[CIG-SHORT] One question & one small pylith bug

[Brad Aagaard]

Shaoyang,

If you want a time-dependent viscosity, then the most appropriate way to 
implement that would be to create a new material object that has 
additional parameters to define the temporal variation of the viscosity. 
See templates/materials for an example of how to implement a new 
material object. You may want to see if the generalized Maxwell or 
power-law rheologies already implemented are sufficient for your needs. 
You should think about what models are physically meaningful too.

Regards,
Brad


On 08/04/2015 10:23 AM, Shaoyang Li wrote:
> Hello Brad,
>
> I want to make a Maxwell viscoelastic model with varying viscosity in time,
> which could be read from user pre-defined time-viscosity list.
>
> I am not sure how hard of this implementation in Pylith.
> Or there is an easy way to achieve it?
>
> Now I am thinking two strategies:
> 1- run viscoelatic model for a time length, export the stresses etc. to next
> viscoelatic model with same configuration but only different viscosity.
>
> 2- make my own time step object that can re-read time step and viscosity
> spontaneously.
>
> Maybe both of them are not easy at all,
> could you please suggest some doable way to go?
>
> Anyway, I started to read codes about power law rheology in Pylith.
>
> In python script, powerlaw_dendb.py, line 143, according equation 5.76
> in manual, reference strain rate should equal reference stress**n * At,
> not "/ At" there in the code.
>
> cheers,
> shaoyang
>