I am solving a numerical modelling stem of autothermal reforming using MATLAB. The problem is ,Pressure and Temperature results are not stable (unsteady state problem) and escalate too much I cannot even solve my species equation. It is a 1-d pseudo homogenous fixed bed model using the matrix approach .I am using finite difference to solve my systems and most of my terms are discretized explicitly.
Any explicit problem, especially heat transfer, has a stability limit in terms of time step, which depends on your discretisation, eg your "element" or grid size. That is the most common reason for diverging results (I assume that is what you mean when you refer to "escalate too much").
I recommend a 3 step approach to fix this problem:
1) Simplify your physics to something really simple, where you can check that there are no obvious issues/typos with the solution approach
2) Gradually add back in more advanced features and see where solution starts to break down
3) Once you identified what causes the problem, try to fix it. If the physics are OK, start changing your time step and mesh size to stay within the stability limits.
I do understand this is a fairly generic approach I described, however it is a good approach to debug any type of simulation model. Hope it helps!
Best regards
David
- Badges
- Science topic