SNESSolve has not converged and Slip sign

PyLith
1

Dear everyone,

I have 2 questions.

  1. This is a basic question but I need to make sure about the slip sign. Following the manual book page 93-fig 6.2, positive value refers to left lateral, reverse, and fault opening slip. But I just realize in the examples/2d/subduction, fault_slip_coseismic.spatialdb file has negative value for the slip. Which indicates slip for normal fault. I check the finite fault data from Hayes, the slip is in positive value. I also check the fault_slabtop_coseismic.spatialdb in the examples/3d/subduction/spatialdb, the slip is also in positive value which indicates reverse slip. Which one is right?

  2. I am currently learning to simulate spontaneous rupture using step5 in 2d/subduction. I use my own geometry and obtain an error as :

Nonlinear solve did not converge due to DIVERGED_FUNCTION_COUNT iterations 3022
[0]PETSC ERROR: --------------------- Error Message --------------------------------------------------------------
[0]PETSC ERROR:
[0]PETSC ERROR: SNESSolve has not converged

How can I solve this issue?

Thank you for your assistances.

Best Regards,
Alvina

2

Dear Alvina,

  1. If you look at the labels for the parameters in the 2D spatialdb, you will see that the slip is labeled as ‘final_slip_left_lateral’. In 2D, we only have ‘left-lateral’ slip, even if your fault is reverse or normal. This can be a bit confusing at times, but it did not make sense to have 3 slip components for 2D.

  2. This is difficult to answer without knowing more about your problem setup. The first question is what sort of quality does your mesh have? Ideally, you would like a maximum condition number of 2 or less for the cells in your mesh, particularly the cells adjacent to the fault.

Cheers,
Charles

3

@albarodriguez,

If you set

snes_max_funcs = 100000

for the PETSC parameters, it can solve your problem related to DIVERGED_FUNCTION_COUNT, although there could be other issues related to your mesh and problem set up as indicated by @willic3. See the solution here: https://lists.mcs.anl.gov/pipermail/petsc-users/2018-September/036258.html

5

Dear Alvina,
Recently,I am trying to simulate spontaneous rupture by RateStateAgeing,using my own 3D geometry model,but a PETSC error " SNESSolve has not converged" is always reported,How can I solve the problem? Thank you.