometry line on the crack
front.
direction for the crack front node and for each node on the plane of
self-similar crack advance.
Generator line 1
B3
Crack front
Crack sleeve
B2 Remainder
B1
B5
B6
B3
B1
B4
B2
B5
B3
B1
B2
B4
B6
B9
B11
B5
B1
B7 B2
B3 B6
B4
B8
B10
B12
In all cases, the model is decomposed into three parts, the crack
front, the crack sleeve and the remainder.
The crack front is that region of the model immediately adjacent
to the crack front edge. This part of the model will be meshed using
mapped meshing, as discussed below.
The crack sleeve is a transition region between the crack front
and the remainder of the model. The crack sleeve encloses the
crack front, and the crack sleeve should include the cracked surface
not represented in the crack front itself. This part of the model will
be meshed using free-form meshing.
The remainder of the model is constructed as usual, and is
meshed using free-form meshing.
Each of these parts is represented with one or more geometry
bodies, as shown. Due to the geometric modeler requirements, two
bodies must be used to represent the closed loop crack front in Fig
10.6-5(b). Also, two bodies must be used to represent the enclosed
crack front in Fig 10.6-6(a), and four bodies must be used to
represent the enclosed closed loop crack front in Fig 10.6-6(b). For
the remainder of the model, up to four bodies are used also, but
these bodies could be merged into a single body, provided that the
faces adjacent to the crack sleeve body remain compatible with the
crack sleeve body faces.
Although a circular crack front is shown in the Figures, the
CRACK-M features can be used whenever the geometry is
topologically equivalent to the Figures.
Crack front body: A typical body used to represent the crack front
is shown in Fig. 10.6-7. The body is prismatic, with one edge
corresponding to the crack front edge, and one face corresponding
to the cracked surface. Any geometric modeler tools can be used to
create the crack front body, provided that the body be topologically
equivalent to the body shown in Fig 10.6-7.
Crack sleeve body: Requirements on the crack sleeve body are: the
crack sleeve body should completely surround the crack front
body, the crack sleeve body should have one face on the cracked
surface.
Again, any geometric modeler tools can be used to create the crack
sleeve body.
CSBODY1
CFBODY1
CSFACE1
CFFACE1 CFEDGE1
CSBODY1
CFBODY1
CSFACE1
CFFACE1 CFEDGE1
CFFACE2 CFEDGE2
CSFACE2
CFBODY2
CSBODY2
NDIVTAN1
divisions
NDIVRAD
divisions
NDIVTAN1
divisions
NDIVTAN2
divisions
a) Crack on boundary
1 single node
1/4 L 3/4 L
1/4 L 3/4 L
a) 1/&r singularity for linear elastic analysis
3 coincident
untied nodes
1/4 L 3/4 L
1/4 L 3/4 L
Crack tip
z
y
a) Physical problem
Line contour
Integration
point
4 4 3 4 4
3 3 3
1 2 1 2 1 1
2 2
Path number 1 Path number 2 Path number 1 Path number 2
4 3 4 3 4 3
4 3
1 1
1 2 1 2 2 2
Path number 3 Path number 4 Path number 3 Path number 4
4 4 3 4 3
3 4 3
1 2 1 2 1 2 1 2
Path number 14 Path number 21 Path number 14 Path number 21
4
3 4 3 4 3 4 3
1 2 1 1 2 1 2
2
Path number 32 Path number 43 Path number 32 Path number 43
J-LINE ELEMENT
elementi groupi
• Corner nodes in the virtual shift are shifted by the full value of
the virtual shift vector. Mid-side nodes are shifted by vectors
interpolated from the corresponding corner nodes, so that a mid-
side node between two corner nodes in the virtual shift is shifted by
the full value of the virtual shift vector, and a mid-side node
Crack front
Nodes subjected to a virtual shift
10.8.1 Domain of the virtual shift, CRACK-M features are not used
J-VIRTUAL-SHIFT NODE
nodei
Zone III
Zone II
Zone I
J-VIRTUAL-SHIFT ELEMENT
elementi groupi
J-VIRTUAL-SHIFT LINE
linei
and
J-VIRTUAL-SHIFT SURFACE
surfacei
Point
Radius
Figure 10.8-4: Description of a virtual shift by point and radius
Elements in rings 1 to 3
Ring origin
The origin of the ring can either be a single node or a set of nodes.
These origin nodes are selected by parameters RING-TYPE and the
data input lines:
Disk used to select virtual shift nodes in 3-D analysis: For 3-D
analysis, when the virtual shift is defined by POINT or RING, the
virtual shift nodes are selected in two stages. In the first stage,
nodes are selected as described above. In the second stage, the
nodes selected by the first stage are tested against a disk, and only
the nodes that lie within the given disk are selected. There are
additional parameters in the J-VIRTUAL-SHIFT commands for
defining the disk:
J-VIRTUAL-SHIFT … ,
N3DSH NORMAL NX NY NZ THICKNESS,
POINT RADIUS
Node on the central plane. This is taken as the node on the crack
front at generator line N3DSH. If N3DSH=0, this node is found by
taking all of the nodes in stage 1, then finding the first node that
lies at geometry point POINT; if POINT=0, then finding the first
node that lies on the crack front. (N3DSH is not used when RING-
TYPE=AUTOMATIC.)
THICKNESS
NORMAL
Generator line 1
RADIUS
Crack front node
at generator line
N3DSH
Figure 10.8-6: Definition of a disk used to select nodes in 3-D virtual shift
Disk
THICKNESS
RADIUS
• It is clear that the disk option can be used only when there is a
mapped mesh on the plane of self-similar crack advance, and when
this mesh is specified using the CRACK-PROPAGATION
command, as in Section 10.6.3.2.
For 2-D models, this option can be used only when there are nodes
on the line of self-similar crack advance and when these nodes are
specified in the CRACK-PROPAGATION command. The virtual
shift vector lies on the line of self-similar crack advance and has
magnitude of 1/100 the distance between the crack tip and the next
corner node along this line.
For 3-D models, this option can be used only when there is a
mapped mesh on the plane of self-similar crack advance, and when
the nodes on this mesh are specified in the CRACK-
PROPAGATION command. The virtual shift vector has direction
given by generator line N3DSH, and has magnitude of 1/100 the
distance between the crack front node and the next corner node
along this generator line. (See above for description of parameter
N3DSH.)
Fig. 10.8-8 shows typical virtual shifts for station IDIVCF. Fig
10.8-8(a) shows virtual shift IDIVCF and Fig 10.8-8(b) shows
virtual shift NDIVCF+IDIVCF.
Nodes at IDIVCF+1
Nodes at IDIVCF
Nodes at IDIVCF-1
The command
can be used to create a listing that gives, for each virtual shift, the
crack front node, ring number, number of nodes and virtual shift
vector. (The ring number is the number of nodal rings contained in
the virtual shift.)
d
F
Initial crack tip
b) V = CMOD.
a) Some definitions of the CTOD. = load point displacement
L = initial distance between
load points
Crack front
Crack front
Crack growth
control parameter
Current value
of crack growth
control parameter
Crack advance
a) Temperature independent
Interpolated resistance
curve for temperature +
Current value
of crack growth
control parameter
Crack growth
control parameter
Crack advance
a) No change of crack advance
Crack growth
control parameter
Crack advance
b) Change of crack advance -- node shift/release
Crack growth
control parameter
Crack advance
Input specification:
FRACTURE TYPE=PROPAGATION
CRACK-PROPAGATION LINE
linei front-pointi nvshfti
Each data input line specifies the (x,y) points for the resistance
curve for a given temperature. MPOINT is the number of points in
each resistance curve. In this command, “x” is the crack advance
and “y” is the crack growth control parameter. If there are no
temperature-dependent materials, then only one curve need be
entered and thetai is not used. When there are temperature-
dependent materials, then at least two curves need to be entered.
The values of “x” must be the same for each resistance curve.
J-PARAMETER_1: (obsolete)
TEMPERATURE_CORRECTION,
PRESSURE_CORRECTION,
DYNAMIC_CORRECTION: The corrections Gth , G pr , Gdyn
Structure considered
l l
l l
l l
Main
structure
l l
l l
Substructure without
internal masses
l l
substructure 2 substructure 1
substructure 3 substructure 4
substructure 2
substructure 3
substructure 4
reuse
reuse
Z
B
U
B B 3
B - X Y
B
Z
contact
X Y
master I master II
substructure I
substructure II
substructure I
Y
X
substructure II
• Fig. 11.1-9 shows a long columnar block built from one AUI
geometry volume and 10 reuses of the meshed volume. Due to the
limitations of the current analysis, it is required that dummy shell
elements (Youngs modulus=0.001) are defined on the top/bottom
surfaces in order to define boundary conditions and loadings. The
input steps of Fig. 11.1-9 are as follows:
REUSE …
B B
B B B B
B B
U1 U2 U3
B - - -
substructure
whole structure master structure with 10 reuses substructure - reuse 1
X Y substructure-reuse 2
B
B
B
main structure
6 dofs
5 dofs
6 dofs
5 dofs
main structure nodes
6 dofs
substructure-reuse 1
5 dofs
U U U
1 2 3 1 2 3
B - - - - - -
main structure
Fundamental
Cyclic
cyclic part
boundary 2
Cyclic
boundary 1
o
360
-
N
Cyclic
symmetry axis
• Note that “solid” cyclic structures with nodes along the cyclic
symmetry axis can also be modeled.
Z p Z
Gravity
p1 p
Side 2
Side 2
F2 F F
F1
Side 1 Side 1
Y p Y
F F
X X F p
F3
p2
• In cyclic frequency analysis, the user can specify either the total
number of frequencies (default) or the number of frequencies per
harmonic. If the total number is specified, ADINA will internally
determine how many frequencies to solve per harmonic. A
conservative (large) value will be used to ensure that no frequency
is skipped.
If both element birth and death options are used, the element
is added to the total system of finite elements at the time of birth
and remains active until the time of death. The time of death
must be greater than the time of birth. The element is taken out
of the total system of finite elements at all times larger than the
time of death.
Installation of Column to be
temporary support replaced
Ground level
l
l l
l l
l l l
l
l
l l
l
Tunnel to be
excavated
Figure 11.4-1: Analyses that require the element birth and death
options
element for which the birth option is active. Note that if TBIRTH
is input for the range shown (where TBIRTH > tb and TBIRTH is ≤
tb + ∆t), then the stress-free configuration of the element is at time
tb, and the element is first active at time tb + ∆t. Note that the
stress in the element is based on the displacements measured from
the configuration at time tb, irrespective of the input value of
TBIRTH, provided that TBIRTH is input in the range shown in
Fig. 11.4-2(a). See also the example given below.
Time
0 t t+t
Stress free First solution time for which
state the element is active
(a) Birth option active
Time
0 t-t t
First solution time for which
the element is inactive
(b) Death option active
tR
t F int = t R
1
Element 1:
K1=AE/ 0L
t int
F1 =K1(tL - 0L)
tL
b) Solution at time t
after its addition to the system. Also note that the stiffness of
t
element 2 is based on the length L.
At time t + ∆t, the force in element 2 is determined based on its
deformation with respect to its stress-free state, see Fig. 11.4-3(d).
Hence, the total increment in displacement from time t to time t +
∆t determines the force in element 2. Identically, the same solution
would be obtained using any value for TBIRTH which satisfies the
relation TBIRTH > t and TBIRTH ≤ t + ∆t.
tF =0
Element 2
tL
tR
Element 1
tL
Element 2:
K2=AE/ tL
t+t int
F2 =K2(t+tL - tL)
t+tR
Element 1:
K1=AE/ 0L
t+t int
F1 =K1(t+tL - 0L)
t+tL
Table 11.5-1: Elements and material models that include "death upon rupture"
Truss 2-D 3-D Beam Iso- Shell Pipe
solid solid beam
Plastic-bilinear, plastic-multilinear 9 9 9 - 9 9 9
Mroz-bilinear - 9 9 - - - -
Thermo-plastic, creep, plastic-creep,
multilinear-plastic-creep, creep- 9 9 9 - 9 9 9
variable, plastic-creep-variable,
multilinear-plastic-creep-variable
Plastic-orthotropic - 9 9 - - 9 -
Plastic-cyclic 9 9 9 9 - 9 -
Moment-curvature - - - 9 - - -
User-coded - 9 9 - - - -
• The initial strain option is available for all solid and structural
elements, but only for nonlinear element groups. If you apply the
initial strain option to an element group that would otherwise be
linear, the element group is automatically classified as nonlinear.
Case 1 Case 2
Z
Initial configuration
Case 1 Z Case 2
Y
Initial configuration (no
applied or body forces)
• For the truss, beam, iso-beam, plate, shell and pipe elements, the
element initial strains are specified as element data (possibly
assigned to the geometry in the AUI).
• For the 2-D and 3-D solid elements, the element initial strains
are specified using strain field coefficients A, B, C, D, E, F and the
following formulas:
X = ( I + ε0 )
t0 −1
0
t
in which ε 0 is the initial strain tensor and 00 X is the initial
deformation gradient. Then a multiplicative decomposition of the
deformation gradient is used at every step, i.e.:
t
0 X = t0t X t00 X
t
• The components of t0 X are output when deformation gradients
are requested to be output.
• An initial stress option can be used with all element types. The
following material models can be used: elastic-isotropic, elastic-
orthotropic, thermo-elastic, thermo-orthotropic, Drucker-
Prager, Cam-clay, Mohr-Coulomb.
The initial stress capability allows initial stresses corresponding
to an unknown loading to be input. These stresses can be provided
as nodal point initial stresses or element initial stresses.
Whenever the constitutive relation is used, the initial stresses
are added to the current stresses caused by externally applied
loading. The initial stresses are included in the printed and saved
stresses.
t
F= ∫
t
BT ( t τ + τ I ) d tV (11.6-1)
t
V
RI = FI = ∫
0
BT τ I d 0V
0
V
K(
i −1)
∆U ( ) = t +∆t R − t +∆t F (
t +∆t i i −1)
+ RI
t +∆t
in which R is the externally applied load vector.
• Please note:
Restrictions
integration analysis.
• The basic formula for the strain energy density under small
displacement and small strain conditions is
dW = τ ij deij (11.7-1)
where τ ij is the Cauchy stress tensor and eij is the small strain
tensor.
dW = τ ij deijm (11.7-2)
• The basic formula for the strain energy density under large
displacement and small strain conditions is
• The basic formula for the strain energy density under large
displacement and large strain conditions, using the ULH
formulation, is
• The basic formula for the strain energy density under large
displacement and large strain conditions, using the ULJ
formulation, is
Numerical evaluation
1
W = tW + ( tτ ij + t +∆tτ ij )( t +∆t eij − t eij )
t +∆t
(11.7-6)
2
• The AUI can combine the element group quantities over the
whole model or over a selected zone in the model. The AUI also
includes the concentrated masses in the calculations. The moments
and products of inertia are calculated and printed with respect to
the center of mass of the whole model or zone.
11.9 Restart
• Restart is a useful feature in the ADINA System. It can be used
when the user wishes to continue an analysis beyond its previous
end point, or change the analysis type, loads or boundary
conditions or tolerances. A restart analysis is set via the MASTER
command. Recovery of results from a restart file without
continuing the analysis is also possible.
• All relevant solution data needed for a restart run are saved in a
file (extension .res) in case they are needed in a restart analysis.
The frequency of data writing to a restart file is set via the
MASTER command.
• To perform a restart analysis, the .res file from the first model
must be copied to the second model. When running in interactive
mode, the AUI will first look for the restart file, and if it is not
found, the user will be prompted to locate the restart file. When
running in batch mode the .res file must be copied before starting
the second analysis.
• Note that once the second analysis starts, it will overwrite the
.res file with new data. Therefore, if the user wishes to redo the
second run, then the .res file must be copied again from the first
model. This does not apply to when restart data is appended to the
.res file and when the second run is a frequency analysis (no restart
data in this case).
Result recovery
Frequency (2)
9 9 9 x 9
Analysis
Modal
9 9 9 x x
Analysis
Notes:
(1) Must restart to frequency analysis first
(2) Can also be in the same run (no restart needed)
• The sparse solver and the 3D-iterative solver may both be set to
in-core or out-of-core.
• The disk files that may become large are those with the
following file name extensions:
11.12 Miscellaneous
• The program can write intermediate files to disk for external
processing, and can read various inputs (such as stiffness and mass
matrices) directly from a user-provided file (such as pipe pressures,
forces, or displacements).
where
• When you anticipate that you will use the mapping option, you
must create a mapping file (extension .map) in the first run. In the
AUI user interfaces, choose Control→Mapping and select “Create
Mapping File” or from the AUI command line, use the command
MASTER MAP-OUTPUT=YES. You must also save the AUI
database file.
Create a new data file and run ADINA. You do not have to
provide the restart file from the previous run to the current
ADINA run, but you must provide the mapping file from the
previous run to the current ADINA run.
a user can change the zoomed region and still use the full
model initial solutions for every zooming calculation, thereby
enabling the user to experiment with different designs of the
region of interest
The .dat file for the zoomed part must have a different name
from the full model .dat file.
Only 2-D solid and 3-D solid elements can be used in the
zoom modeling.
These zero pivots will stop the solution, unless the zero pivots are
prevented from occuring.
K ii = (1 + ε STAB ) K ii
(a) Bolt modeled with beam-bolt elements (b) Beam-bolt elements used for bolt shank
F = k ( t L − Lb )+ Fb (11.17-1)
F = k ( t ∆ + ∆ b )+ Fb (11.17-2)
Rest of structure
t
∆ + ∆ b = 0 , the bolt axial force F computed from (11.17-1)
equals the axial force Fb .
Bolt element
Rest of structure
• For all other iterations, in which the bolt is not being adjusted,
the bolt axial force - length relationship is given by (11.17-1),
where Fb and ∆ b are in general non-zero and are held constant.
• The basic ideas used in the 3D-bolt are shown in Fig 11.17-4.
This figure shows a very simple bolt model of four elements, with
the ends of the bolt fixed. For simplicity, it is assumed that the
Poisson's ratio is zero, so that the cross-section of the bolt doesn't
change during the deformations.
Fig. 11.17-4(a) shows the 3D-bolt group along with its bolt plane.
This bolt plane is an additional input for the 3D-bolt group. As
seen, the bolt plane determines the bolt direction and the bolt split
faces. (For more information about the bolt plane, see Section
11.17.5.)
Fig 11.17-4(b) shows an incompatible displacement, called the
bolt displacement, applied to the nodes on the bolt split faces.
Because the bolt model is fixed at its ends, the elements must
increase in length, and, to satisfy equilibrium, the elements must all
be under uniform tension. The nodal point forces acting at the
nodes on the bolt split faces are combined into a single bolt force.
Fig 11.17-4(c) shows the bolt element group after deformations,
as visualized during postprocessing. The elements appear to be
different lengths because the bolt displacement is not included in
the plot. However, the strains are the same in all elements.
In this example, it is clear that the bolt force is determined by
the bolt displacement, and that the bolt force increases as the bolt
displacement increases.
Bolt direction
(normal to
bolt plane) Bolt split faces
Bolt plane
Bolt
displacement
Consistent
nodal point
forces
(b) Bolt displacement and consistent nodal point forces
• Fig 11.17-5 shows the same problem, but with one end of the
bolt free. Now when the bolt displacement is applied, the elements
on the free end move stress-free, and the resulting plot is shown in
Fig 11.17-5(c). This example justifies the use of the term "bolt
displacement".
When the bolt end is free, it is clear that the bolt force is zero
for any value of the bolt displacement.
Bolt
displacement
(c) As visualized
K KUB ∆u R − F
KBU KBB ∆v = R − F (11.17-3)
b b b
Rb − Fb
< tol
(
max Rb , Fb )
is used.
K KUB ∆u R − F
=
KBU α KBB ∆v α v presc − v
(11.17-4)
b ( b b )
K KUB ∆u R − F
KBU α KBB ∆v = 0 (11.17-5)
b
vbpresc − vb
< tol
(
max vbpresc , vb )
is used.
BOLT-OPTIONS TABLE=NO
For example
At time 2.0, all bolts are loaded using force loading. In the first
bolt-loading step, half of the bolt load is applied, and the second
bolt-loading step, the entire bolt loading is applied.
BOLT-OPTIONS TABLE=YES
BOLT-OPTIONS TABLE=YES
BOLT-TABLE 1 TYPE=FORCE TIME=2.0
1 2 0.5
1 3 0.2
2 4 0.3
3 2 1.0
3 3 1.0
3 4 1.0
At time 2.0, bolts 2, 3 and 4 are loaded using force loading. In the
first bolt-loading step, bolt 2 is loaded by 0.5 times its specified
bolt force, bolt 3 is loaded by 0.2 times its specified bolt force and
bolt 4 is not adjusted. In the second bolt loading step, bolts 2 and 3
are not adjusted, and bolt 4 is loaded by 0.3 times its specified bolt
force. In the third bolt loading step, all three bolts are loaded by
their specified bolt forces.
• The results at any bolt step can be saved to file. If bolt tables are
used, the saved bolt steps all have the same solution time, and are
distinguished by their bolt sequence numbers.
• For the TLA and TLA-S automatic time step options, the bolt
loading procedure can only be applied at the beginning of the
analysis or at restart. The same holds for the LDC method.
• Bolt loading can be used with ADINA heat flow, for both one-
way or fully coupled TMC solutions.
• The 3D bolt element group also uses a bolt plane as part of its
definition. In the general case, the bolt plane is defined using the
BOLT-PLANE command and the bolt plane is specified using the
BOLT-PLANE parameter of the EGROUP THREEDSOLID
command.
The bolt plane gives the bolt direction and the approximate
location of the bolt split faces. The bolt direction is used to
determine the direction of the bolt displacement and the direction
of the bolt force. The bolt plane is used to determine the bolt split
faces. Note that the bolt split faces do not necessarily coincide with
the bolt plane. Rather, the program determines the bolt split faces
to lie "near" the bolt plane.
Bolt plane
Option=ZPLANE
x=(0,0,COORDINATE)
P2
Bolt direction
x
P1
Option=XPLANE Option=YPLANE
x=(COORDINATE,0,0) x=(0,COORDINATE,0)
Centroid
Automatically determined
bolt plane
• The AUI outputs the bolt direction and bolt cross-sectional area
for each bolt during data file generation. This information can be
used to confirm that the bolt plane is defined correctly.
• When the bolt force is specified for a system in which one end
of the bolt is free (for example, the system shown in Fig 11.17-5),
the global system of equations (11.17-3) is singular. This is
physically correct as there is no non-zero bolt force that can satisfy
equilibrium.
To prevent numerical difficulties in this case, parameter BOLT-
OPTIONS FACTOR3 can be used. FACTOR3 provides an
additional stiffness relating the bolt force and bolt displacement
displacement. The default for FACTOR3 is zero, and the
dimension of FACTOR3 is stiffness. If FACTOR3 is set nonzero,
its value should be much smaller than the bolt stiffness itself.
Alternatively, stiffness stabilization can be used to prevent
numerical difficulties.
BOLTPOINT B2 BOLT=2
• For the beam-bolt elements, the bolt forces are available for
post-processing using the variables BOLT_FORCE,
BOLT_SHEARFORCE-S, BOLT_SHEARFORCE-T,
BOLT_TORSION, BOLT_MOMENT-S, BOLT_MOMENT-T.
The bolt shortening is also available using variable
BOLT_SHORTENING. These variables are stored along with the
elements. For example
t +∆t
T (i ) −t +∆t T ( i −1)
t +∆t (i )
2
< TOLL (12.1)
T
2
t +∆t
U (i ) −t +∆t U ( i −1)
t +∆t (i )
2
< TOLL (12.2)
U
2
• For the direct TMC option, the program solves one system of
equations which includes displacement and temperature increments
and there is no TMC iteration process; only the standard Newton-
Raphson iteration procedure is applied for which convergence or
displacements/forces and temperatures are checked separately
using equations (12.1) and (12.2) respectively.
θ structure
(k )
= (1 − λ1 ) θ heat
( k − 2)
+ λ1 θ heat
( k −1)
- diffusive radiosity
- thermal eigenvalue calculations
- seepage
- analogous field problems
- Joule heating
12.2 Modeling
• One finite element model is used for mechanical and heat flow
solutions with the following features:
• For shell elements, heat flow loads can be applied either on top,
bottom or mid-surface. Note that different temperature loads can be
applied to the top and bottom surface. When temperatures are
applied to the midsurface they are assigned to both top and bottom
shell surfaces.
• Line search can also be used in the heat transfer solution in both
heat transfer only and coupled heat transfer analyses.
• Mesh glueing can be used for temperature DOFs for 2-D and 3-
D solid elements. If the GLUEMESH command is specified in the
input and a TMC analysis is requested, then, in addition to
displacement DOFs, temperature DOFs are also constrained
between glued domains. If heat flow only option is used, then only
temperature DOFs are constrained.
v = −k ⋅∇π (12.1)
where
σ = Ce − απ 1 (12.2)
where
θ = α ev + βπ (12.3)
where
ev = soil skeleton volumetric strain
β = second soil consolidation parameter
∇ ⋅σ + f b = 0 (12.4)
∂θ
− q b = −∇ ⋅ v (12.5)
∂t
∂ ev ∂π
∇ ⋅ ( k ⋅∇π ) = α +β − qb (12.6)
∂t ∂t
We can now see that the stress equation (12.4) and the diffusion
equation (12.6) constitute a coupled diffusion-stress equation
system.
∂u
0 0 ∂t K u K uπ u f
K T + = (12.7)
πu M ∂π 0 K π π q
∂t
where
K u = ∫ BTu C Bu dV
V
K uπ = − ∫ BTu α H dV
V
K π u = − ∫ HT α Bu dV
T
K π = ∫ BπT k Bπ dV
V
M = ∫ HT β H dV
V
K u u = f − K uπ π (12.8)
∂π ∂u
M + K π π = q − K πT u (12.9)
∂t ∂t
Equation (12.8) is solved using the ADINA model, for a given fluid
pressure field, and equation (12.9) is solved, for a given
displacement field. An iterative procedure can therefore be used to
solve this equation system.
K UU K UΦ U FUB + FUS
K = (12.10)
ΦU KΦΦ Φ FΦS
where
K UU = mechanical stiffness
KΦΦ = electric permittivity matrix
K UΦ =piezoelectric matrix
KΦ U = KΤUΦ
FUB = body force loading vector
FUS = surface force loading vector
FΦS = surface charge density vector
U = vector of structural displacements
Φ = vector of electric potentials
- Method QT, where heat fluxes are transferred from the fluid
model to the solid model. Note, that this option requires a
prescribed temperature in at least one point of the solid model.
• The BTFSI method is used in one way FSI coupling, i.e., the
fluid model is executed first. Time step size depends on fluid flow
conditions only. For every solution step, fluid stresses and
temperatures (or heat fluxes) are stored in a file to be used when
displacements and stresses are calculated in the solid model.
• Solution time steps can be different for fluid flow and solid
model calculations. If the fluid flow solution time span is smaller
than the solid model solution time span, then the last saved fluid
stresses and temperatures (or heat fluxes) are used.
• For the BTFSI coupling the models are prepared exactly the
same way as for the FSI coupling without temperatures, i.e., there
is no need to model solids in the fluid model, temperatures (or heat
fluxes) are transferred via FSI boundaries and then heat flow
solution option in the solids is used to calculate temperatures and
heat fluxes in the solid model.
• The AUI must know the answers to the following four questions
before it can obtain a result.
When (for which load step, mode shape, etc) is the result to
be evaluated?
You can supply answers to each of these questions within the AUI.
"What" is controlled by variables. "Where" is controlled by model
points, model combination points, model lines, zones and result
grids. "When" is controlled by response depictions and
response-range depictions. "How" is controlled by smoothing
depictions and result control depictions.
Spatial transformations:
The AUI can combine the results from all of the elements
attached to a node into a single result (smoothing).
The AUI can interpolate the results from the nodes to points
within elements.
Time transformations:
Resultant transformations:
Combination transformations:
matter which rule to use; both will give the correct results. If the
combination transformation is nonlinear and the resultant
transformation is linear, use “resultants before combinations”. If
the combination transformation is linear and the resultant
transformation is nonlinear, use “resultants after combinations”. If
both transformations are nonlinear, then neither rule will give the
correct results.
Element face: Variables that are output at the element edges (2D
analysis) or element faces (3D analysis) of ADINA-F elements.
The tables give notes for those variables that are not self-
explanatory.
1) These are the components of the displacements and rotations transformed by the
AUI into a user-specified coordinate system. More information is given at the end of
this section.
2) The A, B and C components refer to the skew coordinate system (if one is used).
The V1, V2 components refer to the shell rotation directions (for shell midsurface
nodes with five degrees of freedom). The X, Y and Z components always refer to the
global coordinate system. The AUI can transform results that were calculated by
ADINA in one coordinate system to another coordinate system, see Table 13.1-24.
This comment holds for velocities, accelerations and eigenvectors as well as
displacements.
1) These are the components of the velocities, angular velocities, etc., transformed by
the AUI into a user-specified coordinate system. More information is given at the
end of this section.
3) This is the 1st time derivative of the fluid potential degree of freedom.
6) These are the 1st time derivatives of the ovalization degrees of freedom.
7) This is the 1st time derivative of the pore pressure (see Table 13.1-7 for the pore
pressure).
9) These are the 1st time derivatives of the warping degrees of freedom.
4) This is the 2nd time derivative of the fluid potential degree of freedom.
5) These are the 2nd time derivatives of the ovalization degrees of freedom.
6) This is the 2nd time derivative of the pore pressure (see Table 13.1-7 for the pore
pressure).
7) These are the 2nd time derivatives of the warping degrees of freedom.
1) These variables are computed by the AUI from data output by the solution
program.
2) Variables that begin with FE_ denote quantities that are output directly from the
solution program. For example, FE_EFFECTIVE_STRESS is the effective stress as
output from ADINA for certain elements/material models.
These variables should be distinguished from variables defined in the AUI as
predefined resultants (see Note 1).
(
σ e = 1/ 2 (τ xx − τ yy ) + (τ xx − τ zz ) + (τ yy − τ zz ) + 6 (τ xy2 + τ xz2 + τ yz2 )
2 2 2
)
However, the AUI interpolates and smoothes FE_EFFECTIVE_STRESS using the
same algorithms that it uses for any other stress component.
5) These are the principal stresses and therir directions, as output by the ADINA
program. Note that the concrete and curve description models, these are the true
principal stresses only before cracking has occurred. After cracking, the directions
are fixed corresponding to the crack directions and these variables are no longer
principal stresses.
For the data fitted concrete material model, FE_SIGMA-P1_ANGLE and
FE_SIGMA-P2_ANGLE for 2-D solid elements, and FE_SIGMA-
P1_DIRECTION-(XYZ), FE_SIGMA-P2_DIRECTION-(XYZ) and
FE_SIGMA-P3_DIRECTION-(XYZ) for 3-D solid elements define the normal
directions of the crack planes; FE_SIGMA-P1, FE_SIGMA-P2 and FE_SIGMA-
P3 are the stress values at the time of crack formation along the normal directions of
the crack planes.
For 2-D solid elements, FE_SIGMA-P3 is always out of the plane of the element
and FE_SIGMA-P1, FE_SIGMA-P2 are always in the plane of the element
(FE_SIGMA-P1 ≥ FE_SIGMA-P2) only for the concrete and curve description
material models.
6) These are the principal stresses as computed by the AUI from the stress
components (SIGMA-P1 ≥ SIGMA-P2 ≥ SIGMA-P3).
7) This is the maximum shear stress, defined as one-half the difference between the
maximum and minimum principal stress, as computed by the AUI from the stress
components.
8) This is the 2-norm of the stress tensor (the larger of the absolute values of the
maximum and minimum principal stress) as computed by the AUI from the stress
components.
9) These are the stresses transformed by the AUI into a user-specified coordinate
system. More information is given at the end of this section.
Note that these variables have nothing to do with the initial stresses entered in the
initial strain directions.
10) These are the stress components in the axes of the material coordinate system, for
orthotropic materials. Note that the AUI cannot transform stresses output in other
coordinate systems to the material coordinate system.
11) These are equal to the global stress components for those elements in which the
stresses are output in the global coordinate system, equal to the isoparametric
coordinate system stress components for those elements in which the stresses are
output in the isoparametric coordinate system, and equal to the material coordinate
system stress components for those elements in which the stresses are output in the
material coordinate system. These variables are mainly used for constructing
resultants that are invariants (that is, do not depend on the coordinate system
direction). For example, resultant variable EFFECTIVE_STRESS is defined in
terms of STRESS-II, STRESS-JJ, etc.
12) These are the stress components in the axes of the isoparametric coordinate
system. For shell elements, component directions R and S refer to directions r and
s (see Section 2.7.1 and Figure 2.7-6). Note that the AUI cannot transform stresses
output in other coordinate systems to the isoparametric coordinate system.
13) These are the stress components in the global coordinate system. Note that the
AUI cannot transform stresses output by ADINA in other coordinate systems to the
global coordinate system.
15) This is the value of the pore pressure, used in the porous media formulation.
17) In 2-D analysis, the BB, AB and CC components are defined; in 3-D analysis, the
BB, CC, AB, AC, BC components are defined.
1) The off-diagonal strain terms are engineering strains; i.e. twice the tensorial
strains.
2) These are the Cartesian components of the left stretch tensor V, as output by
ADINA. See Section 3.1, equation 3.1-5 for the definition and meaning of V.
The principal stretches can be computed in the AUI by defining a resultant using
functions SIGP1, SIGP2 and SIGP3. For example, you can define, for 3-D problems,
The left stretch tensor is also used in the AUI for plotting the stretches as element
vectors.
3) These are the strain components transformed by the AUI into a user-specified
coordinate system. See the end of this section for more information.
Note that these strains have nothing to do with the initial strains entered in the
initial strain directions.
4) The same comments as are given for the stress variables (notes 10, 12, 13 of Table
13.1-7) apply to the strain variables.
5) The definition of the accumulated effective plastic strain is given in Section 3.4.1.
6) These are the components of the deformation gradient tensor 0t X , as output from
ADINA. Note that there are 9 components, since the deformation gradient tensor is
in general not symmetric.
7) These are the Cartesian components of the left stretch tensor V, as computed by
the AUI from the deformation gradients output by ADINA. See Section 3.1,
equation 3.1-5 and Figure 3.1-2 for the definition and meaning of V.
The principal stretches can be computed in the AUI by defining a resultant using
functions SIGP1, SIGP2 and SIGP3. For example, you can define, for 3-D problems,
The left stretch tensor is also used in the AUI for plotting the stretches as element
vectors.
8) These are the Cartesian components of the right stretch tensor U, as computed by
the AUI from the deformation gradients output by ADINA. See Section 3.1,
equation 3.1-1 for the definition and meaning of U.
The principal stretches can be computed in the AUI by defining a resultant using
functions SIGP1, SIGP2 and SIGP3. For example, you can define, for 3-D problems,
9) In 2-D analysis, the BB, AB and CC components are defined; in 3-D analysis, the
BB, CC, AB, AC, BC components are defined.
10) These are the maximum, intermediate and minimum principal logarithmic
strains. They are available in large strain analysis.
12) These variables are used with the SMA material defined in Section 3.14.
13) These are the maximum intermediate and minimum principal strains, as
computed from variables STRAIN-II, STRAIN-IJ, etc.
14) These variables apply to the data fitted concrete material model. They are the
strain values at the time of crack formation along the normal directions of the crack
planes.
( )
1) This is evaluated by the AUI as 0.5 M r r − M s s , which is an invariant (does not
depend on the directions of r and s .
( )
2) This is evaluated by the AUI as 0.5 Rr r − Rs s , which is an invariant (does not
depend on the directions of r and s .
5) These are the consistent nodal point forces corresponding to the stresses within the
element. The AUI does not transform these from the coordinate system in which
they are output by ADINA.
6) These are computed by the AUI from the data output by ADINA.
−M s s Mr s
7) This is evaluated by the AUI as the maximum principal value of .
Mr s M r r
Rr r Rr s
8) This is evaluated by the AUI as the maximum principal value of .
Rr s Rs s
−M s s Mr s
9) This is evaluated by the AUI as the minimum principal value of .
Mr s M r r
Rr r Rr s
10) This is evaluated by the AUI as the minimum principal value of .
Rr s Rs s
11) Each of these variables is the square root of the sum of the squares of the
components, as computed by the AUI.
14) For 2D, 3D, plate, shell and spring elements, these variables are output in the
coordinate system of the node to which the element is attached (global system or
skew system). For example, if a 3D element is attached to nodes with skew systems,
variable NODAL_FORCE-X is the nodal force in the skew A direction.
15) These are the components of the contact forces, transformed by the AUI into a
user-specified coordinate system. More information is given at the end of this
section.
NORMAL_TRACTION-(XYZ) Contact
RESTRAINING_TRACTION-(XYZ) Drawbead segment 4
SURFACE_TRACTION-(XYZ) Mesh surface 2
TANGENTIAL_TRACTION Contact 3
TANGENTIAL_TRACTION-(XYZ) Contact
2) These are computed by the AUI as predefined resultants using the formulas
SURFACE_TRACTION-X = <STRESS-XX>*<SURFACE_NORMAL-X> +
<STRESS-XY>*<SURFACE_NORMAL-Y> +
<STRESS-XZ>*<SURFACE_NORMAL-Z>
SURFACE_TRACTION-Y = <STRESS-XY>*<SURFACE_NORMAL-X> +
<STRESS-YY>*<SURFACE_NORMAL-Y> +
<STRESS-YZ>*<SURFACE_NORMAL-Z>
SURFACE_TRACTION-Z = <STRESS-XZ>*<SURFACE_NORMAL-X> +
<STRESS-YZ>*<SURFACE_NORMAL-Y> +
<STRESS-ZZ>*<SURFACE_NORMAL-Z>
See Table 13.1-17 for the definitions of the surface normal variables.
6) These variables are output when the “new” contact segments are used.
7) NODAL_NORMAL_TRACTION is the magnitude of the normal traction.
8) These variables are output when the “new” contact segments are used in frictional
contact.
10) These are the components of the nodal normal and tangential tractions,
transformed by the AUI into a user-specified coordinate system. More information is
given at the end of this section.
1) These are the components of the reactions, moment reactions, etc., transformed by
the AUI into a user-specified coordinate system. More information is given at the
end of this section.
2) Reactions are only output for those nodes with fixities. Reactions that are
numerically zero might not be output by ADINA.
4) These are the components of momentum flux output on ADINA-F element edges
(2D analysis). On boundaries with no flow, momentum flux is equal to reaction.
5) These are the components of momentum flux output on ADINA-F element faces
(3D analysis). On boundaries with no flow, momentum flux is equal to reaction.
HEAT_FLUX_SURFACE = <HEAT_FLUX-X>*<SURFACE_NORMAL-X> +
<HEAT_FLUX-Y>*<SURFACE_NORMAL-Y> +
<HEAT_FLUX-Z>*<SURFACE_NORMAL-Z>
See Table 13.1-17 for the definitions of the surface normal variables.
KINETIC_ENERGY_FLUX_SURFACE =
KINETIC_ENERGY_DENSITY*VOLUME_FLUX_SURFACE
MASS_FLUX_SURFACE = ELEMENT_DENSITY*
(<X-VELOCITY>*<SURFACE_NORMAL-X> +
<Y-VELOCITY>*<SURFACE_NORMAL-Y> +
<Z-VELOCITY>*<SURFACE_NORMAL-Z>)
See Table 13.1-17 for the definitions of the surface normal variables. You must
define variable ELEMENT_DENSITY before using variable
MASS_FLUX_SURFACE.
4) These are computed by the AUI as predefined resultants using the formulas
MOMENTUM_FLUX-X_SURFACE =
<X-VELOCITY>*<MASS_FLUX_SURFACE>
MOMENTUM_FLUX-Y_SURFACE =
<Y-VELOCITY>*<MASS_FLUX_SURFACE>
MOMENTUM_FLUX-Z_SURFACE =
<Z-VELOCITY>*<MASS_FLUX_SURFACE>
You must define variable ELEMENT_DENSITY before using these variables. See
Table 13.1-17 for the definitions of the surface normal variables.
VOLUME_FLUX_SURFACE = <X-VELOCITY>*<SURFACE_NORMAL-X> +
<Y-VELOCITY>*<SURFACE_NORMAL-Y> +
<Z-VELOCITY>*<SURFACE_NORMAL-Z>
See Table 13.1-18 for the definitions of the surface normal variables.
6) Each of these variables is the square root of the sum of the squares of their
components, as computed by the AUI.
7) These are the components of the fluxes, transformed by the AUI into a user-
specified coordinate system. More information is given at the end of this section.
8) This is the mass flux output at ADINA-F element edges (2D analysis). The mass
flux is positive when the mass flux is in the same direction as the element edge
inwards normal.
9) This is the mass flux output at ADINA-F element faces (3D analysis). The mass
flux is positive when the mass flux is in the same direction as the element face
inwards normal.
2) These are the components of the current density, transformed by the AUI into a
user-specified coordinate system. More information is given at the end of this
section.
1) This variable is used to plot the cracked surface in 3-D crack propagation analysis
(ADINA fracture mechanics analysis). It is equal to 1.0 if the node is on the crack
front or has been released; otherwise it is not defined. This variable can be plotted in
a band plot to shade the area that has cracked during the analysis.
2) This variable indicates the current boundary condition applied to a node on a crack
propagation surface (used in ADINA fracture mechanics analysis). It has the
following values: 0=node is released, 1=node is on the crack front and is fixed,
2=node is on the crack front and is constrained, 11=node is unreleased and is fixed,
12=node is unreleased and is constrained.
The crack boundary flag is not defined on nodes that are not on the crack
propagation surface.
3) These variables are used in ADINA fracture mechanics analysis, virtual crack
extension method, see Chapter 10.
5) This value of the J-integral does not include temperature, pressure or dynamic
corrections.
7) These variables are used in ADINA fracture mechanics analysis, line contour
method.
8) This is the average of the J-integrals for the outer and inner paths within elements
that use 2×2 Gauss integration.
10) This variable gives, for each node, the amount of participation in the virtual shift,
from "undefined" (node is not included in virtual shift) to 0.0 (node is included, but
does not shift) to 1.0 (node is included and shifts completely). It can be used in a
band plot to show the virtual shift.
2) The failure variables are used in analyses employing the anisotropic failure criteria
within shell element groups, see Section 2.7.8.
3) Each failure criterion variable stores the value of the failure criterion used in the
analysis.
4) The failure flag has the following values: 0=material at the point considered has
not failed, 1=material at the point considered has failed.
5) All of the failure flags with a qualifier have the following values: 0=material at the
point considered has not failed in that mode, 1=material at the point considered has
failed in that mode.
6) This variable returns the number of cracks at the point, as computed by the AUI
from the variable CRACK_FLAG. If the material is crushed, the value of
CRACK_FLAG is 4.
7) The plastic flag has the following values: 1=elastic conditions, 2=plastic
conditions.
For the Mohr-Coulomb material, the plastic flag can have the following values:
1=elastic conditions; 2=Mohr-Coulomb plastic deformation; 3=plastic deformation at
the yield surface vertices (Drucker-Prager yielding); 4=tension cutoff. Note that, for
the Mohr-Coulomb material, variable MOHR-COULOMB_FLAG should be used
instead of PLASTIC_FLAG. MOHR-COULOMB_FLAG has the same numerical
values as PLASTIC_FLAG, and the AUI knows the allowed values of MOHR-
COULOMB_FLAG.
8) This variable is used for the Drucker-Prager material model. It has the following
values: 1=elastic conditions, 2=Drucker-Prager yielding, 3=cap yielding, 4=vertex
yielding, 5=tension cutoff yielding. Note that, for the Drucker-Prager material,
variable DRUCKER-PRAGER_FLAG should be used instead of PLASTIC_FLAG_2.
DRUCKER-PRAGER_FLAG has the same numerical values as PLASTIC_FLAG_2,
and the AUI knows the allowed values of DRUCKER-PRAGER_FLAG.
9) This variable has the following values: 0=no wrinkling, 1=wrinkling in one
direction, 2=wrinkling in two directions.
10) This variable is the value of the yield function of failure surface function.
13) SMA_FLAG is used with the SMA material. It has the following values: 1=elastic
state; 2=martensite transformation; 3=austenite transformation.
14) CAM-CLAY_FLAG is used with the Cam-clay material. It has the following
values: 1=elastic state, 2=hardening, 3=softening, 4=critical.
17) DF-CONCRETE_FLAG is used with the data fitted concrete material model. It
has the following values: 1 = elastic unloading; 2 = elastic reloading; 3 = nonlinear
loading; 4 = one-crack; 5 = two-crack; 6 = fully cracked; and 7 = crushed.
1) These are the components of the eigenvectors transformed by the AUI into a user-
specified coordinate system. More information is given at the end of this section.
1) These are the initial coordinates (coordinates of the original configuration of the
model), when used with responses of type load-step. These are the reference
coordinates (coordinates of the reference configuration), when used with responses of
type mode-shape, response-spectrum, harmonic and random (recall that these
responses include a reference time).
2) These give the changes in the X, Y and Z coordinates between the original
configuration and the current configuration. They are used when the coordinates of
the model change during the analysis, for example, in fracture mechanics crack
propagation analysis.
3) These variables, when used in a line list or line graph, are defined as follows:
4) The lever variables are defined as follows: xL = ( x - xREF )N with similar definitions
for the y and z lever variables. ( x, y ,z ) are either the original or current coordinates,
( xREF, yREF , zREF ) are the coordinates of a reference point and N is an integer greater
than or equal to 1. The choice of original or current coordinate, ( xREF, yREF , zREF )
and N are entered as part of a result control depiction.
The purpose of the lever variables is to make it easier to compute moments about
a reference point.
5) These are the current coordinates (coordinates of the current configuration of the
model), when used with responses of type load-step. These are the reference
coordinates (coordinates of the reference configuration), when used with responses of
type mode-shape, response-spectrum, harmonic and random (recall that these
responses include a reference time).
6) These variables are used to access the face outwards normal of mesh integration or
mesh extreme points.
These variables can also be used when plotting bands (for example, a resultant can
contain these variables when plotting bands). When plotting bands, these variables
refer to the normal at the current time (when the response is of type load-step), or at
the reference time (when the response is of type mode-shape, response-spectrum,
harmonic or random).
7) These are the initial coordinates (coordinates of the original configuration of the
model), when used with any response type.
8) These variables can be used when plotting bands (for example, a resultant can
contain these variables), except when plotting bands on cutting surfaces. When
plotting bands, these variables refer to the normal of the original configuration of the
model.
11) These are the current area and original area of an ADINA-F element edge (2D
analysis). For planar analysis, unit thickness is assumed. For axisymmetric analysis,
one radian is assumed.
12) These are the current area and original area of an ADINA-F element face (3D
analysis).
1) These variables are available from ADINA at the local nodes of ADINA shell
elements.
6) THINNING=-(THICKNESS_CHANGE/THICKNESS_ORIGINAL).
TIME Loc-independent
TIME_FUNCTION_* Loc-independent
4) These variables refer to modal participation factors for ground motions acting in
the X, Y and Z directions.
6) The physical error bound is used to measure the accuracy of the computed
eigenvalues, see Section 6.2.
1) These variables are available for ADINA models when mass properties
calculations are requested in ADINA-IN for all element groups in the model. They
represent the given properties for the entire structure (or for that part of the structure
selected in the AUI) and are therefore location-dependent.
4) The contact state flag has the following values: 0=unknown, 1=free, 2=sticking,
3=sliding. This variable is defined only for node to node contact.
5) This variable stores the number of the coordinate system into which
stresses/strains are transformed, see Section 13.1.3.
6) The fluid dof type has the following values: 0=free velocity potential, 1=fixed,
2=constrained, 3=free velocity potential, 4=free P0 degree of freedom.
7) This is the largest principal inertia, given in a coordinate system with origin at the
center of mass and axes parallel to the global coordinate system.
8) This is the principal axis associated with the largest principal inertia.
9) This is the intermediate principal inertia, given in a coordinate system with origin
at the center of mass and axes parallel to the global coordinate system.
10) This is the principal axis associated with the intermediate principal inertia.
11) This is the smallest principal inertia, given in a coordinate system with origin at
the center of mass and axes parallel to the global coordinate system.
12) This is the principal axis associated with the smallest principal inertia.
13) This is computed by the AUI as a predefined resultant using the formula
KINETIC_ENERGY_DENSITY = 0.5*ELEMENT_DENSITY*
(<X-VELOCITY>**2 + <Y-VELOCITY>**2 + <Z-VELOCITY>**2)
16) These are the moments of inertia, given in a coordinate system with origin at the
center of mass and axes parallel to the global coordinate system. The XX component
of the inertia tensor is equal to the XX moment of inertia.
17) This is the sum (over all pipe elements attached to the node) of the pipe internal
cross-sectional area, based upon the current geometry.
18) This is the sum (over all pipe elements attached to the node) of the internal pipe
skin area, based upon the current geometry.
19) These are the products of inertia, given in a coordinate system with origin at the
center of mass and axes parallel to the global coordinate system. The XY component
of the inertia tensor is the negative of the XY product of inertia.
20) Rotational system has the following values: 0=rotational results are stored in the
global coordinate system, 1=rotational results are stored in a skew system,
2=rotational results are stored in a shell midsurface coordinate system (the shell node
has five degrees of freedom).
21) Translational system has the following values: 0=translational results are stored
in the global coordinate system, 1=translational results are stored in a skew system.
24) These are computed by the AUI as predefined resultants using the formulas
SURFACE_MOMENT-X =
<Y-LEVER>*<SURFACE_TRACTION-Z> -
<Z-LEVER>*<SURFACE_TRACTION-Y>
SURFACE_MOMENT-Y =
<Z-LEVER>*<SURFACE_TRACTION-X> -
<X-LEVER>*<SURFACE_TRACTION-Z>
SURFACE_MOMENT-Z =
<X-LEVER>*<SURFACE_TRACTION-Y> -
<Y-LEVER>*<SURFACE_TRACTION-X>
See Table 13.1-18 for the definitions of the lever variables and Table 13.1-10 for the
definitions of the traction variables.
25) NODAL_CONTACT_GAP is output when the “new” contact segments are used.
27) These are computed by the AUI as predefined resultants using the formulas
REACTION_MOMENT-X =
<Y-LEVER>*<Z-REACTION> -<Z-LEVER>*<Y-REACTION>
REACTION_MOMENT-Y =
<Z-LEVER>*<X-REACTION> -<X-LEVER>*<Z-REACTION>
REACTION_MOMENT-Z =
<X-LEVER>*<Y-REACTION> -<Y-LEVER>*<X-REACTION>
See Table 13.1-18 for the definitions of the lever variables and Table 13.1-11 for the
definitions of the reaction variables.
1) These variables are used in conjunction with the user-supplied material model.
2) This variable has the constant value “1”. It is included so that you can determine
the total area of a domain that is defined as a mesh integration point.
3) Resultants that you define in the AUI are added to the list of user-defined
variables. The variable location type of a resultant is determined by the variables that
define the resultant.
• The AUI can transform node field and node discrete results
calculated by ADINA in one coordinate system to another
coordinate system as shown in Table 13.1-24.
Variable prefix
Coordinate system of X,Y,Z A,B,C V1,V2
ADINA/ADINA-F
result
Global 1 1 3
Skew 2 1 3
V1,V2 3 1 1
B −1
C = Cmin + ( Cmax − Cmin )
A −1
a grid value of 0.
Label point: You specify the label point in the element at which
results were computed by the solution program.
If the results were calculated at the integration points, then
the label is a 1, 2, 3 or 4 digit number giving the integration
point as shown in Table 13.1-26.
If the results were calculated at the element local nodes, then
the label is the local node number.
No extrapolation is performed by the AUI when using the
label option.
ADINA-F:
Element type Result locations Label number
2-D fluid Int. pts. 10(INR) + INS*
3-D fluid Int. pts. 100(INR) + 10(INS) + INT*
Notes:
(end of table)
When the element point lies on the shell midsurface, when the
results extrapolated from the element integration points are
evaluated at the point, the t coordinate and layer used for the
evaluation are selected by the parameters of the result control
depiction.
3) List or plot the mass flux at the model result point. The output
mass flux is the sum of the mass fluxes at all of the element
edges or faces of the element face-set or edge-set.
You can also list the individual mass fluxes at the element edges or
faces of an element edge-set or element face-set. For example,
given an element edge-set or face-set:
Integral: ∫ zdS
S
∫ zdS
Averaged: zav = S
∫z
2
dS
Mean-square: zms = S
∫z
2
dS
S
Root-mean-square:
S
∫ ( z − z ) dS
2
av
S
Variance:
S
∫ ( z − z ) dS
2
av
S
Standard deviation:
S
2
z−z
∫S zav av dS
Relative variance:
S
2
z−z
∫S zav av dS
Relative standard deviation: (also
S
referred to as the coefficient of variation)
extreme) points. These are all based upon the face outwards
normal. The face outwards normal can be accessed through
variables SURFACE_NORMAL-X, SURFACE_NORMAL-Y,
SURFACE_NORMAL-Z. Useful quantities for which predefined
resultants are defined are VOLUME_FLUX_SURFACE,
MASS_FLUX_SURFACE, SURFACE_TRACTION-X,
SURFACE_TRACTION-Y, SURFACE_TRACTION-Z,
MOMENTUM_FLUX-X_SURFACE, MOMENTUM_FLUX-
Y_SURFACE, MOMENTUM_FLUX-Z_SURFACE,
KINETIC_ENERGY_DENSITY,
KINETIC_ENERGY_FLUX_SURFACE,
HEAT_FLUX_SURFACE. Any of these variables can be used
along with a mesh integration point. For example, you can
determine the volume flux of flow through a pipe by defining a
mesh integration point corresponding to a cutting surface, then
graphing or listing the variable VOLUME_FLUX at the mesh
integration point.
You can also obtain the surface area of the mesh integration
point by evaluating variable ONE at the point (variable ONE is
defined to have the value 1.0).
You can also obtain the approximate volume of the mesh by use
of the divergence theorem (also known as Gauss’ theorem or
Green’s theorem), which is:
∫ ∇iA dV = ∫ Ain dS
V S
Rules: The rules are the extreme type, the multiplying factor,
the search grid order and the domain configuration. The
extreme types can be absolute maximum, maximum or
minimum. Use the multiplying factor to multiply the result after
the extreme value is found. Use the search grid order to indicate
the number of sampling points within each element face. Use
the domain configuration to indicate whether to search for the
extreme value using the original or current deformed
configuration of the model.
Please note that you do not specify the quantity within the
mesh extreme point. You specify the quantity when you
reference the mesh extreme point within another dialog box.
Once you have defined a mesh extreme point, you can use it
within any command that accepts a model result point. For
example, you can plot the time history of an extreme quantity
using Graph→Response Curve (Model Point).
The variables that can be used for mesh integration points
can also be used for mesh extreme points.
Smoothing
Results from 2-D, 3-D (solid or fluid) and plate elements can
be smoothed. Results from shell elements can be smoothed on
the midsurface of the shell elements if shell element midsurface
calculations are turned on in the result control depiction.
13.1.4 When (for which load step, etc.) are results available
accum=0.0
∫ v(t ) dt
t1
Time averaged: vtav =
t2 − t1
t2
∫ v(t )
2
dt
t1
Time mean-square: vtms =
t2 − t1
t2
∫ ( v(t ) − v )
2
tav dt
t1
Time variance: vtvar = = vtms − vtav
2
t2 − t1
t2 2
v(t ) − v
∫t vtav tav dt v
Time relative variance: vtrvar = 1 = tvar
t2 − t1 2
vtav
N
Sample sum: ∑v
i =1
i
∑v i
Sample averaged: vsav = i =1
∑v 2
i
Sample mean-square: vsms = i =1
∑(v − vsav )
2
i
Sample variance: vsvar = i =1
= vsms − vsav
2
2
vi − vsav
N
∑
v
v
= sav = svar
i =1
Sample relative variance: vsrvar 2
N vsav
• The AUI includes dialog boxes for each of the response types.
Each dialog box defines a response of the desired response type.
You then use the response in another command when you want to
plot or list the results using that response.
For response types response spectrum, harmonic, random,
response combination, random or envelope, the AUI does not
actually compute the response until the response is referenced by
another command.
• The breaks in the bands are easily seen if the bands are drawn
using two colors that alternate (so that the resulting plot looks like
a photoelastic image). This option is available in the AUI when a
band table of type repeating is requested.
For convenience, the AUI has an Error Plots icon that allows
you to easily plot an error indicator. For example, suppose that a
band plot of unsmoothed pressure is displayed. When the Error
Plots icon is clicked,
• See the AUI Primer, Problem 13, for an example showing the
evaluation of meshes using band plots.
Index
2 A
2-D solid elements, 15 Acceleration variables, 1083
3-D plane stress, 15 Accumulated effective plastic strain,
axisymmetric, 15 395, 424
element output, 27 Active column solver, 812
formulations, 23 ADINA heat transfer capabilities,
generalized plane strain, 15 1059
mass matrices, 26 ADINA System documentation, 4
material models, 23 ADINA-TMC
numerical integration, 24 Piezoelectric analysis, 1067
plane strain, 15 soil consolidation analysis, 1063
plane stress, 15 Alignment elements, 305
recommendations for use, 33 element output, 348
triangular, 22 Analysis zooming, 1034
nd
2 Piola-Kirchhoff stresses, 361 Analytical rigid targets, 679
Anand material model, 603
3 Arc length method, 830
3-D solid elements, 37 Arrival time, 743
element output, 50 Arruda-Boyce material model, 503
formulation, 46 3-D analysis, 503
mass matrices, 50 axisymmetric analysis, 503
material models, 45 plane strain analysis, 503
numerical integration, 47 plane stress analysis, 503
prisms, 44 selection of material constants, 504
pyramids, 44 ATS method, 741, 824, 842
recommendations for use, 56 dynamic analysis, 827
tetrahedra, 44 low speed dynamics, 823
3D-iterative solver, 816 Automatic step incrementation, 818
3D-shell elements, 194 ATS method, 824
LDC method, 830
5 Axisymmetric shell elements, 123,
125, 133
5 degrees of freedom node, 157 element output, 131, 134
formulations, 132
6 linear, 129
6 degrees of freedom node, 159 mass matrices, 134
material models, 132
nonlinear, 132
elastic, 366 E
inelastic, 366
Earthquake loading, 853
total, 363
Effective plastic strain, 395
Deformation-dependent distributed
Effective-stress-function algorithm,
loads, 749
409
Deformation-dependent pressure
Eigenvalue problems, 799
loads, 749
Eigenvector variables, 1103
Determinant search method, 804
Eight-parameter creep law
Direct time integration, 303, 849
variable coefficients, 446
Director vectors, 147
Eight-parameter creep law, 444
Disk files, 1029
Elastic-isotropic material model, 372,
displacements, 1032
373
factorized linear stiffness matrices,
Elastic-orthotropic material model,
1030
372, 374
forces, 1032
2-D solid elements, 374
global mass, 1030
3-D solid elements, 379
global stiffness, 1030
fabric, 575
pipe internal pressures, 1031
plate elements, 379
temperature gradients, 1032
shell elements, 380
temperatures, 1031
Electric field variables, 1098
Displacement variables, 1081
Electromagnetic loads, 779
Displacement-based finite elements,
Element ‘death upon rupture’, 1010
17, 39
Element birth/death, 818, 1003
Displacement-based fluid elements,
Element local node field variables,
246
1079
3-D, 247
Element local node variables, 1079
axisymmetric, 247
Element locking, 192
element output, 248
Element output for
irrotational conditions, 246
2-D solid elements, 27
planar, 247
3-D solid elements, 50
Distributed loads, 748
alignment elements, 348
beam, 753
axisymmetric shell elements, 133
deformation dependent, 749
beam elements, 83
iso-beam, 755
cohesive elements, 356
pipe, 755
displacement-based fluid elements,
shell, 758
246
Draw bead variables, 1078
general elements, 228
Drucker-Prager material model, 558
iso-beam elements, 134
Dynamic analysis, 849, 860
moment-curvature beam elements,
95
nonlinear spring/damper/mass
Friction
axisymmetric analysis, 506
basic models, 673
plane strain analysis, 506
user-supplied models, 675
plane stress analysis, 505
Friction delay, 670
selection of material constants, 506
Full Newton iterations, 818, 821, 840
line searches, 821
I
G Ilyushin material model, 412
Implicit time integration, 850, 860
Gap override, 665
Newmark method, 850, 860
Gasket material model, 585
trapezoidal rule, 851, 862
General elements, 226, 854
Wilson theta method, 850
element output, 228
Improperly supported bodies, 696
Geometric imperfections, 802, 1038
Incompatible modes finite elements,
Geotechnical material models, 552
21, 43
Global mass matrix, 864
Inelastic deformations, 366
Green-Lagrange strains, 358, 360
Inertial properties, 1022
Gurson material model, 390
center of mass, 1022
fluid, 1084
Trapezoidal rule, 851, 862 force, 1092
TRS material, 495, 544 , 1108
True strains, 361
frequency/mode