Virtual Fields Method Uniaxial Tension Models
Note
The thickness direction is the Z direction and the loading direction must be the Y direction. User provided meshes and data should conform to these directions.
Warning
SIERRA/SM is inherently 3D and, as implemented, the VFM models may produce non-negligible through the thickness stresses depending on the boundary value problem being simulated. In the virtual internal work calculation, all through thickness stresses are currently ignored. This leads to errors in calibration, however, if the boundary value problem conforms well to the plane stress assumption these errors should only be on the order of 1% or less for the parameter values.
Warning
The experimental full-field data must be properly aligned with the mesh. We currently do not assist in aligning the mesh coordinate system with the experimental data coordinate system. Work with the experimentalist to do so or perform alignment as a preprocessing step.
MatCal currently provides two models developed in support
of Virtual Fields Method capabilities. These models are
MatCal’s VFMUniaxialTensionHexModel
and VFMUniaxialTensionConnectedHexModel.
They can be used to reduce the computational cost of models
for MatCal studies when
the following are true:
The loading and specimen geometry from a material characterization test allow for a plane stress assumption to be valid for the specimen’s stress state over the specimen gauge section.
Full-field, in-plane displacement data over the entire specimen gauge section are obtained.
Global load-displacement data are measured.
These models have most of the MatCal standard
model features as described in MatCal SIERRA Solid Mechanics Standard Models.
The following features are disabled for the
VFMUniaxialTensionHexModel:
activate_thermal_coupling()for conduction problems
These features are disabled for all VFM models:
VFMUniaxialTensionHexModel.set_allowable_load_drop_factor
VFMUniaxialTensionHexModel.activate_self_contact
In this section, we will provide more information about how the meshes are generated, specifics on simulation boundary conditions, and what is output from these models.
For more specifics about the VFM method theory and MatCal implementation details see Virtual Fields Method.
VFM model geometry and mesh generation
Both of these models require only two geometric parameters for their geometry to be built and meshed, and they are meshed in very similar ways. These two parameters are passed to the models’ constructors and include a surface mesh filename and the thickness of the specimen being simulated. In the constructor, these parameters have the names:
surface_mesh_filename
thickness
The provided surface mesh should only include the portion of the test specimen where full-field data is collected. To be valid for VFM, this should include all of the visible surface of the tested component between the grips which we will refer to as the gauge section. Since most DIC software is unable to calculate displacements near the free-edges of the specimen, the GMLS tool we employ for mapping field data will extrapolate the displacements linearly over the small regions near free-edges where DIC data is not provided.
Note
This user provided mesh must be in the X-Y plane with the Y axis as the loading axis.
The only difference between the two models is
the connectivity of the mesh elements and
the portion of the volume meshed.
The VFMUniaxialTensionHexModel
most closely emulates classical VFM by taking
the user provided quadrilateral mesh of the desired surface
and creates a hexahedral element from each surface element that
has no connectivity. The hexahedral elements are created
by extruding the surface element in the Z direction.
The total thickness of
each hexahedral element created is the thickness of the tested specimen.
The resulting disconnected hexahedral mesh most
closely emulates classical VFM because each
element effectively operates as a material
point simulator.
The VFMUniaxialTensionConnectedHexModel
creates a hexahedral mesh from the user provided
quadrilateral surface mesh that
maintains the correct connectivity such that it is
the three-dimensional equivalent of the quadrilateral mesh.
The thickness of the created mesh
is half the thickness of the tested specimen.
This mesh is created by extruding the entire
surface mesh in the Z direction and updating
the mesh with the correct connectivity.
VFM model boundary conditions
The boundary conditions for the models are described in this section. Since these models can support thermomechanical coupling, the boundary condition descriptions have been separated into the following two subsections associated with the solid mechanics and thermal models.
VFM solid mechanics boundary conditions
The solid mechanics boundary
conditions for these models are applied
to best approximate the plane stress assumption
required by VFM.
Both of these models have their in-plane
displacements fully prescribed. These
in-plane displacement boundary conditions are
mapped onto the meshes using MatCal’s
interface to the PyCompadre’s GMLS
algorithm. See Full-field Interpolation and Extrapolation
for more details. Both meshes are only one
element thick and both in-plane
faces for each element have their
in-plane nodal displacements prescribed
by two-dimensional interpolation and extrapolation
from the experimental data.
For both models, the out-of-plane displacements
for the nodes on one of the in-plane faces of each
element is fixed. The other out-of-plane displacements for
the remaining nodes are free and must be calculated using
a finite element solve. For the
VFMUniaxialTensionHexModel,
these boundary conditions closely emulate plane stress deformation.
For the
VFMUniaxialTensionConnectedHexModel,
these boundary conditions loosely represent plane stress deformation, while
more accurately capturing through thickness stresses and maintaining
a continuous displacement field on the free surface. Since only half the
specimen is simulated for this model with an out-of-plane symmetry boundary
condition, two elements are used through the thickness which should
provide better estimate of the through thickness stress.
Note
We acknowledge that the boundary conditions are imperfect, however, believer they are the best available given limitations in what is available in Sierra and experimentally. In verification problems on synthetic data, these models produce similar accuracy and are able to reproduce material model parameters that were used to generate the synthetic data within 1%.
The in-plane displacements are determined from full-field
data supplied by the user. The data are passed to the model
through the add_boundary_condition_data()
method. Currently, only a single matcal.full_field.data.FieldData object can be passed
for each state to the add_boundary_condition_data()
method. If an averaged set of displacements is desired for repeats from a single state,
combine the repeat data sets before hand. If the points from each data set are collocated
in space and time,
use NumPy average to create an average field. If not, use the
meshless_remapping()
function to map all data to a common mesh and common time steps.
User must make sure the data are properly aligned
in both space and time before doing so.
VFM thermal model boundary conditions
There are no thermal boundary conditions current supported
for the VFM models. All models are fully insulated.
As previously mentioned,
the VFMUniaxialTensionHexModel
only supports adiabatic heating with the
volumetric heat source being the heating due to plastic work.
The VFMUniaxialTensionConnectedHexModel
supports the addition of conduction with temperature independent
thermal properties.
VFM model specific output
By default, the VFM models require the following exodus element output:
global time
element first_pk_stress - The volume weighted average of the element integration point first Piola-Kirchhoff stress
element centroid - element centroid location for determining the appropriate virtual velocity gradient values to use for the virtual internal work calculation
element volume - the current element volume, only the volume from the first time step is used as the reference configuration element volume.
If coupling is activated or the temperature is prescribed from the data or state, the following temperature output is in the exodus output:
element avg_temperature - The volume weighted average of the element integration point temperature (always)
nodal temperature - temperature at the nodes (only for models with conduction)