[Plum_Flavoured_Biscuit]

Basically,



If I could be directed to a Forum where CHEMKIN 18.2 is discussed in detail that would be great, but basically my problem is this.


I have run, experimentally, an axissymetric Counterflow flame that was fed by 3 SLPM CH4 and 5 SLPM air. Rich flame, I know, but it burns. However, I can simulate this in Cantera using the GRI 3.0 Mechanism. It's CHEMKIN that's the problem. As far as I am aware, for nozzles that are 11.2mm ID and a gap of 11.2mm (for strain unity) with a combustion pressure of 1 atm with inlet temperatures of 300K I should get the following units to insert into CHEMKIN:


FUEL (CH4) --> 3 SLPM = 3.58500E-05 kg/s = 0.364 kg/m^2.s = 0.508 m/s for that nozzle.


Oxidiser (Air) --> 5 SLPM = 1.07750E-04 kg/s = 1.094 kg/m^2.s = 0.846 m/s for that nozzle.


It's a strained flame, sure. But if you take the example for an axissymetric flame on the Cantera website http://www.cantera.org/docs/sphinx/h...ion_flame.html and do a few switches of flows, It runs for that Chemical Kinetic Mechanism, which is the GRI 3.0 Mechanism.


If I try this in CHEMKIN 18.2, I get a flat line. If I use the Cantera example flows, 0.72 kg/m^2/s air and 0.24 kg/m^2/s fuel (but CH4 instead of C2H4) I get nada as well, no combustion.



I can however make this work with flows 1/5 of the ones above, for a gap of 2cm. It has worked for 0.048 kg/m2-sec fuel and 0.144 kg/m2-sec air (air being 78% N2, 21% O2 and 1% Ar) but is being difficult where it should work.


First thoughts: My flow numbers are wrong. Those Mass Fluxes used for Chemkin are wrong.



Second thoughts: kg/m2-sec (chemkin) isn't the same as Kg/m^2/s (cantera.)


Now, look, I understand that it's unlikely anyone will read this, But if in the very least, can I be directed to somewhere that deals in CHEMKIN 18.2?
Thanks to those of you who have looked.
Regards T

[Plum_Flavoured_Biscuit]

To those that might be led here...

I'm still having trouble, but it's been made significantly easier by me realising something that Cantera does automatically, but Chemkin does not.

...Make the grid out of more points. The Default is 6. Try 20, try 50. If it solves the flame on 10 points, even if there's no curve, Chemkin rolls with it.
Cantera does this automatically, Chemkin does not.

Lesson half-learned so far.

If anyone finds a Chemkin-Forum, please let me know.

[Kummi]

Hello Tommy,
I am not familiar with your problem. However, after reading your contents, I came to know that you are dealing with CHEMKIN GRI mechanism by simulating in CANTERA.

Have you ever tried to solve your problem in CHEMKIN (software) itself, because there are many kinetic reaction models for detailed study of the flame chemistry and structure in 1D. Because once if you solve in CHEMKIN, you will have an idea about the basic flow behavior with validated reaction rates and kinetic parameters.

While surfing, I have gone through the document - attached here for your reference,

https://thesis.library.caltech.edu/1500/3/thesis.pdf

Good luck

[Plum_Flavoured_Biscuit]

Hi Kummi

No worries, finding people who are familiar with Chemkin has proved less than interesting. For an industrial tool, it's not too well discussed online.

I have solved the problem in CHEMKIN, and that's what led me to make the post in the first place, as CANTERA would solve, but CHEMKIN would not.

However it turns out that I was missing something - In Cantera, if a flame solves with a flat line, Cantera tries again with more points. Chemkin just runs with it. Hence, increase the first grid from the default 6 points to say, 20, or 50, and a lot more combustions will work and emerge with temerature peaks for CFC flames.

Thank you for the Caltech Document, it is appreciated.

Regards
T