[Dronzer]

Hi,

I hope you all are doing good.
I am trying to simulate the flow through a wind tunnel and measure the aerodynamic force coefficients of an airfoil placed at the test section of the wind tunnel. I need to do this at various angles of attack.
I know that adjusting the flow direction (and keeping the airfoil horizontal) can do it (PFA pic1). But, this does not seem to be the right method for the simulation of the flow through the wind tunnel (PFA pic2), especially at higher angles of attack.
Does anybody know how to adjust the angle of attack in the wind tunnel simulation without remeshing the geometry for each angle of attack?
Do I need an overset mesh for this?
Please let me know. Thanks in advance.

[vinerm]

There are at least three ways to do it. One is with moving mesh, another one with dynamic mesh, and another one with overset.

You can cut the region around the airfoil in a circular shape. You have to define interface between the circular region and the rest of the fluid region of the tunnel. For each angle of attach, you can rotate the mesh of the circular region and then recreate the interface. Rotation has to be done using Moving Mesh. You can define a rotational velocity, doesn't matter what value, and then view the mesh motion for a certain period of time. The period of time must be such that the product of time-period and velocity specified gives you required rotation. Then, disable the mesh motion and run the simulation. Interfaces will reduce the accuracy a little bit. But if you keep interface far away from the airfoil, then it will have very less effect.

Second method is by remeshing but within Fluent. You need triangular mesh (tetrahedral in 3D). This would not require an interface, hence, no compromise on accuracy. Using dynamic mesh, you can rotate the airfoil to required angle.

Third option is overset, similar to the first one.

[Dronzer]

Hi,

Thanks for the detailed reply.

I preferred a structured mesh (using ICEM CFD, PFA Pic1) as I thought it would help the calculation. Therefore, I will go for either option 1 or 3.

Can I perform a steady-state simulation for this? (Since I will have to input rotation rate and time)

In any case, I will do a few trials. I hope I can update the thread if I have any other concerns about the simulation.

Thanks again.

[vinerm]

If you want to use structured mesh and variation in AoA is not very large, then I'd suggest method 2. You can use only smoothing option wherein Fluent will not remesh but compress and expand the cells slightly. Since the variation is to be handled by compression and expansion of cells, therefore, variation in AoA should not be very large. How large depends on mesh cell sizes. I suppose within a variation of , it should work fine.

[Dronzer]

Originally Posted by vinerm

If you want to use structured mesh and variation in AoA is not very large, then I'd suggest method 2. You can use only smoothing option wherein Fluent will not remesh but compress and expand the cells slightly. Since the variation is to be handled by compression and expansion of cells, therefore, variation in AoA should not be very large. How large depends on mesh cell sizes. I suppose within a variation of , it should work fine.

Thank you. I will look into that too. Appreciate it.

[talhanaimrao]

Hi. I want to perform simulation from -30 to 30 degree AoA with 5 degree increment using dynamic meshing. How should I proceed?