computational grid generation on fluent
 

Hi! guys,

I am a new user of fluent. I am working on a simple aerofoil shape trying to analyse the flow over it when it is near the ground. ie: ground effects

any effects if i change the grid definition? i am currently using tri elements and type :pave would there be much different if i use other type of elements? I am running it on spalart-allmaras model.

what are possible ways of improvements?

could some guys who are good in such stuff.. provide so feedback

thanks alot really appreciate it

terry

Re: computational grid generation on fluent
 

For airfoil flows you should definitely use quads/hexas (if your geometry allows that). This will resolve the boundary-layers and wakes better since you can assure that the cells are aligned with the flow. The 1-equation SA model is okay for attached flows in this type of applications.

Re: computational grid generation on fluent
 

Hi! Jonas

Thanks for the response

appreciate it, but if we are considering just the forces for life and drag over the foil, what are some of the concerns that we should look out for? its a one dimensional problem and we are using high reynolds number, is it still suitable to use a turbulence model like the SA? its steady state flow... at 50m/s across the aerofoil.

thanks again Terry

Re: computational grid generation on fluent
 

Computing the drag correctly can be a bit tricky since you have to get the skin-friction correct. The lift is easier. I haven't done any validations of the SA model for airfoils so I can't say how well it will perform on the drag. You can always compare it to a more advanced two-equation model (I'd recommend the Realizable k-epsilon model with the Wolfstein 1-eq model in the boundary layers - Fluent calls it "two-layer zonal model"). Lift should be okay as long as you have attached flow. If this is an airfoil in free air make sure you use a very large domain, this is important. Search the forums here and you'll find a lot of discussion about this, including references, if my memory doesn't fail me now.

You should of course first compute a wing for which you have data to compare your results with - see the Resources/Reference section for a lot of nice cases to run. I think that both the ERCOFTAC database and the NPARC database includes airfoil cases.