The total pressure

[7xinyu]

Hi,

I'm using LaMEM to run a 2D subduction model with phase transition from basalt to eclogite. I set an internal velocity 5cm/yr on the horizontal plate to initializing subduction. The viscosity field seems normal, but the pressure field looks very strange, very different from the example in LaMEM. Besides, I use the eclogite phase transition in LaMEM. However, even the temeprature of basalt parts is higher than 1000 degree, there is no phase transition. I use 32 cpu cores.

I use 20 SNES iterations, and the SNES residual is ~100.

The following are Petsc options and result pictures.

`
-snes_npicard 10
-snes_monitor
-snes_atol 1e-4
-snes_rtol 1e-5
-snes_stol 1e-16
-snes_max_it 20
-snes_max_funcs 50000
-snes_max_linear_solve_fail 10000
-snes_type newtonls

-snes_ksp_ew
-snes_ksp_ew_rtolmax 1e-2

-snes_PicardSwitchToNewton_rtol 1e-4
-snes_NewtonSwitchToPicard_it 15
-snes_NewtonSwitchToPicard_rtol 1e-3

-snes_linesearch_monitor
-snes_linesearch_type cp
-sens_linesearch_maxstep 1.0

-js_ksp_type fgmres
-js_ksp_max_it 25
-js_ksp_converged_reason
-js_ksp_monitor
-js_ksp_rtol 1e-4
-js_ksp_atol 1e-4`

[boriskaus]

Which LaMEM example are you referring to?

Also, it would be helpful if you can upload your input script for us to have a chance to figure out what you did. I don’t know how you set the phase transition, for example

[7xinyu]

In the subduction2D with free surface, the pressure field is similar to lithospheric pressure (the following figure), but the pressure field of my result above looks a little strange.

The matlab script and input file are also atteched. I use LaMEM with PETSc 3.16.4.

Subduction2D_input.txt
CreatePhases2D.txt

[boriskaus]

Looking back at this, it may have been caused by your mantle having Phase=0 (see issue #14).
This has been addressed in PR #15.

@7xinyu: would be great if you can try whether this fixes it.