Convergence issues

[Timothy Simmons]

Dear all,

I am using GPyro (0.8186), and am encountering the seemingly common issue of "Gas Yj not converged" at a timestep. It reduces the timestep repeatedly until it reaches 0.5e-5, and then does does reduce the timestep any further and essentially fails. How can I reduce allow it to reduce the timestep below this value, I presume that there is a "minimum timestep" type parameter, but haven't been able to find it?

Any help would be appreciated.

Thanks,

Tim

[Timothy Simmons]

To add some context, 

I initially tried recreating Problem 6 in the latest version of the workbook (1D pyrolysis of PMMA), and that worked without an issue. I am now trying to test some experimental results which use significantly higher flux (4MW/m^2) for 3 seconds (Paper is Scale Dependence of Material Response at Extreme Incident Radiative Heat Flux (2018), Alex Brown et al). I have attached my input deck. 

An additional question I have regards the outputs. In the workbook (Problem 2, but elsewhere as well) there is a plot with "Pyrolysis rate" on the y-axis, how can I output this quantity, and could anyone explain what the  "M/M0" and "MPPI" outputs are, are these the "fraction of initial remaining at current timestep" and "mass of each condensed species remaining" respectively?

Thanks,

Tim

[Chris Lautenberger]

Tim,

 

I’ve moved Gpyro from a locally-hosted Trac/Subversion site to GitHub:

 

https://github.com/reaxfire/gpyro

 

There are pre-compiled Windows binaries in the repository so please try the current version (v0.8200) and see if you encounter the same convergence issue.  If you do, submit a ticket on GitHub and attach your input file(s) and I’ll take a look. Thanks.

 

Best,

 

Chris

--
You received this message because you are subscribed to the Google Groups "Imperial Gpyro Working Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gpyro-working-g...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/gpyro-working-group/a20d3227-25bd-4ea6-9efb-2b2e58f4e9d6n%40googlegroups.com.

[Timothy Simmons]

Hi Chris,

Thanks for getting back to me, and for providing me with the link to the git repo. I generally work on Linux, and there appears to be an issue with the make file - when trying to run "make_gnu.sh" it cannot find any of the targets, but trying on the previous version of 8.1 it does work, is there a difference between the two in terms of how to compile it on linux? 

That said, I have run it on a windows machine, and it ran successfully, though it had to reduce the timestep to 6e-5s on several occasions sas the gas phase wasn't converging. What are the differences in the new version?

Thanks

Tim

[Chris Lautenberger]

Git doesn’t store empty folders so even though the build folders were committed they’re not in the Git repository.  It’s an easy fix though, I just modified the build scripts to create the build directories manually. Long story short, if you clone the current Git repository, Gpyro should build under linux. If you encounter any problems submit a ticket.

 

The only thing I’ve changed is to relax the gas-phase mass fraction sum constraint which is what usually causes non convergence problems.

 

Best,

 

Chris