[cfdblender]

Running an ground vehicle aerodynamic simulation and everything was fine until changed from SEGREGATED flow and energy to COUPLED flow and energy - then every simulation diverged, no matter what I tried (mesh improvements, different turbulence models, etc) ... until I switched back to segregated and it works again.

But why? Could it be the Mach number? (my simulation runs at M=0.175)

[FMDenaro]

Originally Posted by cfdblender

Running an ground vehicle aerodynamic simulation and everything was fine until changed from SEGREGATED flow and energy to COUPLED flow and energy - then every simulation diverged, no matter what I tried (mesh improvements, different turbulence models, etc) ... until I switched back to segregated and it works again.

But why? Could it be the Mach number? (my simulation runs at M=0.175)

It could depend on the stiffness for low Mach number....you used a pre-conditioned coupled solver?

[cfdblender]

Originally Posted by FMDenaro

It could depend on the stiffness for low Mach number....you used a pre-conditioned coupled solver?

No, not using a pre-conditioned coupled solver? Would the pre-conditioning be done by running another simulation to obtain preliminary some data to configure the solver?

[FMDenaro]

Generally, the pre-conditioning is performend on the set of equations by changing the state vector.
For example in Fluent:

http://aerojet.engr.ucdavis.edu/flue...g/node1002.htm
http://aerojet.engr.ucdavis.edu/flue...g/node1003.htm