[CIG-SHORT] dynamic benchmark problem

Mon Apr 15 11:03:20 PDT 2013

Roby,

The SCEC benchmarking site http://scecdata.usc.edu/cvws/cgi-bin/cvws.cgi 
shows the results from various codes. There are various implementations 
and discretizations. Unfortunately, they don't show the effect of 
discretization size very well. The benchmarks were designed to be run at 
100m, but some benchmarks are more robust than others. The branching 
benchmarks tend to more dependent upon discretization size than others 
due to the effects of dynamic stresses.

See TPV24-25 for examples (I ran these at both 100m and 200m 
resolutions). In TPV24-25 it is clear that there is a greater 
sensitivity to discretization size when the rupture is just on the verge 
of propagating. A finer discretization will tend to yield larger peaks 
in dynamic stresses, thereby making it easier for triggering of ruptures 
on favorably oriented faults.

Regards,
Brad

On 4/15/13 10:52 AM, Roby Douilly wrote:
> Thank you Brad, I will try changing the shear wave speed and wave period
> to see. But I do have a question
> regarding this benchmark. I run the 200m resolution and in this case the
> rupture didn't propagate in the branch
> fault. The reason I was trying to run the 100m resolution is to see if
> the rupture could propagate on the branch fault.
> Do you think that the rupture will propagate on the branch fault at 100m
>   and also is 200m to large for the rupture to
> jump on the branch fault?
>
> Roby
>
> On Apr 15, 2013, at 1:46 PM, Brad Aagaard <baagaard at usgs.gov
> <mailto:baagaard at usgs.gov>> wrote:
>
>> Roby,
>>
>> It has been several years since I ran this benchmark. I am not sure if I
>> have actually used the 100m settings for this benchmark. I have used
>> this mesh in TPV24-25. In those cases, when I ran the 100m resolution
>> case, I used the following scales for nondimensionalizing the problem:
>>
>> normalizer.shear_wave_speed = 3333*m/s
>> normalizer.wave_period = 0.3*s
>>
>> NOTE: You will need a machine with lots of memory or a cluster to run
>> the simulation at 100m. Remember that in 3-D using a mesh with 1/2 the
>> grid spacing increases the problem size by a factor of 8x.
>>
>> Regards,
>> Brad
>>
>> On 4/15/13 10:32 AM, Roby Douilly wrote:
>>> Brad,
>>>
>>> I run it with this command:
>>>
>>> pylith tet4.cfg tpv15.cfg tpv15_tet4_100.cfg
>>>
>>>
>>> Roby
>>>
>>>
>>> On Apr 15, 2013, at 1:28 PM, Brad Aagaard <baagaard at usgs.gov
>>> <mailto:baagaard at usgs.gov>
>>> <mailto:baagaard at usgs.gov>> wrote:
>>>
>>>> Roby,
>>>>
>>>> Please give the *exact* command line arguments you are using. You could
>>>> be leaving out a .cfg file or using outdated files.
>>>>
>>>> Brad
>>>>
>>>> On 4/15/13 10:24 AM, Roby Douilly wrote:
>>>>> Hello,
>>>>>
>>>>> I am trying to run the scec dynamic benchmark tpv15 for 100m. Each time
>>>>> I tried to run it, I got the error message below.
>>>>> There is no error message when I run it for 200m. Is there a way to
>>>>> avoid this error.
>>>>>
>>>>> Thanks,
>>>>>
>>>>> Roby
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> RuntimeError: Determinant of Jacobian (9.85786e-07) for cell 1668 is
>>>>> smaller than minimum permissible value (1e-06)!
>>>>>
>>>>> --------------------------------------------------------------------------
>>>>> MPI_ABORT was invoked on rank 0 in communicator MPI_COMM_WORLD
>>>>> with errorcode -1.
>>>>>
>>>>> NOTE: invoking MPI_ABORT causes Open MPI to kill all MPI processes.
>>>>> You may or may not see output from other processes, depending on
>>>>> exactly when Open MPI kills them.
>>>>> --------------------------------------------------------------------------
>>>>> --------------------------------------------------------------------------
>>>>> mpirun has exited due to process rank 0 with PID 17553 on
>>>>> node chilmark.eas.purdue.edu <http://chilmark.eas.purdue.edu>
>>>>> <http://chilmark.eas.purdue.edu>
>>>>> <http://chilmark.eas.purdue.edu> exiting
>>>>> without calling "finalize". This may
>>>>> have caused other processes in the application to be
>>>>> terminated by signals sent by mpirun (as reported here).
>>>>> --------------------------------------------------------------------------
>>>>> /opt/pylith/bin/nemesis: mpirun: exit 255
>>>>> /opt/pylith/bin/pylith: /opt/pylith/bin/nemesis: exit 1
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> ---------------------------------------------------------------------------------
>>>>> 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 <mailto:rdouilly at purdue.edu>
>>>>> <mailto:rdouilly at purdue.edu>
>>>>> <mailto:rdouilly at purdue.edu>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> CIG-SHORT mailing list
>>>>> CIG-SHORT at geodynamics.org <mailto:CIG-SHORT at geodynamics.org>
>>>>> <mailto:CIG-SHORT at geodynamics.org>
>>>>> http://geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>>>>
>>>>
>>>> _______________________________________________
>>>> CIG-SHORT mailing list
>>>> CIG-SHORT at geodynamics.org <mailto:CIG-SHORT at geodynamics.org>
>>>> <mailto:CIG-SHORT at geodynamics.org>
>>>> http://geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>>
>>>
>>>
>>> Roby
>>>
>>> ---------------------------------------------------------------------------------
>>> 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 <mailto:rdouilly at purdue.edu>
>>> <mailto:rdouilly at purdue.edu>
>>>
>>>
>>>
>>>
>>>
>>>
>>> _______________________________________________
>>> CIG-SHORT mailing list
>>> CIG-SHORT at geodynamics.org <mailto:CIG-SHORT at geodynamics.org>
>>> http://geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>>
>>
>> _______________________________________________
>> CIG-SHORT mailing list
>> CIG-SHORT at geodynamics.org <mailto:CIG-SHORT at geodynamics.org>
>> http://geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>
>
>
> Roby
>
> ---------------------------------------------------------------------------------
> 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 <mailto:rdouilly at purdue.edu>
>
>
>
>
>
>
> _______________________________________________
> CIG-SHORT mailing list
> CIG-SHORT at geodynamics.org
> http://geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>