Dear users,

I am currently working on my thesis about estimating power curve of a 3 bladed HAWT at the university using FLUENT. But, there are problems with the problem setup and the results .

Since I need multiple points of power obtained from 1 blade (I will multiply by 3 because of the periodic modeling), I use multiple reference frames model and spalart allmaras turbulence model. I define constant angular velocity to the fluid domain, absolute angular velocity of 0 rad/s for the wall domain surrounding the fluid domain and the turbine blade. (reference:fan tutorial) And I change inlet velocity of air and expect to find the power from moment*angular velocity. There is a certain characteristic form of the power curve where the curve has a peak and falls. But using this model, the power values I find are all unrealistic values, exceeding the power available in the wind which is already known.

My question is should I subtract the moment for the case of same angular velocity and zero inlet velocity condition from all of my cases. Because if I do this I get better results. Or is it totally wrong?

In addition, I use same angular velocity for all cases and change the inlet velocities. This is why I tried to subtract the zero inlet velocity but same angular velocity case. Because there isn't experimental data of the turbine, I cannot foresee the angular velocities and inlet velocities for each case. Bottomline, for all cases, w is the same value while Vin changes...

Thanks in advance...

Hello ,

I am workin also on this project using Fluent , u ca contact me to my email whenever u wan t , we can help each other .MR frame is the bbest method to solve this problem , concerning what u did (0 rad/s ..) i think that it true (this also what i will do). Concerning ur question , i think that bets metho dto do it since AS i know there is no way to get it . My question for u is : How did u do the meshing of the wind turbine (I well know that u should add the domain flow which is modeled by a Brick then subtracting volumes , then meshing ...); But did u do it directly like this , or u take some points on ur graph , then constracting area.. Best regards, Huss/Lebanon