Hi all, I'm studying the airflow around a vehicle (in 2D and 3D) for different inlet-velocity cases. Now, I'm working with the 2D case. The problem is that I'm not able to reach convergence. I'm using simple turbulence models (k-e standard, RNG and Realizable) and I only reach convergence if I use a 1st order scheme + low vel-inlet. I'm sure that the grid is ok (high quality, I believe).

The fact is that I've only used steady simulations. Can anybody tell me if it could be an unsteady case?

Fer

Hi Fer, What range of reynolds number u r using. If its high change to compressible solver.

hope this helps bharat

I think Reynolds isn't too high; well I'm trying with velocities in the range between 10 and 50 m/s. Well, the fact is that in some areas of the car the flux velocity is about 1.7-2*vel_inlet (easy to calculate aproximately using, for example, Bernoulli's equation). Could those zones motive that problems with convergence?

Fer

what are your other boundary conditions? For your 2D case you should get convergence with higher order schemes instead of just 1st order. Something is not right in your simulation.

Hi Fer. I'm having similar problems so I made a Mass-Weighted Average plot of the outlet velocity and it showed me a fluctuation instead of getting stable as one could expect if it would converge, so I am starting to think that I'm in an unsteady case.

But when I run an unsteady simulation, after 1500 iterations start to fluctuate periodically, and also the Mass-Weighted average of the outlet velocity: could this be a convergence criterion? All unsteady problems shows these periodic fluctuations? And those fluctuations means that I have arrived to the convergence?

ThAnxs to all!

Well, my boundary conditions are: - Static Wall for the car surface and for the ceiling. - Velocity-intlet at the inlet edge. - Outflow at the outlet edge. - Translational wall at the floor (moving at the same vel as the vel-inlet).

Perhaps there are problems using the outflow, but I don't know. In that case, tell me what's the most convenient to use in that edge, please.

Thanks

Fer

Use a pressure boundary condition instead of outlet. See if you get some better results