CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM (http://www.cfd-online.com/Forums/openfoam/)
-   -   Body in windtunnel - high preassure behind the body? (http://www.cfd-online.com/Forums/openfoam/72629-body-windtunnel-high-preassure-behind-body.html)

Andyjoe February 12, 2010 07:42

Body in windtunnel - high preassure behind the body?
 
Hello,

I'm investigating the airdrag of a objekt in the wind tunnel. I'm using the simpleFoam solver, running at a speed of 200 m/s. The inlet preassure and the initial preassure field is set to 0. I used this preassure values, because I basically use the motorBike tutorial, and 0 preassure is the value used there.
My result is far away from that, what I expected: Behind the body, the preassure becomes bigger, then in the stream before the body.
I think the reason might be, that the airspeed behind the body decreases, and lower speed means higher preassure. But I don't think, that this has anything to do with reality The calculated drag of the body is 0,022 which is far away from the measured which is about ~0.5. So I wonder whats going wrong, because the motorBike tutorial is a tutorial about calculating the forces, so the settings should be ok. Has anyone an idea, what could go wrong? Thanks in advance.

thekay February 12, 2010 09:13

Maybe it would be better if you used a compressible solver for this amount of speed? In my eyes it is way too large for simpleFoam.

Andyjoe February 12, 2010 09:22

Thanks for replying. What kind of solver would you commend? sonicFoam only works at speeds above Ma = 1. 200m/s is the absolute maximum speed that occurs, U_inf is 170m/s. Thats fast, but only about Ma ~ 0.5, so from the physical point of view it would be correct to assume that the air is incompressible. And I don't think this assumption has influence of the simulation, because the solver don't know that air is incompressible, so everything is consistent. And I think I have seen simpleFoam solutions in which base drag occurs. But maybe I'm wrong.
In the first post I have forgot to mention, that I'm using the k-omega-sst model.

Thomas Baumann February 16, 2010 05:10

Hi Andyjoe,
I'm not sure if you're right using an incompressible solver. The machnumber is too high. I think it should be smaller than 0.2 for using incompressible solvers.

Regards Thomas

Andyjoe February 16, 2010 11:57

3 Attachment(s)
I already tried to use a velocity of 20 m/s. I also checked the viscosity, because it seems that it would be to low.
I atached three images of the run with 20 m/s, argunaIIIanimU shows the velocity distribution, argunaIIIanimP the pressure distribution and argunaIIIHeck shows the velocity vectors at the base of the rocket. So there are vortices, altough the force coefficients given by libforces are unrealistic. The cd value should be about 0.015, while the value measured in reality is about 0.7. The dimensions of the simulated rocket are equal to the dimensions of the real rocket, 3.2m.


All times are GMT -4. The time now is 19:12.