Simulation over a NACA0012
 

Hi,I'm a final year student and I'm working at a project in which I have to simulate the flow over a NACA0012 at different angles of attack and Mach number without and with vortex generators. The fact is that I have just started practicing with Ansys,so,as you can imagine I am having few problems.I have a vague idea of what I have to do,because I have done a simulation for a body in a fluid domain already.But for the airfoil I'm struggling a bit.Any directive from you guys with a lot more experience?Just a starting int would help me a lot.For example for Mach number and alpha,do I have to consider typical values for a flight conditions,and from them get air temp,density.... If somebody has done something similar could plese help me? Cheers for now.

Franny

Re: Simulation over a NACA0012
 

Okay. I'm not expert in this field, but I do use vortex generators to improve flow non-uniformity in a waterjet tunnel. You normally used vortex generators on air wing to delay the flow separation. I suggest you to get a sense of what vortex generators will do on an airfoil. You may need to optimise the number, shape, and location of the vortex generators. Too many vortex generators will increase the losses, and that is not efficient, considering today high fuel price. You can also compare their performance at different flight conditions.

Re: Simulation over a NACA0012
 

I would model them all in air at constant temperature, makes things a lot easier, stick with the standard air at *degress. Start with slower speeds so that you can increase the mesh sizes near the surfaces.

Re: Simulation over a NACA0012
 

Hi,thanks for your ints!

TB,before the vortex generators I still have a bit to go,and I will consider your suggestion.

Jon,At the beginning of my simulation (just for the naca profile) I was thinking to set the problem ,at sea level conditions,thus at T=15+273 ,and at different angle of attack and Mach's nember.From this , calculate the Re ,the thickness of my boundary layer and the pressure gradient to see where separation happen.What it really worry me is how to relate my nalitical results withe the mesh size.Do I need just to play with the mesh spacing,the edge lenght,the radius of infl,the expansion factor,numb of inflated layer? Cheers.

Fra

Re: Simulation over a NACA0012
 

I assume you are running rans models? The most critical point is your Y+ value, you can see this in post when you have completed a run. these values are a measure measure of how close the first node is away from the wall. Certain models have different Y+ ranges they can cope with. If you are using the SST model then your Y+ must be approximately 2. If you are using the K-E i believe the Y+ must be greater than 30. CAn't remember exactly but they are in the manual.

Simply if you y+ is too big reduce the distance to your first node. and watch your aspect ration of these cells to ensure it doesn;t become rediculous. Jon