[nfurkan]

Hello all,

I am trying to model a system involving three reversible reactions (6 components and 2 inerts) according to an article in a channel on Fluent 19.1. I have already written the diffusivity and kinetic theory UDFs and the model works well at 493K with 0.023 m/s velocity magnitude. The results I got perfectly match the results MATLAB code gives, which is written specifically for the same subject.

When I, however, increase the temperature to 503K, Fluent simply does not work and gives divergence error; float point exception. No matter what I tried, I cannot solve the problem. It also works when I increase the velocity magnitude to 0.2 m/s at 503K, but then I cannot get enough conversion.

Here are some information about the model;
-I disable the energy equation, so it is an isothermal system.
-Hybrid initialization does not work with the case, so I use standard one with catalyst inlet as the reference conditions.
-BCs and other parameters are set correctly since it works perfectly at 493K and temperatures below.
-Mesh minimum orthogonal quality is 0.9967 and cell squish is 6.6E-3.
-I use SIMPLE method, all other parameters are at their default values.

Here are what I tried so far;
-I tried to reduce under-relaxation factors, it did not work.
-I tried different methods, it did not work.
-I tried high order term relaxation at both 0.25 and 0.10 values, it did not work.
-I tried transient solver, it did not work.
-I tried different cycle types, it did not work.
-I tried longer channel to allow higher speeds as high as 0.10 with identical mesh quality, again did not work.

I do not understand why a complex model works perfectly at some temperature but does diverge at 10K higher one, and I cannot find a solution to it. I have been searching forums quite a long time but could not find a similar situation yet.

If you have any idea how I can overcome the problem, it would be much appreciated.