[Richard23]

Hello, I am currently working on my senior design project for mechanical engineering which involves designing and analyzing a darrieus-type vertical axis wind turbine, and right now I am attempting a sliding mesh to simulate the revolution of the turbine in order to monitor and extract the moments (torque) of the airfoils for calculating the coefficient of power. I'm relatively inexperienced at using ANSYS so I am having some problems with my simulations and I think it is likely related to the setup.

The dimensions involve a 1.2 diameter turbine, chord length of 0.3 m, blade thickness of 0.1 m, and the surface area of the blades is 0.013 m^2. Currently I am using the k-epsilon realizable model, an inlet velocity of 5 m/s, mesh motion to rotate my inner domain (rotation speed is adjusted for various TSRs), and a moment monitor for all three airfoils with the moment center at (0, 0). I've set the number of time steps and time step size to allow three revolutions. However, when the simulation is done calculating I export the moment values into excel and take the average, the moment I end up with is way too large. When I multiply that moment by the angular velocity I set I end up with an efficiency well over 100% for higher TSRs.

I'm not sure what I am doing wrong because based on all the research I've done this is how the analysis is conducted. I've read that by adjusting the reference values it has resolved this issue that people have had, but unfortunately I do not know how to properly set the reference values. For instance, I am not sure if area corresponds to the area of the turbine or area of the blades, and if length refers to rotor diameter or width of the blade.

Any recommendations that you have to help resolve my issues would be very much appreciated as I am not sure what to do moving forward, and I apologize in advance if I am not being descriptive enough.


*Edit*

If anyone is curious, I managed to figure out what reference values I needed to use for a 2D VAWT simulation.

Area = Rotor Diameter*Depth
Depth = 1 m (for 2D)
Length = Rotor Radius

Adjusting these values still did not fix my issue, however, I did manage to fix the high moment values. I created a steady simulation and used the solution of it to initialize the sliding mesh simulation. After I did this, the coefficient of moment values were more realistic for each TSR (up until the higher TSRs which returned a negative average coefficient of moment which from what I've read is supposed to happen).