CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   Help with shock waves (http://www.cfd-online.com/Forums/cfx/18386-help-shock-waves.html)

Adam August 31, 2000 12:51

Help with shock waves
 
Hi there

We are using CFX-5.4 to simulate shock wave interactions at high Mach numbers (upto M = 3.2).

Experimentally, the pressure in the test section of the wind tunnel is ~200 kPa. When setting up the Fluid Domain in CFX, a Reference Pressure must be specified. Also, for supersonic inlets a Relative Pressure must be entered when creating the Boundary Condition. The problem is this :

The CFX-Solver crashes whenever the Reference Pressure is NOT set to atmospheric pressure (101.3 kPa). Thus, we cannot simulate the actual pressure in the test section. We've tried to set the Reference Pressure to 101.3 kPa and the Relative Pressure to 88.7 kPa, thus resulting in a pressure of 200 kPa, but this results in a "Fatal overflow in linear solver." Setting the Reference Pressure to 200 kPa and the Relative Pressure to 0 kPa also results in an overflow.

Can anyone help ?

John C. Chien August 31, 2000 17:57

Re: Help with shock waves
 
(1). You need to find out whether the code is properly formulated for the supersonic flow at Mach 3.2. (2). Actually, the compressible flow is normally formulated using density-based , transient method. And is easier to compute than the subsonic flows. (3). I don't know what method is used in the code. It is likely that it is not formulated for the Mach 3.2 conditions. It is just my personal feeling. I could be wrong. (4). Were you able to find the Mach 3.2 applications in the vendor's website?

Dan Williams September 16, 2000 18:16

Re: Help with shock waves
 
It's always good to set your reference pressure, for compressible flows, to a ballpark value of what you expect. I'd go as far as to set the reference pressure to 200kPa and set your inlet pressure to 0 relative (if that is what you want). This is better for linear solver performance.

Linear solver overflows can occur for many reasons. Did you try simplifying the physics / boundary conditions to make sure you don't have a bad mesh?

Dan.

Adam September 19, 2000 11:24

Re: Help with shock waves
 
Hi Dan

Thanks for the help.

Regards, Adam.


All times are GMT -4. The time now is 18:06.