[CIG-SHORT] problem about setting the spatial database

[Brad Aagaard]

Tu,

I strongly recommend using a initial stress and friction model 
parameters that are continuous. If you use fields that have steps in 
them, the solution is often resolution dependent. A Gaussian is a 
reasonable option for imposing a continuous stress field. In such a 
case, I would create a grid for the entire fault using a SimpleGridDB.

For an example, see 
https://github.com/geodynamics/pylith_benchmarks/blob/master/dynamic/scecdynrup/tpv102/create_friction.py. 
See http://scecdata.usc.edu/cvws/tpv101_102docs.html for the description 
of the parameters.

Regards,
Brad


On 3/11/18 5:48 PM, tu xiang wrote:
> Dear Charles,
> Yes, it is better to write a script to generate the database. However, I
> do not know the process of writting the script because I do not know how
> to use the actual fault coordinates to set up the desired pattern.
> For example,
> If the nucleation zone is a circular area,  do I need to discretize the
> circular area into a grid of (x,y) coordinates and write the coordinates
> into the database? The value of the rest area of the fault which is
> outside the circular area will default to be zero?  Will the values of
> slip\stress from zero to the circular area change linearly?
> looking forward to your reply
> Best regards,
> Tu Xiang