[CIG-SHORT] fault opening problem

[Brad Aagaard]

Roby,

I just tried a case where I knew the cohesive zone was not well-resolved 
(artificially small slip-weakening parameter d0) and was able to create 
some fault opening errors. So the error appears to be tied more to 
numerical noise and user error than any problem with the fault 
implementation.

Brad


On 11/21/2012 05:50 PM, Roby Douilly wrote:
> Thank you Brad.
> Sometimes it works but some other times I got the error depending on the initial fault tractions.
>
> Roby
> On Nov 21, 2012, at 8:43 PM, Brad Aagaard wrote:
>
>> Roby,
>>
>> I was able to create a simple curved fault and it runs without any problems. However, I get similar errors as yours in a 3-D model but only at the ground surface where I have very low fault normal tractions. I will let you know if I am able to verify that there is a bug.
>>
>> Brad
>>
>>
>> On 11/21/2012 03:29 PM, Brad Aagaard wrote:
>>> Roby,
>>>
>>> It looks like there might be a bug related to nonplanar geometry and
>>> fault slip causing opening when using explicit time stepping. I have not
>>> yet been able to investigate this problem fully.
>>>
>>> If you are building from source, a workaround is to remove the assert on
>>> line 379 of libsrc/pylith/faults/FaultCohesiveDyn.cc:
>>>    assert(fabs(tractionNormal) < _zeroTolerance);
>>>
>>> You will still get the warning about fault opening but the code will run.
>>>
>>> I will try to fix this ASAP but I will likely not be able to get it done
>>> before AGU.
>>>
>>> Regards,
>>> Brad
>>>
>>>
>>>
>>
>
>
>
>
> ---------------------------------------------------------------------------------
> Roby Douilly
> Graduate Student
> Department of Earth, Atmospheric, and Planetary Sciences
> Purdue University
> 550 Stadium Mall Dr.
> West Lafayette, IN 47907-2051
> rdouilly at purdue.edu
>
>
>
>
>