Liquid phase with RMG4 and RMG3 #297
Description
Hello,
I am starting my work on liquid phase and I have started with RMG-Java instead of Python (because of the constant concentration issue). However, I found something strange between RMG4 and RMG3.
As you mentioned on the RMG website, calculation of the enthalpy of solvation with RMG3 is incorrect.This problem should appear with temperatures different than 298K. However, I did a simple comparison of the liquid phase example (octane autoxidation) at 298K with RMG3 and RMG4, and at the end I don't have the same thermochemistry with the same species. See the following differences of the 2 chem.inp files.
Just cheking species coming from the GRI-MECH library :
you can see that they are different (but they shoudn't at 298K for the reasons explained before. Correct me if I am wrong).
So I made a diff between some directories of my 2 RMG versions to check that my libraries and groups have the same values. This is True for :
thermo_libraries/SolventLibraries/Library.txt
thermo_librairies/GRI-MECH3.0/Library.txt
thermo_librairies/GRI-MECH3.0/Dictionnary.txt
thermo_librairies/primaryAbrahamLibrary/Dictionary.txt and library.txt
thermo_group/Group_library.txt (except sulfur containing groups)
Then when you just look at the kinetic parameters obtained for the first reactions they are very different. Energy barrier heights drastically change from E+10 to E-10 for reactions involving reactants. Are those changes in agreement with RMG4 improvements?
I tried to check some kinetic libraries and groups but there is too many changes (new families, new groups...).
This has a huge impact on the mechanism obtained with RMG4.
I checked with several solvents and I obtained the same RMG4 behavior from octane to hexadecane at 5 different temperatures.
In addition, RMG3, which is supposed to not work properly at temperatures other than 298K, is able to reproduce more realistic fuel consumption delays compare to RMG4 mechanisms which are several orders of magnitude too long then too short after a given temperature. It seems there is something wrong with the solvation in RMG4. What do you think?
Did you ever report this issue before. Do you think that it is possible to find it in the RMG-Py version after the constant concentration issue is solved?
Thanks