[CIG-SHORT] Initial stress diagnostic output

Wed Nov 23 14:52:51 PST 2016

Hi Demian,

To use PyLith to generate your initial stresses, look at examples/2d/gravity.  Your case should be even easier, since I believe you are using tetrahedra, correct?  In that cases the stresses are all at the element centers and you won’t have to worry about quadrature points.

Cheers,
Charles

> On 24/11/2016, at 10:32 AM, Demian Gomez <demiang at gmail.com> wrote:
> 
> Hi Charles,
> 
> I tried the approach that we discussed yesterday but when I turn off gravity I get large displacements (1e5) pointing in the opposite direction of gravity. I think this is reasonable, since there is no force to prevent the displacements that result from the initial stresses. The displacement vectors are not completely perpendicular to the surface of my domain as you move towards the edges due to the zero displacement boundary condition imposed at the bottom of the model (see attached figure; cross section of the model). Nevertheless, I don't think this boundary condition should affect the solution if I input the equilibrium stresses and turn gravity on.
> 
> I think this test is evidence that shows that there has to be something not quite right about the interpolation scheme, since the displacement vectors when turning off gravity seam reasonable. I'll try to make my SpatialGridDB finer to see if things improve. I also read somewhere (can't remember exactly where, maybe one of your presentations?) that it's possible to run Pylith once (without imposing slip at the fault, I believe) to get the exact equilibrium stresses and then run the quasi-static problem with slip at the fault. This is the method I should use since my next step is to move from a uniform sphere to a layered sphere with a subducting slab, and calculating the exact stresses to write them into a spatialgriddb will be tricky. The question is: how do I input the resulting stresses from the first run into the second run? I don't think that reading the HDF output file and making a spatialdb will help in this case, as we've seen during the experiment.
> 
> Thank you!
> Demián
> 
> --
> Dr. Demián D. Gómez
> Postdoctoral Researcher
> The Ohio State University - School of Earth Sciences
> 275 Mendenhall Laboratory
> 125 South Oval Mall
> Columbus, Ohio 43210
> Cell: +1 (901) 900-7324 <tel:%2B1%20%28901%29%20900-7324>
> email: gomez.124 at osu.edu <mailto:gomez.124 at osu.edu>
> 
> On Tue, Nov 22, 2016 at 3:12 PM, Charles Williams <willic3 at gmail.com <mailto:willic3 at gmail.com>> wrote:
> Hi Demian,
> 
> Let us know what happens.  The problem is that with interpolation, etc., your stresses may not be in equilibrium.  You will be able to tell.  If there are no displacements, then you should be in equilibrium.  Otherwise, you will get displacements.
> 
> Cheers,
> Charles
> 
> 
>> On 23/11/2016, at 8:06 AM, Demian Gomez <demiang at gmail.com <mailto:demiang at gmail.com>> wrote:
>> 
>> Hi Charles,
>> 
>> Are you suggesting that I should turn off gravity and run my problem with zero displacements at the fault? This will output the stress field that I introduced as initial stress without altering anything, right? Didn't think of that... thanks. 
>> 
>> What do you mean by "You should be able to fix this by constraining the displacements."? My initial stresses are s_xx = s_yy = s_zz = rho*g*depth and s_xy = s_yz = s_xz = 0, which should be the lithostatic stress in equilibrium. My model is spherical built in ECEF coordinates so s_zz is not aligned with g. However, since s_xx = s_yy = s_zz and s_xy = s_yz = s_xz = 0 the initial stress state is independent of the coordinate system. I am therefore assuming that my problem has to be related to the interpolation from the spatialdb to the domain or something similar.
>> 
>> Thanks again,
>> Demián
>> 
>> --
>> Dr. Demián D. Gómez
>> Postdoctoral Researcher
>> The Ohio State University - School of Earth Sciences
>> 275 Mendenhall Laboratory
>> 125 South Oval Mall
>> Columbus, Ohio 43210
>> Cell: +1 (901) 900-7324 <tel:%2B1%20%28901%29%20900-7324>
>> email: gomez.124 at osu.edu <mailto:gomez.124 at osu.edu>
>> 
>> On Tue, Nov 22, 2016 at 12:43 PM, Charles Williams <willic3 at gmail.com <mailto:willic3 at gmail.com>> wrote:
>> Dear Demian,
>> 
>> I may be mistaken, but I believe that if you turn off gravity for your problem the remaining stresses should be the initial stresses.  The one issue is that if the stresses aren’t in equilibrium there will be some displacement, which will in turn alter the stresses.  You should be able to fix this by constraining the displacements.
>> 
>> Cheers,
>> Charles
>> 
>>> On 23/11/2016, at 4:44 AM, Demian Gomez <demiang at gmail.com <mailto:demiang at gmail.com>> wrote:
>>> 
>>> Dear all,
>>> 
>>> I am still trying to incorporate gravity to my model. After I run my quasi-static simulation (with zero displacement at the fault) I'm getting ~16 m of displacement with a very suspicious spatial variation. Therefore, something is inconsistent but I can't figure out what it is. Is there a way to output the initial stress as a diagnostic variable (like density, lambda, etc)? I didn't find how to do this in the manual. It would be very helpful to see how Pylith is assigning my initial stress spatialdb to the domain.
>>> 
>>> Thanks,
>>> Demián
>>> 
>>> --
>>> Dr. Demián D. Gómez
>>> Postdoctoral Researcher
>>> The Ohio State University - School of Earth Sciences
>>> 275 Mendenhall Laboratory
>>> 125 South Oval Mall
>>> Columbus, Ohio 43210
>>> Cell: +1 (901) 900-7324 <tel:%2B1%20%28901%29%20900-7324>
>>> email: gomez.124 at osu.edu <mailto:gomez.124 at osu.edu>
>>> _______________________________________________
>>> 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 <http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short>
>> Charles Williams I Geodynamic Modeler
>> GNS Science I Te Pῡ Ao
>> 1 Fairway Drive, Avalon 5010, PO Box 30368, Lower Hutt 5040, New Zealand
>> Ph 0064-4-570-4566 I Mob 0064-22-350-7326 I Fax 0064-4-570-4600
>> http://www.gns.cri.nz/ <http://www.gns.cri.nz/> I Email: C.Williams at gns.cri.nz <mailto:your.email at gns.cri.nz>
>> 
>> _______________________________________________
>> 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 <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 <http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short>
> Charles Williams I Geodynamic Modeler
> GNS Science I Te Pῡ Ao
> 1 Fairway Drive, Avalon 5010, PO Box 30368, Lower Hutt 5040, New Zealand
> Ph 0064-4-570-4566 I Mob 0064-22-350-7326 I Fax 0064-4-570-4600
> http://www.gns.cri.nz/ <http://www.gns.cri.nz/> I Email: C.Williams at gns.cri.nz <mailto:your.email at gns.cri.nz>
> 
> _______________________________________________
> 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 <http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short>
> 
> <domain_no_gravity.png>_______________________________________________
> CIG-SHORT mailing list
> CIG-SHORT at geodynamics.org
> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short

Charles Williams I Geodynamic Modeler
GNS Science I Te Pῡ Ao
1 Fairway Drive, Avalon 5010, PO Box 30368, Lower Hutt 5040, New Zealand
Ph 0064-4-570-4566 I Mob 0064-22-350-7326 I Fax 0064-4-570-4600
http://www.gns.cri.nz/ <http://www.gns.cri.nz/> I Email: C.Williams at gns.cri.nz <mailto:your.email at gns.cri.nz>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.geodynamics.org/pipermail/cig-short/attachments/20161124/c23f08aa/attachment-0001.html>