Add missing VOF term in the strong form of the residual #1149

blaisb · 2024-05-20T22:22:53Z

Description of the problem

There was a missing term in the strong form of the residual which is (mu_1 - mu_0) * (grad_u + grad_u^T) * grad_phase
This does not have significant consequences, but I think it is best to remain consistent

Description of the solution

Add missing term

How Has This Been Tested?

All test results barely change, but I had to update them nonetheless
It seems to help a little bit in terms of velocity divergence, but really not much to be honest.

source/solvers/navier_stokes_vof_assemblers.cc

blaisb · 2024-05-20T22:24:40Z

source/solvers/navier_stokes_vof_assemblers.cc

@@ -143,7 +157,7 @@ GLSNavierStokesVOFAssemblerCore<dim>::assemble_matrix(
              const auto &phi_u_j      = scratch_data.phi_u[q][j];
              const auto &grad_phi_u_j = scratch_data.grad_phi_u[q][j];
              const auto &div_phi_u_j  = scratch_data.div_phi_u[q][j];
-              const auto &shear_rate_j = grad_phi_u_j + transpose(grad_phi_u_j);
+              const auto  shear_rate_j = grad_phi_u_j + transpose(grad_phi_u_j);


This should not have been a reference, so this was a bit weird.

AmishgaAlphonius

Funny that it was not thought of before! I only found a small typo to fix :)

source/solvers/navier_stokes_vof_assemblers.cc

lpsaavedra

I have only a small comment. These equations are not in the documentation, am I correct? I guess we should be writing these equations there soon in a "multiphysics coupling" section...

lpsaavedra · 2024-05-22T14:44:28Z

source/solvers/navier_stokes_vof_assemblers.cc

@@ -90,12 +94,22 @@ GLSNavierStokesVOFAssemblerCore<dim>::assemble_matrix(
          calculate_navier_stokes_gls_tau_transient(
            u_mag, viscosity_for_stabilization_vector[q] / density_eq, h, sdt);

+      // Void fraction gradient for strong residual


Suggested change

// Void fraction gradient for strong residual

// Filtered phase fraction gradient for strong residual

lpsaavedra · 2024-05-22T14:48:32Z

source/solvers/navier_stokes_vof_assemblers.cc

@@ -243,6 +256,10 @@ GLSNavierStokesVOFAssemblerCore<dim>::assemble_rhs(
      const Tensor<1, dim> pressure_gradient =
        scratch_data.pressure_gradients[q];

+      // Void fraction gradient for strong residual


Suggested change

// Void fraction gradient for strong residual

// Filtered phase fraction gradient for strong residual

lpsaavedra · 2024-05-22T14:49:30Z

source/solvers/navier_stokes_vof_assemblers.cc

@@ -909,12 +932,22 @@ GLSNavierStokesVOFAssemblerNonNewtonianCore<dim>::assemble_matrix(
          calculate_navier_stokes_gls_tau_transient(
            u_mag, viscosity_for_stabilization_vector[q] / density_eq, h, sdt);

+      // Void fraction gradient for strong residual


Suggested change

// Void fraction gradient for strong residual

// Filtered phase fraction gradient for strong residual

lpsaavedra · 2024-05-22T14:50:58Z

source/solvers/navier_stokes_vof_assemblers.cc

@@ -1107,12 +1140,22 @@ GLSNavierStokesVOFAssemblerNonNewtonianCore<dim>::assemble_rhs(
            u_mag, viscosity_for_stabilization_vector[q] / density_eq, h, sdt);


+      // Void fraction gradient for strong residual


Suggested change

// Void fraction gradient for strong residual

// Filtered phase fraction gradient for strong residual

PierreLaurentinCS

Nothing to say, everything is clean, but I'd like this question answered before merging :
Why don't we add the term also in the strong jacobian ?

hepap · 2024-05-22T15:40:31Z

applications_tests/lethe-fluid/time_dependent_boundaries_vof.output

-  time    x_vof    y_vof    vx_vof   vy_vof
-1.000000 0.513510 0.638807 0.449408 -0.013436
+  time    x_vof    y_vof    vx_vof   vy_vof  
+1.000000 0.512900 0.576332 0.363353 0.013339 


Maybe it's because this test case is weird, but the differences are more significant here.

Yeah I was wondering this also. I'm not exactly sure what's up. I'll double check.

Yeah that test is weird as fuck, I'm not surpirsed the results changed that much, the test has not real physical meaning it's just there to see if the boundary change in time. No alarm to be had.

blaisb · 2024-05-22T18:13:06Z

Nothing to say, everything is clean, but I'd like this question answered before merging : Why don't we add the term also in the strong jacobian ?

Good question. The strong Jacobian term is very expensive to calculate since it has the full hessian. I checked many cases and it never really affects convergence, so for now we can neglect it i think.

Co-authored-by: Amishga Alphonius <[email protected]>

Description The term associated with the viscosity jump in the strong residual for the NS-VOF assembler added in PR Add missing VOF term in the strong form of the residual #1149 led to a ill-posed formulation in phase change cases where the solid is represented by a highly viscous fluid. Hence, we don't consider this term in the current formulation. There was no appropriate application test for lpbf/phase change with vof cases that could have allowed us to identify this ill-posed formulation prior to the merge of PR Add missing VOF term in the strong form of the residual #1149. The LPBF benchmark example prm files were almost up to date, but did not exactly matched the one we used to generate the benchmark results. The mesh files were missing from the repo. Solution This PR removes the term associated with the viscosity jump. A new test is now included to avoid future mistake in lpbf/phase change with vof cases. LPBF benchmark example prm files were modified and are now up to date with the current working version. The mesh files are now available in the repo. Testing The new test simulate the lpbf benchmark case using the extra-extra-coarse mesh. Temperature statistics and mass conservation monitoring are the outputted metrics (to track both the change in HT and NS-VOF solvers). Other NS-VOF tests were updated, and results for those tests returned to their previous values (or almost with small differences). Documentation No modification in the documentation. Co-authored-by: Amishga Alphonius <[email protected]>

…l#1149) Description of the problem There was a missing term in the strong form of the residual which is (mu_1 - mu_0) * (grad_u + grad_u^T) * grad_phase This does not have significant consequences, but I think it is best to remain consistent Description of the solution Add missing term How Has This Been Tested? All test results barely change, but I had to update them nonetheless It seems to help a little bit in terms of velocity divergence, but really not much to be honest. Co-authored-by: Amishga Alphonius <[email protected]> Former-commit-id: 8e1ec70

Description The term associated with the viscosity jump in the strong residual for the NS-VOF assembler added in PR Add missing VOF term in the strong form of the residual chaos-polymtl#1149 led to a ill-posed formulation in phase change cases where the solid is represented by a highly viscous fluid. Hence, we don't consider this term in the current formulation. There was no appropriate application test for lpbf/phase change with vof cases that could have allowed us to identify this ill-posed formulation prior to the merge of PR Add missing VOF term in the strong form of the residual chaos-polymtl#1149. The LPBF benchmark example prm files were almost up to date, but did not exactly matched the one we used to generate the benchmark results. The mesh files were missing from the repo. Solution This PR removes the term associated with the viscosity jump. A new test is now included to avoid future mistake in lpbf/phase change with vof cases. LPBF benchmark example prm files were modified and are now up to date with the current working version. The mesh files are now available in the repo. Testing The new test simulate the lpbf benchmark case using the extra-extra-coarse mesh. Temperature statistics and mass conservation monitoring are the outputted metrics (to track both the change in HT and NS-VOF solvers). Other NS-VOF tests were updated, and results for those tests returned to their previous values (or almost with small differences). Documentation No modification in the documentation. Co-authored-by: Amishga Alphonius <[email protected]> Former-commit-id: b1ec9c1

Description of the problem There was a missing term in the strong form of the residual which is (mu_1 - mu_0) * (grad_u + grad_u^T) * grad_phase This does not have significant consequences, but I think it is best to remain consistent Description of the solution Add missing term How Has This Been Tested? All test results barely change, but I had to update them nonetheless It seems to help a little bit in terms of velocity divergence, but really not much to be honest. Co-authored-by: Amishga Alphonius <[email protected]> Former-commit-id: 8e1ec70

Description The term associated with the viscosity jump in the strong residual for the NS-VOF assembler added in PR Add missing VOF term in the strong form of the residual #1149 led to a ill-posed formulation in phase change cases where the solid is represented by a highly viscous fluid. Hence, we don't consider this term in the current formulation. There was no appropriate application test for lpbf/phase change with vof cases that could have allowed us to identify this ill-posed formulation prior to the merge of PR Add missing VOF term in the strong form of the residual #1149. The LPBF benchmark example prm files were almost up to date, but did not exactly matched the one we used to generate the benchmark results. The mesh files were missing from the repo. Solution This PR removes the term associated with the viscosity jump. A new test is now included to avoid future mistake in lpbf/phase change with vof cases. LPBF benchmark example prm files were modified and are now up to date with the current working version. The mesh files are now available in the repo. Testing The new test simulate the lpbf benchmark case using the extra-extra-coarse mesh. Temperature statistics and mass conservation monitoring are the outputted metrics (to track both the change in HT and NS-VOF solvers). Other NS-VOF tests were updated, and results for those tests returned to their previous values (or almost with small differences). Documentation No modification in the documentation. Co-authored-by: Amishga Alphonius <[email protected]> Former-commit-id: b1ec9c1

blaisb requested review from hepap and AmishgaAlphonius May 20, 2024 22:22

blaisb commented May 20, 2024

View reviewed changes

source/solvers/navier_stokes_vof_assemblers.cc Outdated Show resolved Hide resolved

blaisb commented May 20, 2024

View reviewed changes

blaisb force-pushed the vof_viscous_term branch from 56cda57 to a54fac5 Compare May 20, 2024 22:50

PierreLaurentinCS self-requested a review May 21, 2024 19:47

AmishgaAlphonius approved these changes May 22, 2024

View reviewed changes

lpsaavedra approved these changes May 22, 2024

View reviewed changes

PierreLaurentinCS requested changes May 22, 2024

View reviewed changes

hepap approved these changes May 22, 2024

View reviewed changes

PierreLaurentinCS approved these changes May 22, 2024

View reviewed changes

PierreLaurentinCS added Reviewed and ready to merge Reviewed and waiting for changes and removed Reviewed and ready to merge labels May 27, 2024

blaisb force-pushed the vof_viscous_term branch from e7ddad3 to 47bfa1a Compare May 28, 2024 18:09

blaisb and others added 7 commits May 28, 2024 14:10

Add missing term in strong residual

b594b39

Make assembler fully consistent

e1ff366

Indent

b89caf4

Update tests that need to be updated

19fa639

Update source/solvers/navier_stokes_vof_assemblers.cc

cc8a946

Update tests

74a1c03

Apply suggestions from code review

47bfa1a

Co-authored-by: Amishga Alphonius <[email protected]>

blaisb added Reviewed and ready to merge and removed Reviewed and waiting for changes labels May 28, 2024

blaisb added 3 commits May 28, 2024 15:01

Make test with strict tolerances

dc8fa2a

Increase time for test

371aead

Slightly decrease tolerance

5e95810

blaisb added 6 commits May 29, 2024 07:56

play with the tolerances

4ea4672

Tolerances again

e8299d8

Update timeout

4878ee4

Reset

117a066

Update tolerances

f32d32a

Update test result

5992325

blaisb merged commit 8e1ec70 into master May 29, 2024
8 checks passed

blaisb deleted the vof_viscous_term branch May 29, 2024 17:17

hepap mentioned this pull request Jun 26, 2024

Fix NS-VOF stong residual formulation #1180

Merged

11 tasks

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add missing VOF term in the strong form of the residual #1149

Add missing VOF term in the strong form of the residual #1149

blaisb commented May 20, 2024

blaisb May 20, 2024

AmishgaAlphonius left a comment

lpsaavedra left a comment

lpsaavedra May 22, 2024

lpsaavedra May 22, 2024

lpsaavedra May 22, 2024

lpsaavedra May 22, 2024

PierreLaurentinCS left a comment

hepap May 22, 2024

blaisb May 22, 2024

blaisb May 28, 2024

blaisb commented May 22, 2024

	// Void fraction gradient for strong residual
	// Filtered phase fraction gradient for strong residual

		@@ -1107,12 +1140,22 @@ GLSNavierStokesVOFAssemblerNonNewtonianCore<dim>::assemble_rhs(
		u_mag, viscosity_for_stabilization_vector[q] / density_eq, h, sdt);


		// Void fraction gradient for strong residual

Add missing VOF term in the strong form of the residual #1149

Add missing VOF term in the strong form of the residual #1149

Conversation

blaisb commented May 20, 2024

Description of the problem

Description of the solution

How Has This Been Tested?

Choose a reason for hiding this comment

AmishgaAlphonius left a comment

Choose a reason for hiding this comment

lpsaavedra left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

PierreLaurentinCS left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

blaisb commented May 22, 2024