[CIG-SHORT] attenuation in a dynamic model?

[Brad Aagaard]

Eunseo,

The stable time step in explicit time stepping comes from the CFL 
condition. With a viscoelastic material, the stability is reduced 
because the effective stiffness is higher at high frequencies.

Regards,
Brad


On 08/06/2015 02:01 PM, Eunseo Choi wrote:
> Thanks for the confirmation, Brad!
>
> I'll try a smaller time step but don't understand why it would work.
> The time step I used was a stable one for the dynamic problem, ~0.01 sec,
> which was already much smaller than relaxation times I tried, ~1 sec to ~1
> year.
>
> Eunseo
>
>
> On Thu, Aug 6, 2015 at 1:03 PM, Brad Aagaard <baagaard at usgs.gov> wrote:
>
>> Eunseo,
>>
>> Attenuation for dynamic models is not yet implemented. This has been on
>> the TODO list for a while but many other features have been a higher
>> priority.
>>
>> If you try to use a viscoelastic model with a high viscosity, then you
>> will likely need to reduce the time step. I am guessing that the
>> instability you see is starting at a cell that is slightly more distorted
>> so it has a smaller stable time step.
>>
>> Regards,
>> Brad
>>
>>
>> On 08/06/2015 09:17 AM, Eunseo Choi (echoi2) wrote:
>>
>>> Can PyLith have attenuation in an explicit dynamic problem?
>>>
>>> I couldn't find how to do this anywhere in the v.2.1.0 manual.
>>>
>>>
>>> So, I first tried MaxwellIsotropic3D expecting the relaxation effect
>>> to do the job but even with a very high viscosity, an instability
>>> kicks in around the boundary where a time-dependent traction BC is
>>> applied. Maybe this material model works only in static or
>>> quasi-static problems? Or is it more likely to be a mistake in my
>>> problem setup?
>>>
>>>
>>> I also read in PyLith's TO-DO list that the plan is to use the
>>> generalized Maxwell model for attenuation. Is this the material model
>>> I was supposed to use? Is it working in v.2.1.0 but just not being
>>> discussed in the manual?
>>>
>>>
>>> Thanks,
>>>
>>> Eunseo
>>>
>>>
>>>
>>>
>>>
>>> _______________________________________________ CIG-SHORT mailing
>>> list CIG-SHORT at geodynamics.org
>>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>>
>>>
>> _______________________________________________
>> CIG-SHORT mailing list
>> CIG-SHORT at geodynamics.org
>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>
>
>
> _______________________________________________
> CIG-SHORT mailing list
> CIG-SHORT at geodynamics.org
> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>