[CIG-SHORT] Initial stress conditions

[Brad Aagaard]

Eric,

I will look at the manual to see if we can improve the description of 
using the initial stresses.

My recommended approach is to write down the stress field in the domain 
and make sure your BC are consistent. Then check the stress and fault 
traction output to make sure they are consistent with what you set as 
parameters. If they don't agree, then something is wrong in the 
simulation setup. Carefully examining all the various diagnostic output 
is how I catch most of my errors in setting up simulations.

Regards,
Brad


On 11/06/2013 03:34 PM, Eric Lindsey wrote:
> OK, I think I've resolved it - you were right that the shear tractions in
> the two cases were different; in the case of applied initial stresses, the
> final shear stress throughout the domain (and on the fault) was close to
> zero. So it seems that Pylith simply relaxed all the initial shear stress
> before allowing the fault to slip, because I had a free-slip boundary on
> one side.  I resolved this by adding an extra Neumann boundary condition on
> that side to maintain the shear stress at the boundary. In retrospect this
> is obvious, but somehow it wasn't intuitive to me how these two stress
> conditions would interact.
>
> Thanks again,
> Eric
>
>
>
> On Wed, Nov 6, 2013 at 1:31 PM, Brad Aagaard <baagaard at usgs.gov> wrote:
>
>> Eric,
>>
>> You can examine the fault tractions to see if they are consistent between
>> the case when you use BC and when you use an initial stress field. They
>> should be the same if everything is setup consistently. This is what I have
>> done for the SCEC dynamic spontaneous rupture benchmarks where I needed an
>> initial stress field over the domain for an elastoplastic rheology.
>>
>> Regards,
>> Brad
>>
>>
>>
>> On 11/06/2013 01:22 PM, Eric Lindsey wrote:
>>
>>> It seems I don't understand what the initial stress conditions imposed by
>>>
>>> [pylithapp.timedependent.materials.elastic]
>>> db_initial_stress
>>>
>>> are doing. I've got a dynamic fault with static friction that will slip
>>> when I impose Dirichlet or Neumann BC that introduce a large enough shear
>>> stress (I took most of the configuration from
>>> examples/3d/hex8/step11.cfg),
>>> but if I impose these initial stresses instead of the extra boundary
>>> conditions, I get no slip! What's going on here - am I misunderstanding
>>> the
>>> purpose of initial stress?
>>>
>>> Thanks,
>>> Eric
>>>
>>>
>>>
>>> _______________________________________________
>>> CIG-SHORT mailing list
>>> CIG-SHORT at geodynamics.org
>>> http://geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>>
>>>
>> _______________________________________________
>> CIG-SHORT mailing list
>> CIG-SHORT at geodynamics.org
>> http://geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>
>
>
>
> _______________________________________________
> CIG-SHORT mailing list
> CIG-SHORT at geodynamics.org
> http://geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>