CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   Similarity Theory in External Aerodynamics (http://www.cfd-online.com/Forums/cfx/23505-similarity-theory-external-aerodynamics.html)

 Amod Kumar January 10, 2007 03:17

Similarity Theory in External Aerodynamics

I am simulating external aerodynamic simulation of an automobile body. I am using symmetrical half. As per standard recommended practice, the computational domain should be 8L X 4L X 2L in longitudinal, transverse and vertical direction.

However, I have a machine limitation with mesh size restricted to < 1 million elements. It seems impossible to squeeze the number within this limit.

I resorted to use Similarity analysis to reduce the domain size. I have scaled down the computational domain by a factor of 16 and increase the inlet velocity 16 times to ensure dynamic similarity (constant Re for model and prototype).

Can anyone suggest the imlication of such simulation strategy on final result (vis-a-vis drag forces)?

Thanks,

Amod

 Glenn Horrocks January 10, 2007 18:28

Re: Similarity Theory in External Aerodynamics

Hi,

I would not recommend similarity scaling the simulation. Model the true size and use the mesh size required for a mesh independent solution. If Reynolds number is the only important scaling factor then by reducing the physical size of the object while keeping the Reynolds number the same you will have achieved nothing - the mesh size required will scale with the geometry!

If you do not have the computer resources to do the simulation then get some parallel licenses.

Glenn Horrocks

 Charles January 11, 2007 02:54

Re: Similarity Theory in External Aerodynamics

Have to agree with Glenn here. One of the greatest advantages of CFD is that you can get away from the scaling issues often involved in experimental work. If you now scale your CFD model like an experiment, the exercise becomes rather pointless, unless the objective is to specifically work on the scaling parameters! And if you really can't run bigger meshes, go through a mesh refinement study with the meshes that you can manage, to check how sensitive your solution is. You may be surprised at how effective a relatively coarse mesh may be. In this situation your best option is to spend a lot of time on building a really efficient mesh, probably using mostly hexa cells.

 Amod Kumar January 11, 2007 07:08

Re: Similarity Theory in External Aerodynamics

Dear Charles & Glenn,

Thanks for your time and feedback.

I had similar reservations regarding model scaling. However, I got inspiration from an article on CFD simulation of "Large Outdoor Cabinets" used in telecommunication industry namely BTS (Base Transmission Station) sheltors.

You can e-mail me a blank mail at amod_kumar@rediffmail.com. I will send the article for your further comments.

Thanks and regards,

Amod Kumar

 All times are GMT -4. The time now is 09:39.