[Feature] Enable PoroElasticTransverseIsotropic Model

[FishYNY]

At first, I used this version to run my model. But there is error that getShearModulus() not implemented for this model.
GEOSX version: 0.2.0 (develop, sha1: 94e338c)

• c++ compiler: gcc 9.4.0
• MPI version: Open MPI v4.0.3, package: Debian OpenMPI, ident: 4.0.3, repo rev: v4.0.3, Mar 03, 2020
• HDF5 version: 1.12.1
• Conduit version: 0.8.2
• VTK version: 9.2.6
• RAJA version: 2023.6.1
• umpire version: 2023.6.0
• adiak version: ..
• caliper version: 2.10.0
• METIS version: 5.1.0
• PARAMETIS version: 4.0.3
• scotch version: 7.0.3
• superlu_dist version: 6.3.0
• suitesparse version: 5.7.9
• Python3 version: 3.8.10
• hypre release version: 2.29.0

The error information:
***** ERROR
***** LOCATION: /home/pku/codes/codes09122023/GEOS/src/coreComponents/constitutive/solid/SolidBase.hpp:137
***** Controlling expression (should be false): true
***** Rank 7: getShearModulus() not implemented for this model

I found that "Re-Enable anisotropic constitutive models #2749 ". And I updated the codes and rerun it today. But still there is error. The error information turned into getBulkModulus() not implemented for this model.

GEOS version: 0.2.0, maybe 3c386ca

• c++ compiler: gcc 9.4.0
• MPI version: Open MPI v4.0.3, package: Debian OpenMPI, ident: 4.0.3, repo rev: v4.0.3, Mar 03, 2020
• HDF5 version: 1.12.1
• Conduit version: 0.8.2
• VTK version: 9.2.6
• RAJA version: 2023.6.1
• umpire version: 2023.6.0
• adiak version: ..
• caliper version: 2.10.0
• METIS version: METIS_VERSION
• PARAMETIS version: PARAMETIS_VERSION
• scotch version: 7.0.3
• superlu_dist version: 6.3.0
• suitesparse version: 5.7.9
• hypre version: v2.30.0-44-geab5f9f7f (master)
• trilinos version: 13.4.1
• Python3 version: 3.8.10

***** ERROR
***** LOCATION: /home/pku/codes/codes03032024/GEOS/src/coreComponents/constitutive/solid/SolidBase.hpp:112
***** Controlling expression (should be false): true
***** Rank 0: getBulkModulus() not implemented for this model

[jhuang2601]

@FishYNY I've gone through the code and there is no issue for ElasticTransverseIsotropic model. As the smoke test inputFiles/solidMechanics/elasticHollowCylinder_transverseIsotropic_smoke.xml has already been added to the integratedTest to protect this modelling capabilities, it should be running with the develop repo. Can you test running this smoke example?

[FishYNY]

@FishYNY I've gone through the code and there is no issue for ElasticTransverseIsotropic model. As the smoke test inputFiles/solidMechanics/elasticHollowCylinder_transverseIsotropic_smoke.xml has already been added to the integratedTest to protect this modelling capabilities, it should be running with the develop repo. Can you test running this smoke example?

Sure, I can run the test with this example. However, I am currently working on a fluid-solid coupling model, and I wonder if this will have an effect on the invocation of the ElasticTransverseIsotropic model? Do we have examples where this module is tested within a fluid-solid coupling scenario?

[jhuang2601]

@FishYNY I've gone through the code and there is no issue for ElasticTransverseIsotropic model. As the smoke test inputFiles/solidMechanics/elasticHollowCylinder_transverseIsotropic_smoke.xml has already been added to the integratedTest to protect this modelling capabilities, it should be running with the develop repo. Can you test running this smoke example?

Sure, I can run the test with this example. However, I am currently working on a fluid-solid coupling model, and I wonder if this will have an effect on the invocation of the ElasticTransverseIsotropic model? Do we have examples where this module is tested within a fluid-solid coupling scenario?

Now, I know what is the root cause for your case. In GEOS, current poromechanics wrapper cannot be extended to ElasticTransverseIsotropic model, because of the way how to handle Biot's coefficient and Biot porosity. As of now, ElasticTransverseIsotropic model can only be used to solve mechanical problems, not ready for the fully coupled ones.

[FishYNY]

@FishYNY I've gone through the code and there is no issue for ElasticTransverseIsotropic model. As the smoke test inputFiles/solidMechanics/elasticHollowCylinder_transverseIsotropic_smoke.xml has already been added to the integratedTest to protect this modelling capabilities, it should be running with the develop repo. Can you test running this smoke example?

Sure, I can run the test with this example. However, I am currently working on a fluid-solid coupling model, and I wonder if this will have an effect on the invocation of the ElasticTransverseIsotropic model? Do we have examples where this module is tested within a fluid-solid coupling scenario?

Now, I know what is the root cause for your case. In GEOS, current poromechanics wrapper cannot be extended to ElasticTransverseIsotropic model, because of the way how to handle Biot's coefficient and Biot porosity. As of now, ElasticTransverseIsotropic model can only be used to solve mechanical problems, not ready for the fully coupled ones.

Btw, if the crack is inclined, the transverse direction is along the crack surface rather than along the x/y/z axes. However, I have not seen any parameters like rotation angles within the 'transverseIsotropic' settings. How to achieve arbitrary angle transverse isotropy? Would users need to rotate the stiffness matrix based on the crack orientation beforehand?