Classification and Selection of Solvers for low subsonic flows in CFD

udaych17 · July 24, 2019, 05:04

I am working on an external compressible aerodynamics flow problem which deals with speeds of Mach 0.45. I am uncertain whether to choose a density-based solver or a pressure-based solver. Generally accepted range from incompressible flows is below Mach 0.3 and Pressure-based solvers are considered best for incompressible flows but at speeds of Mach 0.45 will it be a good choice to still opt for Pressure-based Solver

? My understanding is that density-based solvers take more time and memory for solving the problem and hence I want to weigh my decision based on any useful feedback or suggestions.
Thanks in advance!

LuckyTran · July 24, 2019, 10:39

Both will work (and work very well). You can use whichever one you like.

There are no incompressible flows per se in Fluent.... it's solving the full compressible navier-stokes regardless of whether you use the pressure-based or density-based solver.

Pressure-based solvers are good for low mach number flows because density-based ones are (traditionally) unstable there without special treatment. Modern solvers have this treatment to stabilize the solver and you can use the density-based solvers for low Mach flows now too.

Pressure-based solvers also can be used for high Mach number flows (I have used it up to Mach 25). They can be less than optimal compared to density based solvers because the momentum-energy coupling is much more important in compressible flows and pressure-based solvers are (they are not bad per se but they are) not as good as density-based solvers at maintaining this coupling.

Density-based solvers do take more time and memory per iteration to solve. However, for compressible flows where coupling between different equations is limiting convergence, they tend to converge in fewer iterations than pressure-based solvers (and can be cheaper overall).

July 24, 2019, 05:04	Classification and Selection of Solvers for low subsonic flows in CFD	#1
udaych17 New Member Uday Join Date: Mar 2019 Posts: 1 Rep Power: 0	I am working on an external compressible aerodynamics flow problem which deals with speeds of Mach 0.45. I am uncertain whether to choose a density-based solver or a pressure-based solver. Generally accepted range from incompressible flows is below Mach 0.3 and Pressure-based solvers are considered best for incompressible flows but at speeds of Mach 0.45 will it be a good choice to still opt for Pressure-based Solver? My understanding is that density-based solvers take more time and memory for solving the problem and hence I want to weigh my decision based on any useful feedback or suggestions. Thanks in advance!

July 24, 2019, 10:39		#2
LuckyTran Senior Member Lucky Join Date: Apr 2011 Location: Orlando, FL USA Posts: 5,763 Rep Power: 66	Both will work (and work very well). You can use whichever one you like. There are no incompressible flows per se in Fluent.... it's solving the full compressible navier-stokes regardless of whether you use the pressure-based or density-based solver. Pressure-based solvers are good for low mach number flows because density-based ones are (traditionally) unstable there without special treatment. Modern solvers have this treatment to stabilize the solver and you can use the density-based solvers for low Mach flows now too. Pressure-based solvers also can be used for high Mach number flows (I have used it up to Mach 25). They can be less than optimal compared to density based solvers because the momentum-energy coupling is much more important in compressible flows and pressure-based solvers are (they are not bad per se but they are) not as good as density-based solvers at maintaining this coupling. Density-based solvers do take more time and memory per iteration to solve. However, for compressible flows where coupling between different equations is limiting convergence, they tend to converge in fewer iterations than pressure-based solvers (and can be cheaper overall). CMIUCBS likes this.