CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   ANSYS (http://www.cfd-online.com/Forums/ansys/)
-   -   Turbulence. (http://www.cfd-online.com/Forums/ansys/87161-turbulence.html)

ATOTA April 12, 2011 22:17

Turbulence.
 
Hi,

I just want to know how to create turbulence in a simulation before it hits the actual object.

I am testing a section of a bridge deck in a wind tunnel and trying to reproduce it in ANSYS Fluent.

In the wind tunnel, they had small objects on the floor to produce turbulent wind. How do I recreate that in ANSYS. The use of LES model makes it turbulent after hitting the deck but I don't know how to put the flow turbulent even before reaching the deck.

Any suggestions ??

Thanks !

freemankofi April 17, 2011 14:44

I am sure if your question is clear enough. But, I think you might be missing "virtual/simulation" with reality, despite I have no experience with LES modeling.

In fluid dynamics, there is no such thing as "creating turbulence". "creating turbulence" is a layman term, as a flight attendant always mentioned in the plane. I suggest you get yourself some tutorials/educations on CFD before proceeding on this project.

In LES, you're basically interested with large eddies (is that what you called turbulence?) and somehow indirectly accounts for the small eddies. Depending on the "model" use, you might missed out the actual flow physics which you have indicated.

In a wind tunnel, it's an experiment and you typically have NO foreknowledge about the flow. Usually, if you don't "trip" the flow (putting small object or sometimes sand paper), your incoming flow can be laminar or low turbulent intensity.

In CFD, you don't need to trip the flow. You do that by specifying the boundary conditions (BCs). For eddy-viscosity based models, BCs required that you specify the inlet (incoming Wind/air) turbulent properties such as turbulent intensity, eddy viscosity ratio or eddy length scale, etc. In cases, where there's NO experimental data available for this inlet conditions, you use some of the available analytical/empirical formula in literature to estimate them. Here, you need to BC sensitivity tests.

However, you can create a rectangular bump/plate at the flow domain similar to your experiment. Not only this is a waste of resources since it is NOT necessary in CFD, it creates different flow physics which can lead to wrong results.

Refer to the ANSYS CFX/Fluent manual on the "Theory of LES models" and check your boundary conditions. CFX has a well-documented theories on using LES to model turbulent flows. You might want to check your solution initialization also.

Regards,


All times are GMT -4. The time now is 18:22.