[Time4Tea]

Hi, I am running a compressible simulation of internal flow through a valve and I am seeing quite a big discrepancy in the total integral of Acoustic Power, depending on whether I use the pressure-based or density-based solver.

The simulation is 3D, compressible, steady with air. I am using the realizable k-epsilon turbulence model. I am doing a half-model, using a plane of symmetry down the center. The pressure drop is quite high (325 to 50 psi), so there are supersonic regions and shocks forming. I have kept the mesh, physical models, materials, b/cs the same for both cases. Literally the only difference is the solver.



With the pressure-based solver (coupled, second-order momentum, turbulence, energy), I get the following result:


Volume integral of acoustic power = 146.0 W
Integral of surface acoustic power on walls = 140.6 W
Mass integral of turbulent kinetic energy = 99.2 W


With the density-based implicit solver, I get:


Volume integral of acoustic power = 216.7 W
Integral of surface acoustic power on walls = 125.5 W
Mass integral of turbulent kinetic energy = 99.9 W



The total acoustic power is almost 50% higher, although the total turbulent kinetic energy is almost the same. The overall mass flow and velocity field seem very similar between the two models.


Does anyone know what might be causing this discrepancy between the broadband acoustic results of the two solvers?