I ran the example step12.cfg in directory “\3d\hex8”. And I set the friction coefficient to 0. Then I checked the slip rate on fault and found it is about 2 cm/year on all of fault nodes due to the boundary conditions imposed on +x and -x faces, which is correct.
But when I added the buried edges into the “box_hex8_1000m.exo” and specified them in step12.cfg file. Then I found the slip rates on fault nodes aren’t uniform and significantly small than 2 cm/year, and the slip rate on buried edges is 0cm/year. Is it wrong?
The buried edges must be specified in more complex model such as with intersectant faults, otherwise Pylith always halts with some errors. How can I aviod the problem?
If your fault cuts entirely through the domain, then it can slip uniformly. If it does not, then slip must stop somewhere outside of the vertices where slip happens. Prescribing buried edges just lets PyLith know where the slip should stop. I do not understand how any model can avoid this? Also, the fact that peak slip is not attained next to the boundary is just a consequence of the elastic rheology.
The buried edges are only to be used to force the slip to zero along fault edges that are in the middle of the domain. If the fault edges are on the boundary of the domain, then you don’t use buried edges.
Designating fault nodes as buried edges forces the fault slip to be zero on those edges. This provides the appropriate termination of slip on the fault boundaries when the fault is embedded in the middle of a domain. When you have fault with zero friction in the middle of a domain, the fault slip will not be uniform. For a uniform background stress field, the fault slip will be a maximum in the middle of the fault and taper to zero at the edges.