simulation around an hydrofoil using simpleFoam - SA model

 Register Blogs Members List Search Today's Posts Mark Forums Read

 May 28, 2013, 13:35 simulation around an hydrofoil using simpleFoam - SA model #1 New Member   Olivier Mesnard Join Date: Sep 2012 Location: Boston Posts: 1 Rep Power: 0 Hello everyone, I am trying to simulate a steady-state turbulent flow around a chord-normalized hydrofoil. The Reynolds number is Re = 10^6 and the angle of attack is set to 2deg. I am using simpleFoam with the Spalart-Allmaras model. For the initial & boundary conditions, we need to define the eddy viscosity (nut) and the viscosity-like variable (nuTilda). My assumption is that the flow is laminar in the farfield, so I set the turbulent viscosity to zero at the inlet, outlet and on the top/bottom (symmetryPlane). I also make the assumption that in the viscous sub-layer, the flow can be considered as laminar, and so, the eddy viscosity is set to zero. To avoid refining the mesh so much, I also use a wall function "nutUSpaldingWallFunction". My first question is: from which value of y+ do I have to use a wall function? My convergence criteria are set to 10^-5 (see fvSolution file attached). However, the convergence is never reached. When I am plotting the drag and lift coefficients over the iterations, I have a lot of oscillations (see attached file). Even when I am refining the mesh, these oscillations are still here. Moreover, the convergence history looks very poor; my pressure residual is around 0.1 (see attached file). Plotting the vorticity field, it looks like it is not well resolved (see attached file). My second question: what is the best solver solution for the pressure for high Reynolds? If you have any advice to give me, that will be very helpful! Thank you very much. forceCoeffs_2deg_470K.png vort_30000_470K.jpg fvSolution.txt nut.txt residuals.png

 May 29, 2013, 06:47 #2 Senior Member     Vieri Abolaffio Join Date: Jul 2010 Location: Always on the move. Posts: 308 Rep Power: 16 few observations. 1) SA is not really good for that flow regime, you will need more advanced turbulence models to have realistic results, 2) i'ts been a while since i used it but i don't think SA cope well with wall functions. you shuld fully define the viscous sublayer and have a y+ less than 1 3) check your mesh, i bet you have something horrible going on with the BL at approx 60-80% of the chord, on both sides of the foil 4) increase the residuals tolerance. for the pressure try GAMG __________________ http://www.leadingedge.it/ Naval architecture and CFD consultancy

 Tags openfoam, simplefoam, spalart allamras, turbulence, wall function