[CIG-SHORT] R: pressure & pylith

[Brad Aagaard]

Alberto,

Please ALWAYS send questions about PyLith to cig-short at geodynamics.org 
instead of one of us individually.

If your test case 1 worked, then there is probably an error in how you 
are setting up test case 2. If you need to specify information 
cell-by-cell, you should use the integration points, not the cell 
centroids as the stresses are evaluated at the integration points. 
PyLith uses 9.80665 m/s**2 as the default value for the acceleration of 
gravity. If you compute your initial stresses using 9.81 m/s**2, then 
you should change PyLith's default value to 9.81 m/s**2. Other things to 
check include the type of spatial distribution (use data-dim = 3 for 
points in a volume; if you have a logically rectangular grid, then use a 
grid spatial database (SimpleGridDB) for improved performance) and all 
of the other information in the spatial database header.

Regards,
Brad


On 10/24/2016 12:36 PM, Cominelli Alberto wrote:
> Brad,
> We have been working on a test-case where we simulate a 3D system with depth going from zero to 2000 m, with a simple hexahedral grid  of nx * ny * nz elements describing a slanted fault.
> Having in mind the need to fill pilitth with fluid pressure derived stresses, we uprepared two cases:
> 1)  Initialisation with sigma_zz=rho * g  z (rho = 2500 kg/m3 - g = 9.81) , with two values at z=0 and z=-2000 and linear interpolation to sample the database.
> 2) Initialisation with sigma_zz=rho * g  z (rho = 2500 kg/m3 - g = 9.81) , with nx * ny *nz values given at element centroids , plus values at the boundary surfaces.
> Solution should coincides with initial conditions, but if it worked with 1) , we could not make pylith working with 2).
> Actually we modified a bit 2.1.0 code to check the values in the integration points and they look strange.
> I wonder if we can submit the problem to you/your group.
> Regards,
> Alberto.
>
>
>
>
>
> -----Messaggio originale-----
> Da: Brad Aagaard [mailto:baagaard at usgs.gov]
> Inviato: martedì 18 ottobre 2016 16:54
> A: Cominelli Alberto
> Oggetto: Re: pressure & pylith
>
> Alberto,
>
> At the beginning of a PyLith simulation, you can specify initial stresses. I don't know if this is sufficient for your needs.
>
> Regards,
> Brad
>
>
> On 10/17/16 11:17 PM, Cominelli Alberto wrote:
>> Dear Brad Aagaard,
>>
>> I am willing to use Pylith for some couple fluid-flow/geomechanical
>> simulations. In my case I want to run a fluid flow simulator first and
>> then pass pressure on an element basis to Pylith in a one-way mode.
>>
>> I wonder if this  possible with the current pylith versions w/o
>> modifying the code. I am digging in the documentation but so far I
>> could not find a way.
>>
>> Looking forward to hearing from you soon I thank you in advance for
>> any hints.
>>
>> Regards,
>>
>> Alberto.
>>
>>
>>
>>
>> *Eni* SpA
>> Sede Legale
>> Piazzale Enrico Mattei, 1
>> 00144 Roma - Italia
>>
>> Capitale sociale
>> euro 4.005.358.876,00 i.v.
>> Codice Fiscale e Registro Imprese di Roma n. 00484960588 Partita IVA
>> n. 00905811006 R.E.A. Roma n. 756453
>>
>> Sedi secondarie:
>> Via Emilia, 1 e Piazza Ezio Vanoni, 1
>> 20097 San Donato Milanese (Milano) – Italia
>>
>> eni.com <http://www.eni.com>
>>
>> ----------------------------------------------------------------------
>> -- Message for the recipient only, if received in error, please notify
>> the sender and read http://www.eni.com/disclaimer/
>
>
> ------------------------------------------------------------------------------------------------------------------------------------------------------------------------
> Message for the recipient only, if received in error, please notify the sender and read http://www.eni.com/disclaimer/
>