lets say someone obtains poor results using a particular grid and attributes the poor results to the turbulence model. if i get good results using a much finer grid, can i then say the first person is wrong to blame the poor results on the turbulence model or is the performance of a turbulence model a function of grid size? if one has a very very fine grid, does that follow that good results will be obtained with any turbulence model on this grid?

yes fine grid allows to get good or better results ..and after some time the effect of makign grid finer will eb negligeable on results .... but still other person is wrong in blaming turbulence for bad results (eventhough he is unknowing right partially) for two reasons: 1. even coarse grids give substantially good results if grids are very good (not turbulence alone control the results) 2. if he doesn't even know the effect of grid quality on results he might have done some more mistakes in setting up case , which may lead to poor results (blaming turbulence model alone is wrong in this case too)

Hi,

Yes as you state in your problem, there are two things,1) choice of the turbulent model 2) grid size If you have chosen the right model specific to your problem then you should be able to get grid independent results, i.e the variation of results between two grid sizes should be less, I am not sure about the Percentage, but me and my friends think that anything close to 5-7% is fine. Hope this answers your question. Aravind