CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   Main CFD Forum (https://www.cfd-online.com/Forums/main/)
-   -   Mach 4 wind tunnel 2D (https://www.cfd-online.com/Forums/main/156484-mach-4-wind-tunnel-2d.html)

AET July 8, 2015 08:10

Mach 4 wind tunnel 2D
 
4 Attachment(s)
Hi.

I am simulating a steady 2D Mach 4 wind tunnel in order to capture the shock waves created in the measuring region. The wind tunnel has ambient conditions at the inlet and vacuum at the outlet (0.01 Pa). The inlet is a convergent-divergent nozzle and after the measuring region there is a plate whose angle I will change in order to study the behaviour of the shock waves.

The problem is that I am not getting any shock wave or residuals convergence.

I am using Star-CCM+. The mesh is structured (although I've tried changing to a polyhedral mesh, and also refining this). The inlet boundary condition is set as stagnation inlet (I've also tried with a mass flow inlet), with total and static ambient pressures, and total temperature of 300K. The outlet boundary is a pressure outlet, with 0.01 Pa of static pressure and a static temperature of 300K. I am using k-omega turbulence model with default settings.
I have tried several different initial conditions, back pressures and inlet pressures, but I still don't get any proper result.
Some other noteworthy details are:
-Reference pressure is 0 Pa.
-Using Coupled flow, ideal gas, steady, two-dimensional models.
-Courant number 0.5 (tried several ones)
-Tried also grid sequencing.
-I get messages of reversed flow in the inlet and also sometimes in the outlet, and I get minimum absolute pressure corrections in X cells.

Any recommendation? Best practices? I have tried so many different things.
I attach some figures of the mesh, residuals and pressure and mach number scenes.

Thank you in advance.

shuige July 9, 2015 09:28

you need consider y+ ,and the Courant number is 0.5 and i think your case is very easy divergicy, this may caused by the wrong sets of your BC

AET July 9, 2015 13:26

Quote:

Originally Posted by shuige (Post 554648)
you need consider y+ ,and the Courant number is 0.5 and i think your case is very easy divergicy, this may caused by the wrong sets of your BC

y+ are within the acceptable values. I've tried low Courant numbers such as 0.1 and also high such as 2.0 or 5.0. What can be wrong with the BC?

Thank you.

shuige July 10, 2015 23:20

if the Courant numbers is 5 or 3, your case become divergency, it maybe caused by your out BC setting. if you want to capture the shockwave you need choose density-based, and use small size mesh and time step


All times are GMT -4. The time now is 11:02.