Features tested
User subroutines to define isotropic Mises plasticity, Mooney-Rivlin, and neo-Hookean
hyperelastic material properties.
User subroutine for isotropic Mises plasticity model
Elements tested
Problem description
This set of verification problems tests many of the variables that are passed
into UMAT, such as material
properties, step times, and strain increment data. These tests also verify that
the user-defined quantities of stresses, solution-dependent variables, and the
Jacobian matrix are properly transferred to the solution process. These tests
are modifications of the tests described in Rate-independent plasticity. Wherever an elastic-plastic material was
defined in those tests, a user-defined material has been implemented in its
place. The structure being analyzed is a cube made of a single
C3D8 element.
Material:
- Elasticity
-
Young's modulus, E = 200.0E3
Poisson's ratio,
= 0.3
- Plasticity
-
Hardening:
| Yield
stress |
Plastic
strain |
| 200. |
0.0000 |
| 220. |
0.0009 |
| 220. |
0.0029 |
Results and discussion
The tests in this section are set up as cases of homogeneous deformation of a
single element of unit dimensions. Consequently, the results are identical for
all integration points within the element. In each case the constitutive path is
integrated with 20 increments of fixed size.
User subroutine for hyperelastic model
Elements tested
Problem description
This set of verification problems is primarily intended to test the deformation
gradient that is passed into UMAT. Variables in
subroutine UHYPER that are functions
of the deformation gradient are also tested. The structure being analyzed for
the two-dimensional case is a unit square made up of three coincident
CPE4 elements. The
three-dimensional case consists of a cube of unit dimensions made up of three
coincident C3D8 elements. For both
cases the material properties of the first element are specified directly with
general hyperelastic material properties. The same material properties are
defined for the second and third elements through user subroutines UMAT and UHYPER, respectively. The
displacements are prescribed at each of the nodes of the models, thus the
stresses induced in each element are the same.
Material:
- Hyperelasticity: Mooney-Rivlin
-
= 80.0
= 0.0
= 2.013E−4
Results and discussion
The tests in this section are set up as cases of homogeneous deformation of three
elements of unit dimensions. Consequently, the results are identical for all
integration points within each of the elements. In each case the constitutive
path is integrated with 10 increments of fixed size.
User subroutine for total and incompressible hybrid formulations
Elements tested
Problem description
This set of verification problems is primarily intended to test the total and
incompressible hybrid formulations with user subroutine UMAT.
Material:
- Hyperelasticity: neo-Hookean
-
= 100.0
= 1.E-03 or
= 0
Results and discussion
The tests in this section are set up as cases of homogeneous deformation. For
each case the results are compared against native Abaqus models and are found to be identical to the latter.
User subroutine for frequency domain viscoelastic behavior
Elements tested
Problem description
This verification problem is primarily intended to test the implementation of
frequency domain viscoelastic behavior in user subroutine UMAT in terms of the
storage modulus (stiffness) and loss modulus (damping).
Material:
- Linear elastic: viscoelastic
-
= 200.E+09,
= 0.3
= 1.775E-03.
= –5.116E-03
Results and discussion
The test involves the steady-state dynamic response of a cantilever beam
subjected to pressure load. The results are compared to a similar model that
uses a native Abaqus material model. The results for the two problems are identical.
Additional user subroutine problems for hyperelastic model
Elements tested
Problem description
This set of verification problems is primarily intended to test the variables
that are passed into UHYPER. In each test the
material properties are specified using general hyperelastic material properties
for the first element and the strain energy function and the associated
derivatives defined in user subroutine UHYPER for the second
element. Three different sets of material data are used, as described below.
Material 1
Polynomial (N=1), compressible
= 80.0
= 20.0
= 1.E−3
Material 2
Polynomial (N=1), compressible, field
variable dependency included for UHYPER
= 80.0
= 20.0
= 1.5E−3
Material 3
Polynomial (N=1), incompressible,
temperature dependency included
= 0.0,
= 80.0,
= 20.0
= 20.0,
= 75.0,
= 18.0
= 30.0,
= 70.0,
= 16.0
Results and discussion
Since the same boundary conditions are applied on the corresponding nodes of the
two elements, the results are expected to be identical in the two cases, thus
verifying the use of UHYPER with the second
element.
User subroutine for defining different hyperelastic models within the parallel
rheological framework
Elements tested
Problem description
The tests in this section define different hyperelastic material models in
different networks of a nonlinear viscoelastic material model using the parallel
rheological framework (see Parallel Rheological Framework and VISCOELASTIC). For the
purpose of verification, the results are compared with those obtained using an
element overlay approach.
The element overlay consists of two elements sharing the same nodes. The first
element of the overlay has material properties defined by a polynomial
hyperelastic material model (see Table 1). The second element of the overlay uses a neo-Hookean hyperelastic model
with a nonlinear viscoelastic network, defined such that there is no
contribution from the equilibrium network (see Table 2 and Table 3). A third element is introduced to test user subroutine UHYPER in combination with
a parallel rheological framework definition using the properties described in
Table 1, Table 2, and Table 3. The properties and stiffness ratios of the material are defined to provide
the exact same response as the overlay model that uses native Abaqus models. The user subroutine defines the strain energy density function with a
polynomial hyperelastic model for the equilibrium network
(networkid=0) and with a neo-Hookean strain
energy function for the nonlinear viscoelastic network,
(networkid=1). All elements are subjected to the
same uniaxial tension loading and the same boundary conditions.
Material:
Table 1. Polynomial hyperelastic material (N=1).
|
|
80.0 |
|
|
20.0 |
|
|
0.001 |
Table 2. Neo-Hookean hyperelastic material.
|
|
50.0 |
|
|
0.001 |
Table 3. Hyperbolic-sine creep model coefficients.
|
|
2.5 × 10-5 |
|
|
0.3 |
|
|
1.0 |
Results and discussion
The results obtained using user subroutine UHYPER are identical to
the results obtained using native Abaqus material models and the overlay technique.
|