[turbvisc]

Hello,


I'm trying to run a simulation on the Plume-Surface-Interaction in a lunar environment. However the simulation keeps diverging. The total pressure and Mach number are increasing very much in one cell near the outlet (either on the outlet near the ground or near the outer nozzle wall, see picture 2). I have tried various approaches with fully structured and unstructured meshes, both fine and coarse, or coarsening towards the outlets, with resolved or unresolved boundary layers.

I achieved on result on a relativly coarse mesh (see picture 1), but refining, again, leads to divergence.
Also I varied the size of the farfield, without achieving something.



Boundaries:
Pressure Inlet: 4.5 MPa, ~2700 K

Outlet: 1 Pa, Non Reflective Wave Model with Pressure at infinity
No slip walls



Solver settings:
steady state, axissymmetric

species transport
k-w-SST, with Produciton Limiter and Compressibility Effects

Solver: density based, implicit

Flux: AUSM
Discretization: Least Squares Cell Based, second order (also tried with first order)
Higher Order Term Relaxation enabled
CFL and URF: 0.3


hybrid and fmg initialization



I am aware, that the Navier-Stokes equations are only valid for a continuum, this simulation shall be used, to identify the continuum breakdown, to then couple with a DSMC sim. Even though potentially unphysical, literature has shown, that fluent could be used to run such a case (e.g. https://www.researchgate.net/publica..._lunar_landing or https://www.semanticscholar.org/pape...0bc82cc29da622)


Ma_number.png

diverged.png