Currently, there are 3 stress components in SeisSol, with partial redundancy.
STRESS_STR_DIP_SLIP_AM (used in Sumatra SC. the optimally oriented fault loading is limited to rake=0 or 90 degree),
OptimalStress (used in Palu, is a generalization of STRESS_STR_DIP_SLIP_AM because we can specify the rake of the maximum traction. It also features other small differences (1) returned tensor not normalized, 2) TR(sij) used to constrain the magnitude of the stress tensor, and not s_zz, (which is not necessarily more realistic)).
AndresonianStress (simpler parameterization for an Andersonian stress, that is all principal stresses vertical or horizontal).
At the end, so far, I kept STRESS_STR_DIP_SLIP_AM for backward compatibility but then users don't know which component to use. Therefore, we could either continue keeping STRESS_STR_DIP_SLIP_AM but insist on its deprecations (see the current state of PR), or drop STRESS_STR_DIP_SLIP_AM.
Also, I would add an extra parameter s_zz in OptimalStress ( to offer the possibility to constrain the magnitude of the stress tensor with s_zz, which at the end seems to make more sense in most cases).
This would lead to some of our previous setups needing slight updates in the easi files.
(e.g. adding s_zz 0 in OptimalStress, changing STRESS_STR_DIP_SLIP_AM to OptimalStress following the example in the readthedocs).
I hope the aim of the PR is clear, let me know what you think,
Currently, there are 3 stress components in SeisSol, with partial redundancy.
At the end, so far, I kept STRESS_STR_DIP_SLIP_AM for backward compatibility but then users don't know which component to use. Therefore, we could either continue keeping STRESS_STR_DIP_SLIP_AM but insist on its deprecations (see the current state of PR), or drop STRESS_STR_DIP_SLIP_AM.
Also, I would add an extra parameter s_zz in OptimalStress ( to offer the possibility to constrain the magnitude of the stress tensor with s_zz, which at the end seems to make more sense in most cases). This would lead to some of our previous setups needing slight updates in the easi files. (e.g. adding s_zz 0 in OptimalStress, changing STRESS_STR_DIP_SLIP_AM to OptimalStress following the example in the readthedocs).
I hope the aim of the PR is clear, let me know what you think,
Thomas.