[CIG-SHORT] Implicit to Explicit formulation for quasi-static to dynamic simulation

[Brad Aagaard]

Birendra,

PyLith does not currently support changing the time stepping formulation 
in the middle of a simulation. Doing so requires not just changing the 
time stepping object but swapping some of the underlying objects (for 
example, the finite-element integrators ElasticityExplicit to 
ElasticityImplicit) and making sure everything associated with the "new" 
time stepping formulation is initialized properly. This is mostly 
bookkeeping but it is certainly not trivial.

Implementing this in PyLith is high on our TODO list and is planned for 
this year. The earthquake cycle features have gotten delayed due to the 
major restructuring for v2.0, which is setting things up to help us push 
forward on many fronts.

Regards,
Brad


On 08/29/2013 11:22 AM, Birendra jha wrote:
> Hi all
>
> Is it possible to change the formulation from Implicit to Explicit during the simulation? I am modifying the example in
>
> /examples/bar_shearwave/quad4/spontaneousrup_slipweakening.cfg
>
> such that the rupture happens after some time in the simulation. I want to switch from the Implicit formulation to the Explicit formulation during the dynamic rupture phase.
>
> Is it possible to specify a switch time for this, or probably a criterion on the difference between the fault strength and the shear traction such that dynamic simulation begins just before rupture begins? This will save a lot of time compared to using Explicit time step of less than a second during the quasi-static phase.
>
> If there are other ways to achieve this, please let me know.
>
> Thanks and best regards
> Birendra
> _______________________________________________
> CIG-SHORT mailing list
> CIG-SHORT at geodynamics.org
> http://geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>