[CIG-SHORT] UNCHECKED Re: Applying Gravitational Forces with Pylith

Tue Aug 14 23:30:49 PDT 2018

Hi Niloufar,

If you are using Trelis or Cubit to create your mesh, there is a simple way to check the quality.  Otherwise, if you have a VTK file, you can use Paraview to check the quality.  There is a mesh quality filter.  For hexes, I think you should probably use the condition number criterion.

Cheers,
Charles

> On 15/08/2018, at 5:50 PM, Niloufar Abolfathian <niloufar.abolfathian at gmail.com> wrote:
> 
> Thanks, Charles. You should be right. I am using hexahedral cells. The model is a very simple vertical strike-slip fault. But gravity and finite deformation may destort the cells. I have to run the model again to save the vtk files in order to check the paraview. Also now I am trying to check some finer cell sizes and tetrahedral mesh to see how they will affect the results. Thanks for the suggestion.
> 
> I am still not sure how to check the quality of the mesh! Is there any simple way to do that or any resources I can read from?
> 
> Best,
> Niloufar
> 
> 
> 
> On Tue, Aug 14, 2018 at 10:02 PM, Charles Williams <willic3 at gmail.com <mailto:willic3 at gmail.com>> wrote:
> Hi Niloufar,
> 
> I can’t run your problem since I don’t have your mesh; however, it appears that this problem is bombing at about 7300 years.  You have given it a total_time of 30000 years, which is quite a long time.  I would see what the solution looks like running for 7300 years or so (so it doesn’t diverge).  One useful thing would be to look at the deformed mesh (I believe this is called something like ‘warp mesh’ in Paraview).  Use a magnification factor of 1 and see if the mesh is getting too deformed.  It might also be useful to look at the behavior of the solution with time to see if you can detect anything unusual.
> 
> Cheers,
> Charles
> 
> p.s.  Have you checked the quality of your mesh?
> 
> 
>> On 15/08/2018, at 3:55 PM, Niloufar Abolfathian <niloufar.abolfathian at gmail.com <mailto:niloufar.abolfathian at gmail.com>> wrote:
>> 
>> Hi,
>> 
>> I followed the 2D gravity example and I could run my models with gravity and add initial stresses. But in models including finite strain, the nonlinear solver could not converge. I have tried running the model with gravitation on and off, applying depth dependent initial stress and no initial stress, applying multigrid preconditioner and also not applying the preconditioner and also modifying the linear and nonlinear tolerances, but could not converge the model. I have attached the setting when setting when I am using the gravity on, applying initial stress and using preconditioning, I am wondering why it is always giving the error after 6000 years. I really appreciate all your help solving these issues.
>> 
>> Thanks,
>> Niloufar
>> 
>> 
>> 
>> On Thu, Aug 2, 2018 at 8:43 AM, Brad Aagaard <baagaard at usgs.gov <mailto:baagaard at usgs.gov>> wrote:
>> Is this the gravity only problem to compute the initial stresses? If so, then you should remove the fault.
>> 
>> On 08/01/2018 08:59 PM, Niloufar Abolfathian wrote:
>> Thanks, Charles.
>> 
>> My error is due to the convergence of the nonlinear solver. I have made some modifications but not sure if it is a good idea to decrease my residual norms.
>> 
>> Best,
>> Niloufar
>> 
>> 
>> This is the setting I am using:
>> 
>> 
>> [pylithapp.timedependent]
>> formulation =pylith.problems.ImplicitLgDeform
>> 
>> 
>> [pylithapp.timedependent.formulation]
>> output =[domain,subdomain]
>> output.subdomain =pylith.meshio.OutputSolnSubset
>> solver =pylith.problems.SolverNonlinear
>> 
>> [pylithapp.timedependent.formulation.time_step]
>> total_time =50000.0*year
>> dt =2.0*year
>> 
>> [pylithapp.timedependent.interfaces]
>> fault =pylith.faults.FaultCohesiveDyn
>> 
>> [pylithapp.timedependent.interfaces.fault]
>> label =fault
>> edge =fault_edge
>> zero_tolerance =1.0e-11
>> 
>> # Use the slip-weakening friction model.
>> friction.db_properties =spatialdata.spatialdb.UniformDB
>> friction.db_properties.label =Slip weakening
>> friction.db_properties.values =[static-coefficient,dynamic-coefficient,slip-weakening-parameter,cohesion]
>> 
>> friction.db_properties.data =[0.6,0.4,0.2*m,0.0*Pa]
>> 
>> # ----------------------------------------------------------------------
>> # PETSc settings
>> # ----------------------------------------------------------------------
>> [pylithapp.petsc]
>> malloc_dump =
>> 
>> # Preconditioner settings.
>> pc_type =asm
>> sub_pc_factor_shift_type =nonzero
>> 
>> When you encounter convergence errors, a good strategy is to revert to the direct solver. Note that this only works when you don't have a fault.
>> 
>> pc_type = lu
>> 
>> 
>> For most real applications (after you have debugged the setup)l, we recommend using the multigrid preconditioner. See solver_algebraicmultigrid.cfg (no fault) and solver_fieldsplit.cfg )with fault) in examples/3d/subduction.
>> 
>> 
>> # Friction sensitivity solve used to compute the increment in slip
>> # associated with changes in the Lagrange multiplier imposed by the
>> # fault constitutive model.
>> friction_pc_type =asm
>> friction_sub_pc_factor_shift_type =nonzero
>> friction_ksp_max_it =50
>> friction_ksp_gmres_restart =30
>> 
>> # Uncomment to view details of friction sensitivity solve.
>> #friction_ksp_monitor = true
>> #friction_ksp_view = true
>> friction_ksp_converged_reason =true
>> 
>> # Reduce convergence tolerances.
>> ksp_rtol =1.0e-16
>> ksp_atol =1.0e-12
>> ksp_max_it =4000
>> ksp_gmres_restart =100
>> 
>> snes_rtol =1.0e-9
>> snes_atol =1.0e-8
>> snes_max_it =4000
>> 
>> 
>> _______________________________________________
>> CIG-SHORT mailing list
>> CIG-SHORT at geodynamics.org <mailto:CIG-SHORT at geodynamics.org>
>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short <http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short>
>> <error_files.zip>_______________________________________________
>> CIG-SHORT mailing list
>> CIG-SHORT at geodynamics.org <mailto:CIG-SHORT at geodynamics.org>
>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short <http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short>
> Charles Williams I Geodynamic Modeler
> GNS Science I Te Pῡ Ao
> 1 Fairway Drive, Avalon 5010, PO Box 30368, Lower Hutt 5040, New Zealand
> Ph 0064-4-570-4566 I Mob 0064-22-350-7326 I Fax 0064-4-570-4600
> http://www.gns.cri.nz/ <http://www.gns.cri.nz/> I Email: C.Williams at gns.cri.nz <mailto:your.email at gns.cri.nz>
> 
> _______________________________________________
> CIG-SHORT mailing list
> CIG-SHORT at geodynamics.org <mailto:CIG-SHORT at geodynamics.org>
> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short <http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short>
> 
> _______________________________________________
> CIG-SHORT mailing list
> CIG-SHORT at geodynamics.org
> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short

Charles Williams I Geodynamic Modeler
GNS Science I Te Pῡ Ao
1 Fairway Drive, Avalon 5010, PO Box 30368, Lower Hutt 5040, New Zealand
Ph 0064-4-570-4566 I Mob 0064-22-350-7326 I Fax 0064-4-570-4600
http://www.gns.cri.nz/ <http://www.gns.cri.nz/> I Email: C.Williams at gns.cri.nz <mailto:your.email at gns.cri.nz>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.geodynamics.org/pipermail/cig-short/attachments/20180815/c8a8115c/attachment.html>

[CIG-SHORT] ***UNCHECKED*** Re: Applying Gravitational Forces with Pylith

[CIG-SHORT] UNCHECKED Re: Applying Gravitational Forces with Pylith