[jmt]

I am running a simple internal flow case with reactingFoam. The geometry is a back-facing step with periodic boundary conditions in the spanwise direction. To test the case, I turned reactions off and chemistry off. (In fact, I have modified the solver with YEqn.H removed so as not to transport species at all. The behavior persists.) Reactions will eventually be turned on so I am trying to get the reactingFoam, no reactions case to work first.

The inlet temperature boundary is set to 1500 K with the idea for hot gas to propagate into the domain. The Mach number is on the order of 0.01 (U_infinity = 6 m/s).

I have initialized the velocity with the mean velocity from an incompressible case run with pimpleFoam (via mapFields). The inlet velocity is a parabolic profile combined with the LEMOS inflow generator. I have also tried running with this generator turned off and the behavior still occurs.

For grad(U) and div(phi, U) schemes, I have tried Gauss linear and cellLimited Gauss linear 1. The behavior still occurs.

The ICs/BCs are:

* Initial conditions:
+ U: Mapped mean velocity from the ico case
+ p: uniform 100000
+ T: uniform 300
+ nut: uniform 0
+ alphat: uniform 0

* Inlet:
+ U: fixedValue (parabolic profile)
+ p: zeroGradient
+ T: fixedValue 1500
+ nut: calculated
+ alphat: calculated

* Outlet:
+ U: zeroGradient (also have tried pressureInletOutletVelocity)
+ p: totalPressure (also have tried waveTransmissive)
+ T: zeroGradient
+ nut: calculated
+ alphat: calculated

* Front/Back Walls: Cyclic for all conditions.

* Top/Bottom walls:
+ U: No-slip (wall-resolved LES)
+ p: zeroGradient
+ T: fixedValue 300
+ nut: fixedValue 0.0 (WRLES)
+ alphat: fixedValue 0.0

The pressure waves propagate as shown in the gif below. Has anyone encountered this? It is as if a shock occurs at the inlet. The pressure waves blow up the case eventually.