[CIG-SHORT] Question - Dynamic Analysis

[Brad Aagaard]

Farrokh,

In nearly all cases, you shouldn't have to change the tolerance for the 
minimum Jacobian. Instead you should be adjusting the scales used to 
nondimensionalize the problem so that the cells are within an order of 
magnitude or so of nondimensional length of 1.

For dynamic problems, you should be setting the minimum period (maximum 
frequency of the propagating waves) and wavespeed to get the correct 
scales. In spatialdata/units/NondimElasticDynamic.py you can see that 
the length scale is vs * minperiod and the pressure scale is 
vs**2*density (shear modulus).

Regards,
Brad



On 05/12/2016 03:53 PM, Farrokh Sheibani wrote:
> Dear Dr. Aagard,
>
> I work for Dr. Brad Hager in EAPS/MIT. I was wondering if we can set the
> minimum permissible value for Jacobian determinant. The default in
> Pylith is 1e-6, although my mesh has some small cell with det[J] in the
> order of 1e-7 (but in a smoothed mesh and with positive det[J]). I am
> trying to model dynamic rupture in triaxial sample scales. I appreciate
> if you give me some advise regarding this. I am trying to model a sample
> under traiaxial test with a pre-existing natural fracture.
>
> The dynamic analysis doesn't  have Length Scale so I was trying to solve
> the problem by changing density, boundary conditions and shear and P
> wave velocities as well the geometry, although still I get the same result.
>
> Thank you in advance!
>
>
> With best regards,
> Farrokh
>
>
> Farrokh Sheibani, PhD
> Postdoctoral Research Associate
> Department of Earth Atmospheric and Planetary Sciences; Earth Resources
> Laboratory
> Massachusetts Institute of Technology
>
>
>
>
>
>