fireFoam/reactingFoam with ISAT results in processor heat up
 

I have been testing fireFoam & reactingFoam for use with ISAT (TDAC with no reduction) on a uniformly resolved, static mesh based on the smallPoolFire2D tutorial and noticed that the data on one processor heats up differently than all the others. I have also tried performing these runs in serial, where the entire domain heats up accordingly (unphysical for the entire domain to instantly heat up). I have included a screenshot of what this looks like. The following details are the differences from the tutorial:
- 2mm resolution
- laminar combustion and turbulence model
- GRI mech 3.0 + TDAC (only tabulated chemistry - ISAT)

I have ran this in OF7, OF9, OFdev and v2012, with fireFoam (buoyantReactingFoam) and reactingFoam - both show the same result. However, when I turn TDAC off, all 4 trials work fine. I have attached my chemistryProperties file from an OF7 run.

More than happy to also upload the full case for anyone willing to take a gander.

issue ended up being ISAT tolerance too low (3e-3 moved to 3e-6), although I am still at a bit of a loss why the lower tolerance on the tabulation accuracy led to one processor deviating in solution from the others so significantly