[CIG-SHORT] fault edge motion

[Brad Aagaard]

Huihui,

If there is a gap between thw two faults (your diagram makes it look 
like there is) ...

If so, then there are two buried edges one for the red fault and one for 
the blue fault; you will need to create a separate nodeset for each.

If there is no gap, then there is indeed only one nodeset and both 
faults should label it as a buried edge. There will be no slip along 
this line. If you do not use a buried edge, then the topology of the 
mesh along these buried ends is undetermined. The faults could be split 
so that everything remains in the plane of the faults, OR the cells 
could be split off at an angle. It is because of this ambiguity that we 
require the buried edges to be marked. See Slide 54 from Step 02 of the 
troubleshooting tutorial at 
https://wiki.geodynamics.org/_media/software:pylith:tutorials:cdm2017:pylithtutorial_troubleshooting.pdf. 
This indeterminate behavior was true in previous versions of PyLith, but 
you may not have realized it.

If you are attempting to divide a single planar fault surface into a 
prescribed slip region and friction region, this is not currently 
supported. I have some idea on how it could eventually be implemented as 
long as the prescribed slip and friction regions remain the same 
throughout the simulation. The bookkeeping would quite difficult and 
would involve a fair amount of very special coding to setup the fault 
data structures. I am guessing you want this combination approach. I 
have created a GitHub issue on this topic 
https://github.com/geodynamics/pylith/issues/48. This will help keep 
this issue on our radar as it will likely be some time before anyone 
attempts to implement this.

Regards,
Brad


On 10/25/2017 07:46 PM, Huihui Weng wrote:
> Hi Brad,
> 
> The two faults are parallel and adjacent as the attached png file shows. 
> The overlap of the two faults is the fault edge (one line buried inside 
> the domain), which I set as fault edge in the cfg file. Other boundaries 
> of the two faults are intersecting with the domain boundaries, thus they 
> are not buried fault edges. Part of the scripts are as fellow:
> 
> [pylithapp.timedependent]
> interfaces =[fault_up,fault_down]
> [pylithapp.timedependent.interfaces]
> fault_up =pylith.faults.FaultCohesiveDyn
> fault_down =pylith.faults.FaultCohesiveKin
> 
> [pylithapp.timedependent.interfaces.fault_up]
> label =fault_up
> edge =fault_edge
> [pylithapp.timedependent.interfaces.fault_down]
> label =fault_down
> edge =fault_edge
> 
> 
> The nodes of the fault edge (the white line) don’t slip, causing a slip 
> “gap” between the two faults. If I comment either of the lines of "edge 
> = fault_edge” in the script, it fails to run.
> 
> In addition, how about the case if the two faults bend, such as one kink 
> at the fault edge?  In fact, this case is similar with the example in 
> the PyLith manual, i.e., ~/examples/2d/subduction/step03.cfg, except 
> that I change it to 3D and change one kinematic fault to a dynamic 
> fault. Thanks.
> 
> Best regards,
> Huihui
> 
>> On 25 Oct 2017, at 11:26 PM, Brad Aagaard <baagaard at usgs.gov 
>> <mailto:baagaard at usgs.gov>> wrote:
>>
>> Huihui,
>>
>> I do not know what you mean by "two adjacent planar faults". Are the 
>> faults parallel? perpendicular? Do they intersect? Please send a 
>> diagram of your fault geometry and indicate which edges you marked as 
>> buried fault edges.
>>
>> Thanks,
>> Brad
>>
>>
>> On 10/25/2017 02:25 AM, Huihui Weng wrote:
>>> Dear all,
>>> I tried to build a model with two adjacent planar faults, one is 
>>> kinematic fault and the other is dynamic fault. The PyLith of new 
>>> version requires me to provide the fault edge, i.e., the group of 
>>> nodes in the line that connects the two faults. But I find the fault 
>>> edge always keep fixed. How can I set up the model to make the fault 
>>> edge move to fill the “gap” of slip between these two faults? It does 
>>> seem it is OK for 2D models of older Pylith versions. Thanks.
>>> Best regards,
>>> Huihui
>>> _______________________________________________
>>> CIG-SHORT mailing list
>>> CIG-SHORT at geodynamics.org <mailto:CIG-SHORT at geodynamics.org>
>>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>
>> _______________________________________________
>> CIG-SHORT mailing list
>> CIG-SHORT at geodynamics.org <mailto:CIG-SHORT at geodynamics.org>
>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>