Reaction Mechanism Issue (Therm.dat)
 

Hello,

I recently got the Mechanism from Andrae et al. for Toluene reference fuels, C&F 2008, containing 137 species and 633 reactions. I wanted to compare the results with the Jia TRF Mechanism, included in the multi-component fuel example case of the sector engine.
However, there were a few things preventing me from doing so:

The therm.dat file contained species information, that were missing the exponential format:
CH2CHCHCHC.H2
CH2CHCHCO
-> I added the exponent E within the file

Furthermore CONVERGE told me that there was missing species information for HONO, which was caused by a duplicate HNO species:
-> I deleted the duplicate information. The duplicate also had 5 lines instead of 4.

Also a few species were missing the gas phase indicator G. But this was only a warning.
-> I added the G manually.

Finally, CONVERGE told me that there is no gas species of my parcel species C7H8. This is because toluene is named as C6H5CH3 in that mechanism file set:
-> I replaced C6H5CH3 with C7H8 in both mech.dat and therm.dat.

Since I am no expert in the chemistry, I wanted to ask if someone else has run into similar issues? Was my procedure to repair the files correct?

At least CONVERGE is not reporting any errors right now, but i couldnt test the case as i dont have any capacities at the moment.

Thanks for help or further information.

edit:
The message log in Studio still reports an issue:
ERROR: Therm.dat malformed species C7H14O2H. Error high temp

Hi MFGT,
I do have to clean therm.dat in some instances the way you have done. Here are some tips to ensure that you have a clean therm.dat

a) Case Setup Module in Studio is a great help, as it fixes many errors and check for consistencies.

b) Try out the Chemistry Module in Studio as well and let it parse the therm.dat. And see if you get any new warnings or messages or not.

c) Finally, use the utility "cleantherm" to have it consistent with mech.dat. To run this utility, go to your working directory and type the following command :

converge cleantherm

It generates a new file, "therm_cleaned.dat" which needs to be renamed as therm.dat. We talk about this in our Manual.



We also talk about the numbers and characters that we expect to be there with each species, that define species name, elemental composition, low temperature, common temperature, high temperature, end of a line etc.


We expect the following in in line number 1 for species C7H14O2H
C7H14O2H 7/23/98 C 7H 15O 2 G 300.000 5000.000

By the error log from Studio it looks like, there is a non-numeric character instead of the high temperature value (bold-faced above). Change it to a numeric value, say 5000.

Have a look at our tutorial cases in the downloads website, and go through a typical therm.dat file.

Anyhow, keep posting your varied experiences and let the Forum know how it goes for you as far this issue is concerned.

Thanks

Hi Saurav,

the species looked like that:
If i replace the first line with yours, its not reporting the error anymore. I dont know if its because of the common temperature of 1392.00 which is added after the high themperature value?
But basically every other species has a common temperature added in that row.

Furthermore, i think there should be only 14 exponential numbers, but some species have 15?

Hi MFGT,

The spaces in between the numbers matter. Refer to the therm.dat in our tutorial cases. If the numbers/characters are in different columns, then there would be a parsing issue.

Secondly, by default the common temperature for any species will be the one that you specify at the very begining of therm.dat file. If you wanted to specify a different one for a particular species, as you did (1392), just make sure the spaces in between the numbers/characters are correct.

We read 14 exponential numbers, more specifically 4 numbers in line 4.

Thanks

All the hard work for nothing, as it turned out replacing the Jia mechanism with the Andrae mechnaism increased my computational time by a factor of 2.4 :eek:

At the meoment, i can not simply have these large computational times, so i am forced to stick to Jia mechanism (or others of similar size).