[Dan1788]

Hello all,

I was studying the ReactingFoam reaction mechanism in OF 2.2. Instead of having the usual mechanism as given in the tutorial, I included a new one step mechanism given by Westbrook and Dryer (1981) as shown below:

species
5
(
CH4
O2
CO2
H2O
N2
)
;

reactions
{
un-named-reaction-0
{
type irreversibleArrheniusReaction;
reaction "CH4^-0.3 + 2O2^1.3 = CO2 + 2H2O";
A 1.3e+09;
beta 0;
Ta 24354.5;
}
}

Unfortunately however, when I use this mechanism I get an error that is:

--> FOAM FATAL ERROR:

Singular matrix
From function LUdecompose(scalarSquareMatrix& matrix, labelList& rowIndices)
in file matrices/scalarMatrices/scalarMatrices.C at line 60.

FOAM exiting

Can someone please let me know why I am getting this message? Is there something wrong with the format of the mechanism or am I missing something?

Would really appreciate help. Thanks!

[francescomarra]

Dear Daniel,

from my experience, the Westbrook & Dryer mechanism (W&DM) you are trying to use is not so "easy" . I am not able to understand, from the few rows of your output, when the error occurs: when reading the mechanism of during the computation ?

The problem with the W&DM arises from the negative exponent of methane. To check if this is the problem, you could try to lower the preexponential factor and/or use a unitary exponent for the methane. Of course results are not meaningful, but you can check if you are correctly programming your reaction and if the reactingFOAM code is able to deal with negative exponents of species.

The literature offers several other one-step global and irreversible mechanisms, usually valid for a restricted range of equivalence ratios, but with unitary exponents for the concentrations.

Best regards,

Franco

[Dan1788]

Hi Franco,

Thanks a lot for your message.

The error message I posted was coming at the start of the simulation, while reading the mechanism. I changed the methane exponent from -0.3 to 0.3 and the simulation worked fine! Any idea why this happens?

Thanks again!

[francescomarra]

Dear Daniel,

unfortunately I am not very familiar with the OpenFOAM implementation of the chemical terms. However the problem usually arises because of the behaviour of any power function with negative coefficient. When the argument keeps close to zero, the function tends to infinity. Then, or the production term explodes (making unstable the whole computation) or at zero you have a NaN value.
Actually, it is known that the W&DM is correct, but it takes into account only a planar flame, a idealized situation where it easier to keep the correct relative importance of both reactants. Here comes into play the actual implementation of the mechanism.

Best regards,

Franco

[Howard]

Hi, Daniel. I met the same question. Have you solved the problem?

[fabri]

Hi Daniel and Howard,
I met the same problem too.
Have you solved it?

Thank you very much
Fabri

[ordinary]

Hello,

I have also same problem.