[CIG-SHORT] Time stepping to capture the growth of rate-and-state slip instability

Brad Aagaard baagaard at usgs.gov
Thu Mar 8 09:11:24 PST 2018


Ekaterina,

The adaptive time stepping scheme only uses bulk material properties to 
control the time step. We have not implemented feedback from the fault 
constitutive model to the adaptive time stepping scheme.

To have a variable time step size, I recommend using the user specified 
time stepping scheme. This will require you to know a priori when you 
need a finer time step. I think this corresponds to what you mean by 
prescribed time steps.

Regards,
Brad


On 03/08/2018 09:06 AM, Ekaterina Bolotskaya wrote:
> Dear Pylith developers,
> 
> I have a quasi static 2D simulation with a planar horizontal fault with 
> rate-and-state friction. My block is 9m by 6 m.
> I apply y-compression and shear traction to the block as initial 
> conditions and then try to observe the generation of rate-and-state slip 
> instability on the fault (when slip rate increases to 10s of meters per 
> second and then goes back to stable slip rate values).
> 
> I'm using adaptive time stepping:
> 
> [pylithapp.problem.formulation]
> time_step = pylith.problems.TimeStepAdapt
> [pylithapp.problem.formulation.time_step]
> total_time = 0.008*s
> max_dt = 0.00007*s
> adapt_skip = 0
> stability_factor = 1.5
> 
> But currently I don't think that my time step size ever changes: it 
> seems like it keeps the max_dt value throughout the simulation. If I 
> decrease max_dt further the simulation stops converging. Ideally I would 
> like my time step to be ~inversely proportional to my slip rate on the 
> fault, so that I can capture the instability in slip rate. Is there a 
> way to implement this or something like this?
> Should I create a file with prescribed time steps? Would that help?
> 
> Thanks!
> 
> 
> Best regards,
> Ekaterina Bolotskaya
> 
> PhD student in Geophysics,
> Earth, Atmospheric and Planetary Science Department,
> Massachusetts Institute of Technology
> E-mail. bolee at mit.edu
> Mob. +1 (857) 284-2805 <tel:%2B1%20%28857%29%20284-2805>
> +7 (963) 995-36-33 <tel:%2B7%20%28963%29%20995-36-33>



More information about the CIG-SHORT mailing list