alex
|
October 15, 2008 16:44 |
Re: flovent vs fluent
the fact that solution converged, has nothing to do with a much more important question of what it is it converged to.... by adding diffusion (via discretization, relaxation, viscocity dumping, turbulence model or whatever your particular solver calls it) you can converge anything to anything in as many iterations as you want. Why don't you find some experimental data, preferrably for a simple case, preferrably not available to that specific sales person you are talking to and ask them to reproduce the data in their simulation. Then your decision becomes very simple, error.vs.time to run for your specific application....Just from experience, buoyancy driven flows are hard to converge and more often than not have no steady-state....., but the Fluent person has probably already pointed that out...you must build models with a small number of cells if you cannot run stuff in parallel though....
|