[Doflo]

Hello CFDler,

I need your help. For several weeks I try to simulate the following case in OF: It´s a probe in a box. In the first picture you can see the internal mesh (the upper and right plane of the box are defined as uotlet, the other four one are inlet).

Here some facts:
- SimpleFoam (basis motorbike tutorial)
- SST turbulence model
- snappyHexMesh for the probe surface
- boundaries:
Inlet:
U = fixedvalue (39 6 0)
p = zerogradient
Outlet:
U = zerogradient
p = fixedvalue 0
Probe:
U = fixedvalue (0 0 0)
p= zerogradeint

The fvsolution, fvscheme und control dicts are set by default.

Well, my problem is, that the results physically doesn´t make any sense.
So I plotted the residuals (2nd pictures) where you can see, that the case doens´t converge.

I think, one reason could be the boundaries at the corner between probe and outlet, as there is a huge velocitiy peak as you can see in the 3rd picture. Maybe the boundaries on the probe and the outlet don´t get along? Or is there something with the Tolerances?

Have you got some tips, how I can improve my case?

Thanks a lot in advance.
Best regards, Dominik

[Doflo]

Here are three more pictures, which may help to understand the problem.

the inflow is 40 m/s and behind the probe the is a peak with 113 m/s, what can´t be. Moreover, the geometrical symmetrie doesn´t reflect in the results.

My suspicion is, that the boundaries on the probe surface and the outlet boundaries don´t get along with?

Best regards

[alexeym]

Hi,

Please post the following:

1. checkMesh output (as an attachment or in CODE block)

2. explain what do you mean by "default fvSchemes, fvSolution, and controlDict"

3. IC and BC for turbulent properties