Hi Guys,

I ran a simulation in Fluent on simulated wind (velocity inlet) hitting a cubic structure inside another cubic volume. Total number was 1.7 million cells. Using standard KE model with all default settings and unstructured mesh my simulation has coverged after 1200 iterations but I am not sure of the results. All the residuals were fine except epsilon was oscillating a lot especially towards the end.

All other final parameters such as mass flow, velocities, pressures seem to be reasonable, I have checked them against experimental.

Any ideas on how to improve? The gemoetry won't allow a structured mesh, I have tried.

Thanks a bunch

Magic

Oscillating residuals are typically a sign of numerical dispersion or, in this case, a sign that your turbulent mixing is having difficulty stablizing. What values or y+ are you seeing in your model? For a standard KE model you would want to shoot for 30>y+>300. Also make sure your grid is fine enough to resolve your turbulent mixing in regions of interest.