Fame uses fixed grain size(?)

[shamelmerchant]

I am wondering why we have fixed grain size when we write the fame input file

input.append("GrainSize J/mol 4184");

in https://github.com/GreenGroup/RMG-Java/blob/master/source/RMG/jing/rxn/FastMasterEqn.java

even though we calculate the grain size before writing the input

double grainSize = getGrainSize(grainMinEnergy, grainMaxEnergy,
            pdn.getNumUniIsomers()); // [=] kJ/mol 

[rwest]

Um, it does look like a mistake. Introduced in 815a5ed294f86687da78b25ff15ae303689fb2b0, before we had the ability to change it. It may not be as bad as it first sounds, as I think FAME will read both the number of grains and the grain size specification, and use whichever gives the smaller grains (most accurate result). It does mean the current implementation will not let you use grains larger than 1kcal, but at least the whole "increase number of grains and try again" thing should be working.

[shamelmerchant]

Okay, the reason I stumbled upon this is because I have a network which is failing with a negative steady state concentration error and wanted to specify a grain size of 0.5 kcal/mol.