Laminar or Turbulent Flow
 

Hi all,

I have a query regarding a project I finished recently. It was a replicate a Wind Turbine setup in a wind tunnel and simulate the experimental conditions and wake of the turbine, then drawings correlations, if any, with the experimental results.

The simulations correlated strongly with the exp. results however I have a query.

I was questions about the validity of selecting turbulent flow instead of laminar flow as the reynolds number over the individual blades was only approx 18,000-20,000. Given these low values I was told I could have modeled the blades as small flat plates and yielded the same results????

Any one have an opinion on this?

Same results? No way. Similar results? Sure. To get the same results the flat plate drag and lift coefficeints would have to be the same as the sectional data. Even at extremely low Re, this doesn't happen really. Your visualizations would be essentially identical, but thrust, torque and power would all be a little off. I think your model is fine the way it is.

As for laminar vs turbulent. Don't debate this. Even though you may have low reynolds number and even laminar flow over the blades, the wake is most definitely not laminar. At a minimum there will be turbulent mixing of the wake with the freestream as it propagates. Doing a laminar simulation would mean that there is no turbulence equation to govern this behavior.

Nice images by the way...

Cheers MBdonCFD,

It was all because of a query that came up during my defence of the project with a lecturer. I hadnt even questioned my setup before I had entered, and no papers or comments from my supervisor had indicated anything other than a solid setup. Will wait on results and see.

My argument was was exactly as you had mentioned that the wake would be turbulent, and calculating a reynolds number (which I did not include in the final report as it did not influence my decisions on setup) to see whether the flow over the small wind turbine model was laminar or turbulent would have changed that.

Images were easily done with a custom color map, making some scaler values transparent