HEY,

I have to study a 2D wing on fluent for a concrete case of formula one ... so the maximum velocities are between 75m/s and 95m/s. I am stuck on how to define the model of solver I need to use. And I would like to know if there are particular things that we need to look in order to obtain accurate results. And on last point is the y+ velocity component that a teacher told me about but I don't know what it is, and how and where to set up that factor.

I would be very thankful if somebody could help me!!!

Thanks very much!!!!

Velocities that low, since they are in the incompressible region, you should use the Segregated solver. This is because it is a pressure based solver, and the coupled solvers are density based. Using a density based solver is a huge help in refining shocks, but since you're in the incompressible region, density changes are so small that they'll be in the numerical error range and will cause problems on convergence. This is true even if you use constant air properties (incompressible fluid). I'm not sure how the coupled solver works when the density is constant, but I know it doesn't work well!

When your teacher mentioned the y+, I think they were talking about the cell y+ values. This is used for estimating the laminar boundary layer in turbulent flows along walls. If you're using a standard k-epsilon or k-omega turbulence model you should make sure your y+ values stay between 30 and 300 (but I've been told it's actually best to keep it between 30 and 130... different people say slightly different things on this point). If you're using wall treatments, then I believe you want to keep your y+ values less than or equal to 1. Also, if you're using the Spalart Almaras turbulence model, then you want to make sure your y+ value is either close to 1, or above 30. The model doesn't properly predict the turbulence in between y+ of 1 and 30 for this model. Your y+ value is controlled by the cell center height off the surface as well as the local reynolds number. An estimate of your y+ value to help in your initial mesh can be calculated at:

http://geolab.larc.nasa.gov/APPS/YPlus/

As far as turbulence models for a 2D wing, this is what the Spalart-Almaras model was designed for. Also, the k-epsilon realizable model seems to be pretty reliable in these situations.

I highly recommend the Fluent tutorials. Since you're probably on a University license you might have to find your system admin to get your hands on the tutorials. Most of them come with the software, and others are available online at http://www.fluentusers.com/ (the sys admin will have to get a username and password to get access to this site). There's a 2D airfoil at high-speeds... this should help, but remember that they're in the compressible region, so they're using the coupled solver. Also, they set the material density as ideal gas and the viscosity as a sutherland approximation... you'll have to decide for yourself if it's better to do this, or to keep them constant at standard atmospheric conditions. Other than that it should be close to the same problem.

Hope this helps, and goodluck Jason