fix: EFEM bugfixes - effective traction + oldStress #3408

CusiniM · 2024-10-21T22:04:30Z

CusiniM · 2024-10-21T22:06:36Z

src/coreComponents/physicsSolvers/contact/SolidMechanicsEFEMKernelsBase.hpp

    // transp(B)DB
    LvArray::tensorOps::Rij_eq_AikBjk< nUdof, 3, 6 >( Kuw_gauss, matBD, compMatrix );

+    /// FIX: add old Equilibrium operator times oldStress (in Voigt notation)


LvArray::tensorOps::Rij_eq_AikBjk< 3, 3, 6 >( rw, eqMatrix, oldStressVoigt );

and then change uLocal in this kernel to be the increment.

… the solidmechanics efem 2. separate penalty force and fracture pressure contirubtion to the fracture force residual 3. add a sohail's fix to a stencil index issue that causes the crashes of the poromechanics efem-edfm solver

src/coreComponents/physicsSolvers/contact/SolidMechanicsEFEMKernelsBase.hpp

src/coreComponents/constitutive/solid/CoupledSolidBase.hpp

Guotong-Ren · 2024-10-28T21:54:13Z

...Components/physicsSolvers/multiphysics/poromechanicsKernels/SinglePhasePoromechanicsEFEM.hpp

      porousMaterialWrapper.updateStateFromPressureApertureJumpAndTraction( k, 0, pressure[k],
                                                                            oldHydraulicAperture[k], hydraulicAperture[k],
                                                                            dHydraulicAperture_dNormalJump,
-                                                                            jump, traction );


in the previous version, the total traction is passed to the porousMaterialWrapper to update perm. I think only contact traction is needed here. I corrected it here but it may change the results from Integrated Tests. @CusiniM @paveltomin

Now the traction in ContactFields.hpp only represents the contact traction which will be rendered in paraview. Like total stress and effective stress in the rock matrix, user has to do their own calculation to get total force on the fracture. I think this will make the naming more consistent with the data stored here. @CusiniM @paveltomin

I would call it effectiveTraction but yes, this is how it should be IMO. It's also how it is for the other contact solver so it's nice to be consistent.

CusiniM · 2024-10-29T16:45:39Z

...nents/physicsSolvers/multiphysics/poromechanicsKernels/SinglePhasePoromechanicsEFEM_impl.hpp

-  real64 const localJumpFracPressureJacobian = -m_dTraction_dPressure[embSurfIndex] * m_surfaceArea[embSurfIndex];
+
+  // fracture pressure only affects normal direction
+  stack.localJumpResidual[0] += m_fracturePressure[embSurfIndex] * m_surfaceArea[embSurfIndex];


this sign... I guess that if it works it must be correct.

Guotong-Ren · 2024-10-29T18:03:11Z

src/coreComponents/physicsSolvers/multiphysics/SinglePhasePoromechanicsEmbeddedFractures.cpp

-      arrayView2d< real64 > const & fractureTraction = subRegion.template getField< fields::contact::traction >();
-
-      arrayView1d< real64 >  const & dTdpf = subRegion.template getField< fields::contact::dTraction_dPressure >();
+      arrayView2d< real64 > const & fractureContactTraction = subRegion.template getField< fields::contact::traction >();


I gave a different name called fractureContactTraction. if you prefer to effectiveTraction, i can change it. @CusiniM

I think I prefer effectiveTraction. I even think that we should change the name of the field that gets written in teh output. That would actually be more important coz this is just an internal name

i agree. we should refactor it. e.g., two distinct traction fields, one is total while the other is effective. It will make the meaning much clear.

CusiniM · 2024-10-29T18:08:46Z

src/coreComponents/finiteVolume/TwoPointFluxApproximation.cpp

      stencilCellsSubRegionIndex[1] = 0;
      stencilCellsIndex[1] = kes;
-      stencilWeights[1] = 4. * faceArea[fractureRegionIndex] / hydraulicAperture[fractureRegionIndex][0][kes];
+      stencilWeights[1] = 4. * faceArea[kes] / hydraulicAperture[fractureRegionIndex][0][kes];


weird bug...

codecov · 2024-11-06T02:26:04Z

Codecov Report

Attention: Patch coverage is 0% with 21 lines in your changes missing coverage. Please review.

Project coverage is 57.53%. Comparing base (e3445a3) to head (f5fe5ce).
Report is 70 commits behind head on develop.

Files with missing lines	Patch %	Lines
...anicsKernels/SinglePhasePoromechanicsEFEM_impl.hpp	0.00%	8 Missing ⚠️
...sSolvers/contact/SolidMechanicsEFEMKernelsBase.hpp	0.00%	5 Missing ⚠️
...Components/constitutive/solid/CoupledSolidBase.hpp	0.00%	2 Missing ⚠️
...sics/SinglePhasePoromechanicsEmbeddedFractures.cpp	0.00%	2 Missing ⚠️
...omechanicsKernels/SinglePhasePoromechanicsEFEM.hpp	0.00%	2 Missing ⚠️
...ponents/finiteVolume/TwoPointFluxApproximation.cpp	0.00%	1 Missing ⚠️
...ysicsSolvers/contact/SolidMechanicsEFEMKernels.hpp	0.00%	1 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff             @@
##           develop    #3408      +/-   ##
===========================================
- Coverage    57.53%   57.53%   -0.01%     
===========================================
  Files         1092     1092              
  Lines        97848    97858      +10     
===========================================
+ Hits         56299    56301       +2     
- Misses       41549    41557       +8

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

Co-authored-by: Guotong-Ren <133127368+Guotong-Ren@users.noreply.github.com>

EDFM bugfixes: effective traction + oldStress

b35ccb2

CusiniM requested review from castelletto1, matteofrigo5, paveltomin and rrsettgast as code owners October 21, 2024 22:04

CusiniM commented Oct 21, 2024

View reviewed changes

paveltomin requested a review from Guotong-Ren October 25, 2024 13:30

fix stresses used in poromechanics efem solver

5c971a5

Guotong-Ren requested review from dkachuma and joshua-white as code owners October 25, 2024 21:40

CusiniM commented Oct 28, 2024

View reviewed changes

src/coreComponents/physicsSolvers/contact/SolidMechanicsEFEMKernelsBase.hpp Outdated Show resolved Hide resolved

CusiniM commented Oct 28, 2024

View reviewed changes

src/coreComponents/constitutive/solid/CoupledSolidBase.hpp Outdated Show resolved Hide resolved

rename getStress function. remove temporary variable effStress

c22decb

Guotong-Ren self-assigned this Oct 28, 2024

Guotong-Ren added the ci: run integrated tests Allows to run the integrated tests in GEOS CI label Oct 28, 2024

Guotong-Ren changed the title ~~EFEM bugfixes: effective traction + oldStress~~ fix: EFEM bugfixes effective traction + oldStress Oct 28, 2024

Guotong-Ren added ci: run CUDA builds Allows to triggers (costly) CUDA jobs type: bug Something isn't working labels Oct 28, 2024

remove unused variable dFractureTraction_dPressure

0f201fa

Guotong-Ren reviewed Oct 28, 2024

View reviewed changes

CusiniM commented Oct 29, 2024

View reviewed changes

Guotong-Ren reviewed Oct 29, 2024

View reviewed changes

CusiniM added the flag: requires rebaseline Requires rebaseline branch in integratedTests label Oct 29, 2024

CusiniM commented Oct 29, 2024

View reviewed changes

Guotong-Ren added 2 commits October 30, 2024 09:49

rename contactTraction to effectiveTraction

e452d8a

re-baseline

9c71082

Guotong-Ren requested review from cssherman, jhuang2601 and wrtobin as code owners October 31, 2024 18:56

Guotong-Ren removed ci: run CUDA builds Allows to triggers (costly) CUDA jobs ci: run integrated tests Allows to run the integrated tests in GEOS CI labels Oct 31, 2024

paveltomin approved these changes Oct 31, 2024

View reviewed changes

revert back yaml and md

fb39dc0

matteofrigo5 approved these changes Nov 1, 2024

View reviewed changes

Pavel Tomin and others added 2 commits November 1, 2024 13:11

code style

c17fc14

Merge branch 'develop' into bugfix/cusini/edfm-bugfixes

fcbd7cf

paveltomin added ci: run CUDA builds Allows to triggers (costly) CUDA jobs ci: run integrated tests Allows to run the integrated tests in GEOS CI ci: run code coverage enables running of the code coverage CI jobs labels Nov 5, 2024

Merge branch 'develop' into bugfix/cusini/edfm-bugfixes

5db9ba0

paveltomin added 2 commits November 6, 2024 08:31

Update .integrated_tests.yaml

06f3d42

Update BASELINE_NOTES.md

f5fe5ce

paveltomin changed the title ~~fix: EFEM bugfixes effective traction + oldStress~~ fix: EFEM bugfixes - effective traction + oldStress Nov 6, 2024

paveltomin merged commit 9db5dbf into develop Nov 6, 2024
24 of 26 checks passed

paveltomin deleted the bugfix/cusini/edfm-bugfixes branch November 6, 2024 15:58

danielemoretto44 pushed a commit that referenced this pull request Sep 22, 2025

fix: EFEM bugfixes - effective traction + oldStress (#3408)

369f679

Co-authored-by: Guotong-Ren <133127368+Guotong-Ren@users.noreply.github.com>

fix: EFEM bugfixes - effective traction + oldStress #3408

fix: EFEM bugfixes - effective traction + oldStress #3408

Uh oh!

Conversation

CusiniM commented Oct 21, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

codecov bot commented Nov 6, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

CusiniM commented Oct 21, 2024 •

edited

Loading

codecov bot commented Nov 6, 2024 •

edited

Loading