[CIG-SHORT] Initial stresses for large model

[Brad Aagaard]

On 01/28/2017 08:13 AM, Demian Gomez wrote:
> Hi Brad,
>
> Thanks for your answer. I was thinking of maybe making a two step
> problem where at time 0 it would calculate the initial stress and then
> at time 1 put in the slip at the fault. Sort of a pseudo-earthquake
> cycle but with two quasi-static simulations and no time dependent
> problem between them. Do you think this might work?

Using a simulation to calculate initial stresses and then another to 
move forward in time is a good, general strategy. Depending on what 
constitutive behavior you use (viscoelastic, etc) you might need to spin 
up the model to get to steady state conditions to use as your "initial" 
state.

Regards,
Brad