[CIG-SHORT] TimeStepUser

[Birendra jha]

Hi,

I am trying to use TimeStepUser feature of Pylith. Is there an example of using variable, user-specified time step simulation? I intend to read some files at these specific timesteps and want Pylith to only simulate at these steps. In my test.cfg I have:

[pylithapp.problem.formulation]
time_step = pylith.problems.TimeStepUser 
[pylithapp.problem.formulation.time_step]
total_time = 16.49555*year
filename = timesteps.txt
loop_steps = false 

I put the timesteps.txt in the same folder from where I call "pylith test.cfg" which is /build/pylith/pylith-1.6.3/examples/3d/hex8

I see that the simulation enters ElasticityImplicit::integrateResidual() with t=0.0027397 which is 1 day in dimensional unit. Isn't the Elastic prestep from -1 yr to 0 yr? 
After the simulation completes this timestep it just exits without going through my timesteps.txt. What am I doing wrong? 

timesteps.txt file contains:

units = year
0.999315537
1.002053388
0.999315537
0.999315537
1.037645448
0.963723477
0.999315537
0.999315537
0.999315537
1.002053388
0.999315537
0.999315537
0.999315537
1.002053388
0.999315537
0.999315537
0.495550992

I tried with unit = day and corresponding numbers also.

Thanks a lot
Birendra

--- On Tue, 4/10/12, Brad Aagaard <baagaard at usgs.gov> wrote:

> From: Brad Aagaard <baagaard at usgs.gov>
> Subject: Re: [CIG-SHORT] Rate and state friction
> To: cig-short at geodynamics.org
> Date: Tuesday, April 10, 2012, 2:59 AM
> On 04/09/2012 01:11 PM, Hongfeng Yang
> wrote:
> > The current development plan says that rate and rate
> friction needs to
> > be updated to read in prescribed stress conditions in
> order to run SCEC
> > dynamic benchmark problems. I wonder when this is going
> to happen? In
> > addition, is there a way to read in output results from
> interseismic
> > simulations and to run dynamic rupture problems?
> 
> The current development plan mentions adding a Nucleation
> object with a
> spatial and temporal perturbation of tractions from the
> initial value. 
> This type of variation is required for the SCEC dynamic
> rupture 
> benchmarks that use rate- and state-friction. This is listed
> as a likely 
> feature for version 1.8 that has a target release date of
> Fall or Winter 
> 2012.
> 
> You can create a spatial database of initial tractions from
> the final 
> fault tractions of a quasi-static simulation. This involves
> creating the 
> header and writing out the values from the output file. This
> is most 
> easily done with HDF5 output.
> 
> Regards,
> Brad
> _______________________________________________
> CIG-SHORT mailing list
> CIG-SHORT at geodynamics.org
> http://geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>