CFD Online Discussion Forums

CFD Online Discussion Forums (
-   FLUENT (
-   -   Laminar vs. Turbulent Models (

Dahvid Brown October 18, 2000 12:08

Laminar vs. Turbulent Models
I am trying to model a 2-D rotating cylinder in a cross flow. The Reynolds number for my setup is 52,000 which should be in the laminar flow range. I have run both laminar and K-epsilon models in Fluent with mixed results. I get better results for the coefficient of lift with the laminar model, but my results for drag are closer with the turbulent model. If I could pick and choose from the 2 different models, then life would be great, but that approach isn't scientifically sound. My cylinder is 1 inch in radius and the free stream is 15 m/s. I am sure that my results are mesh independent. I appreciate if someone has some suggestions. Thanks.

Rüdiger Schwarze October 19, 2000 03:49

Re: Laminar vs. Turbulent Models
The correct prediction of separation past the cylinder is essential in order to get correct results for the drag and You shouldn't get correct results for separation in the laminar flow field. On the other hand the lift needs a correct description of the laminar boundary layers, thus the laminar model will give better results for the lift coefficient.

Dahvid Brown October 20, 2000 15:23

Also Problem Converging for Laminar
One more question on these models. Why is it that I have no trouble converging in the the turbulent model, but half of the time, I cannot converge with the laminar model. Thanks for your help.

Rüdiger Schwarze October 23, 2000 02:36

Re: Also Problem Converging for Laminar
The laminar model is based on a wrong physics, therefore You shouldn't be surprised that Your iteration won't converge. More precisely You need the "turbulent" viscosity in order to describe the enhanced momentum transfer in the flow field.

All times are GMT -4. The time now is 04:14.