[biro]

Hey there,

I use the SST-turbulence modell an try to refine my mesh to y+ ~1.
Now I have a y+ value of 5-10 and when I just refine the mesh, the solver has problems to start with the simulation. The run will stop within the first 30 iteration.

The Problem are warnings (only within the first iterations) of high machnumbers (> 10) and the p-mass residual has strange values.
Also I have problems with this warning at the start of the run:
+--------------------------------------------------------------------+
| ****** Notice ****** |
| The non-dimensional near wall temperature (T+) has been clipped |
| for calculation of Wall Heat Transfer Coefficient. |
| |
| Boundary Condition : Wall |
| T+ clip value = 1.0000E-10 |
| |
| If this situation persists and you are using the High Speed Model, |
| consider enabling Mach number based blending between low speed and |
| high speed wall functions. You can do so by specifying a Mach |
| number threshold as follows: |
| |
| EXPERT PARAMETERS: |
| highspeed wf mach threshold = 0.1 # default=0.0 (off) |
| END |
+--------------------------------------------------------------------+



In previous simulations I have similar warnings. But after a few iterations everything went well.

I choose "Incl. Viscous Work Term" and "High Speed (compressible) Wall Heat Transfer Model" (specification from my tutor).
Maybe this are the problem ?!
At inlet I have subsonic flow, but maybe the flow reaches Ma > 1 at my blades, which redirect the flow (that is one thin what I'm looking for).

[ghorrocks]

Only include viscous work if it is significant. It is unlikely to be significant in high mach number flows. I am not familiar with the wall heat transfer model you refer to so I would definitely compare it to the default one.

If you must use these models then consider doing a simpler model to start convergence and then use that as an initial condition to the final run with the full settings. Then you have a robust model to start things off, but get the full solution to run to convergence.

[biro]

I dissabled both settings and start a run, but it didn't work. The Mach Number increase and the simulation stops with an error, like before.

As Inital Conditions I use the run with a little bit higher wall distance. What can I do in addition to build a simpler model, which I use as initial condition?

When I analyze this run (with higher wall distance) the max machnumber is approximately 0.8. So it's a subsonic flow.
So I think it's a near wall problem, but I don't understand.

[ghorrocks]

To do a simple model for an initial condition:
* Coarser mesh
* Upwind differencing
* Default physics models (so the default wall HT condition)
* If transient use first order time differencing

[biro]

Where do I found the setting for upwind differencing ?
Solver Control --> Advection Scheme ?

[ghorrocks]

Yes, that is correct.

[biro]

I try getting start the run with all different setting (another HT Model, disabled visc Work terms, Upwind advection scheme, first order turbulence numerics), but every simulation shows the same behaviour and the same error:
The Warning "maximum Machnumber", which increases more and more (> 4e3). And at last: "Floating point exception: Overflow"

I searched for another fault in my setting, but I can't find anyone. There are always the same settings with a step by step refined wall distance in my mesh.
But now nothing works

[ghorrocks]

This problem is related to this FAQ: http://www.cfd-online.com/Wiki/Ansys...do_about_it.3F

The responses in the FAQ apply to your case as well.

[biro]

I try many things with mesh, cfx settings and geometry. And so I think the problem is related to my geometry and better initial conditions are needed.