CFD Online Discussion Forums

CFD Online Discussion Forums (
-   FLUENT (
-   -   Convergence issues for modeling of flow over a car (

blin January 10, 2010 19:31

Convergence issues for modeling of flow over a car

I am an amateur at this and would like to ask some CFD gurus for some help :).

I am trying to model flow over a car body in Fluent (actually, I'm only modeling half of the car split upon its symmetry plane). I have chosen a spalart-allmaras model, with a turbulence intensity of 0.21% and a turbulence length scale of 0.16m.
My discretization values are as follows:

Pressure 2nd Order (0.2)
Momentum QUICK (0.6)
Pressure-Velocity Coupling SIMPLE
Turbulence QUICK (0.7)

The solution converges to residuals of 0.001 when I run it using first order upwind (default) setting. However, when I change it to the above setting the solution does not converge.
I have tried running it using first order upwind, and then change the discretization parameters to the above setting and continue running it without initalization, but that does not converge as well.

I'm really stuck as to what I can do further.

Also, I'm getting a bunch of warnings that says "turbulent viscosity limited to viscosity ratio of 1.000000e+05 in xxxx cells", with the number of cells increasing. I have read on the forums on discussions regarding this topic, but none relating to how to solve it in the spalart-allmaras model, so I'm wondering as to the solution to that as well.

Thanks in advance for any input.

benji2642 January 11, 2010 06:01

What kind of data do you want to extract from your simulation?
Do you just want to visualize the flow?

Why did u choose the QUICK parameter?

blin January 11, 2010 11:58

I want to visualize the flow, along with gather data on the velocities and pressures around the car.

I chose QUICK because it was according to an similar example that I have seen in literature.

All times are GMT -4. The time now is 04:07.