CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   Flowlab - Cylinder turbulent model results (http://www.cfd-online.com/Forums/fluent/44946-flowlab-cylinder-turbulent-model-results.html)

Brendon June 7, 2007 00:09

Flowlab - Cylinder turbulent model results
 
Hi. I am currently working through the Flowlab flow past a cylinder exercise. Just making sure that i am correct in thinking that the turbulent model should be used for 2000 < Re. I'm having difficulty in getting reasonable results for higher Re. The literature specifies that there may be problems above 20,000 due to standard wall functions being imposed.

Also, will increasing the timestep size have a significant impact on the results.

Thanks in advance

Prasad Dudhgaonkar June 7, 2007 16:54

Re: Flowlab - Cylinder turbulent model results
 
What is Flowlab flow? Flow past a cylinder needs careful treatment for adverse pressure gradient, flow separation, vortices and a streamline curvature. Realizable k-e models is reportedly best suited for your case.

You can use PISO P-vel coupling with neighbor correction for large time steps. In any case, it's always desired to keep the time step to the lowest. Large time step may lead to unstable solution or may give erroneous prediction.

Brendon June 7, 2007 21:36

Re: Flowlab - Cylinder turbulent model results
 
Sorry, FlowLab is basically a cutdown student version of Fluent that uses parameterised template files that only allow certain inputs.

There are only two models available, 'laminar' and 'turbulent'. The 'turbulent' model uses a k-e model with enhanced wall functions at Re <20000 and standard wall functions for greater Re. I was wondering if anyone had any success in high Re modeling, or whether this template would only really be suitable for low Re flow.


All times are GMT -4. The time now is 03:35.