Prescribed slip and sliding friction regions on a single fault surface

[baagaard-usgs]

Some users want to prescribe fault slip on a portion of a fault surface (e.g., deep creeping regions) while allowing another portion to slip according to some fault constitutive model. This is not currently supported, due to the indeterminate topology of the mesh when inserting the cohesive cells and dealing with the boundary between the two regions. The boundary between the two regions is a buried edge, so normally it would be marked as a fault edge for each fault and end up having zero slip; however, this is not the behavior that a user wants.

To create the desired topology, I think one would use a single fault surface and have separate (limited to two?) regions with prescribed slip (FaultCohesiveKin) and frictional sliding (FaultCohesiveDyn). Conceptually, one could create a single fault surface and associated cohesive cells then separate them into two regions with separate fault meshes, etc that overlap along the shared edge. I think the bookkeeping in separating the single fault and cohesive cells into regions could be quite complicated. At a first glance, I don't know what to do with the cohesive cell in between the two regions; one edge has vertices with prescribed slip and one edge has vertices with frictional sliding. Do we throw this cohesive cell away?

[knepley]

I think the thing to do here in order to understand is to write out the equations for a two fault cell problem that we want this thing to obey. It sounds like this should be rolled into two different efforts:

1) Change the stress model for faults

2) Parallelize the fault determination

Matt

On Thu, Oct 26, 2017 at 1:19 PM, Brad Aagaard notifications@github.com wrote:

Some users want to prescribe fault slip on a portion of a fault surface (e.g., deep creeping regions) while allowing another portion to slip according to some fault constitutive model. This is not currently supported, due to the indeterminate topology of the mesh when inserting the cohesive cells and dealing with the boundary between the two regions. The boundary between the two regions is a buried edge, so normally it would be marked as a fault edge for each fault and end up having zero slip; however, this is not the behavior that a user wants.

To create the desired topology, I think one would use a single fault surface and have separate (limited to two?) regions with prescribed slip (FaultCohesiveKin) and frictional sliding (FaultCohesiveDyn). Conceptually, one could create a single fault surface and associated cohesive cells then separate them into two regions with separate fault meshes, etc that overlap along the shared edge. I think the bookkeeping in separating the single fault and cohesive cells into regions could be quite complicated. At a first glance, I don't know what to do with the cohesive cell in between the two regions; one edge has vertices with prescribed slip and one edge has vertices with frictional sliding. Do we throw this cohesive cell away?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/geodynamics/pylith/issues/48, or mute the thread https://github.com/notifications/unsubscribe-auth/AAjoidEtbq2uRG0nF1_Wd-ZOYcbfMxWdks5swL8KgaJpZM4QH7Fz .

-- What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead. -- Norbert Wiener

https://www.cse.buffalo.edu/~knepley/ http://www.caam.rice.edu/~mk51/

[DianKusumawati]

Dear Sir @knepley @baagaard-usgs

I would like to ask about the multiple faults scheme.

Is it possible if I set several faults in the .cfg file for the FaultCohesiveKin and FaultCohesiveDyn altogether? For instance, three faults where two of them have a steady slip rate (FaultCohesiveKin), and one of them is locked with frictional sliding parameter (FaultCohesiveDyn), as shown in the figure below.

The figure is from Terakawa and Matsu'ura (2009) https://academic.oup.com/gji/article/178/3/1663/1998838

Any guidance will be very valuable. Thank you

[baagaard-usgs]

PyLith v2.2 supports prescribing slip on some faults while having fault friction on others. PyLith v3 does not yet support fault friction. In PyLith v2.2 prescribing slip on multiple faults is covered in examples/2d/subduction. I don't believe we have an example that shows using both types of faults, but we do have examples showing how to use fault friction.

If I understand the figure, this looks like a single fault surface with a region of fault friction embedded in a region with prescribed slip. This is not supported in PyLith as it would need to be modeled with a single fault surface with different physical processes (prescribed slip and fault friction) on subregions. In order to model something like this, you would need to use a single fault surface that has fault friction. You could make the regions outside the "model region" stable sliding regions.

[DianKusumawati]

Thank you for your comment and suggestion.

I am still new to tectonic modeling and trying to grasp the suggestion or idea. I would like to ask about the stable sliding. Is the stable sliding region you meant like those in the examples/2d/subduction? Could you point out what it means by the stable sliding in this strike-slip model region?

Thank you