Hi RK,
Thank you for your response. I am currently using 2020 R2. I noticed the following three things:
1- When opening the chemkin files, I get a message that "default tolerances have changed to improve accuracy" and I am asked whether I want to keep the old tolerances or not. (So I think partially it may be due to the tightening of the tolerances). Do you by any chance have the log for what the old tolerances used to be?
2- I noticed it happens for one of the kinetics models (AramcoMECH 1.3), but it is not a function of size (# of species/# of rxns) as it is not as bad for Aramco Mech2.0 or AramcoMECH 3.0, whereas the latter two are larger models.
3- With AramcoMECH 1.3 I also get several (8/28) failed runs, and I am forced to change the max simulation time (some times more than once) and rerun those data points to get them to converge.
I haven't tested the reproducibility yet.
Regards Ashkan