[vegardh]

I'm working on setting up a case with internal transonic/supersonic flow in rhoSimpleFoam in OpenFOAM v8. The goal is to assess the capabilities and limitations of rhoSimpleFoam in this flow case, compared to more appropriate solvers.

It seems logical to enable the transonic option in fvSolution, but I'm not really sure what this option actually implies physically. I've been trying to interpret the source code in pEqn.H, but if anybody has a quick answer to what the SIMPLE/SIMPLEC-algorithm does differently when solving the pressure, using the transonic option, that would be of great help!

Best regards,
Vegard

[mAlletto]

The transonic option enables a divergence term in the pressure equation which transforms the pressure equation from elliptic to hyperbolic. For more information see the book https://www.springer.com/gp/book/9783319168739 and this wiki https://openfoamwiki.net/index.php/RhoSimpleFoam

[dlahaye]

My current view is that of a solver hierarchy in the following three steps:

1/ simpleFoam: stationary isothermal incompressible flow

2/ buoyantSimpleFoam: stationary non-isothermal incompressible flow (includes energy equation, excludes high Mach number effects)

3/ rhoSimpleFoam: stationary compressible flow (includes both energy equation and high Mach number effects)

The transonic switch set to false excludes transonic effects, and reverts rhoSimpleFoam back to buoyantSimpleFoam. This is nice to have because the transonic effects include a non-symmetric term in the pressure equation. This renders the pressure equation more expensive to solve.

To me it was instructive to run the squareBend tutorial with lower and high inlet velocity and with the transonic switch set to both true and false.