![naca 0012 airfoil naca 0012 airfoil](https://0.academia-photos.com/attachment_thumbnails/37856798/mini_magick20180818-29325-15o344j.png)
(2015) show that variations in discretization schemes have a great influence on the evolution of the variables towards convergence, but that the value towards which they should tend as the mesh is finer, should be the same. I used the PCG solver with FDIC preconditioner for pressure -GAMG was very unstable- and PBiCGStab with DILU preconditioner for the asymmetric matrices.įor asymmetric matrices, we set a tight tolerance of less than 1e-10, while for pressure -as it is more expensive to calculate- we run with a relative tolerance of 0.001.ĭiskin et al. Since the solution was initiated with the coarsest mesh and then each successive solution mapped to the next grid level, it was not necessary to start with lower order schemes. As with MUSCL, both schemes are second order accurate discretizations of the advective terms. Although these schemes are also available in OpenFOAM, I was getting unstable simulations, so I used linearUpwind instead, for velocity, energy, and turbulence, and linear for the rest. (2015) are second order accurate via a MUSCL scheme. The solutions from all three codes in Diskin et al. Note: older versions of OpenFOAM (v2.x and older) need changes to the turbulence models to match the definitions found in NASA’s website. In any case, this variant should provide negligible differences in most cases. $$ \tilde$ the model remains identical to the standard version. On the other hand, in the farfield we have also the following conditions: Which results in a chord of L = 1.80645 m. Finally, we can obtain the airfoil chord from Reynolds number, air velocity, and kinematic viscosity as Fourthly, we can get additional dry air properties such as kinematic viscosity $\nu_\infty$ = 15.68 x 10 -6 m 2/s. Thirdly, now we can use the Mach number to get the flow velocity, $U_\infty$ = 52.08 m/s. As a result we get a speed of sound of $a$ = 347.2 m/s. Where $R$ is the specific gas constant (287.058 J⋅kg −1⋅K −1 for dry air). Secondly, once we know temperature, we can compute the speed of sound as Firstly, we convert the reference temperature from Imperial to SI (540 R = 300 K). That’s not the case for OpenFOAM, so the conditions need to be translated.
![naca 0012 airfoil naca 0012 airfoil](http://airfoiltools.com/images/airfoil/goe459-il_l.png)
The conditions above states are given for nondimensional CFD codes such as CF元D or FUN3D. The heat capacity ratio ($\gamma$) is 1.4. The Prandtl number Pr is taken to be constant at 0.72, while the turbulent Prandtl number Pr t is taken to be constant at 0.9. There, we get that the conditions for the case are M = 0.15, Reynolds number per chord is Re = 6 million, alpha = 10 deg, reference temperature = 540 R. (2015) and the TME website as reference for this case. I’ll be using the results from Diskin et al. The structured PLOT3D grids can be easily converted to OpenFOAM format with the plot3dToFoam tool. As a result, the grid contains approximately 1 million cell. The minimum length -after scaling- at the wall is 7 x 10 -7 and average stretching rate normal to the wall is ~1.02 for the points near the wall. In all available grids, the farfield extends about 500c. The grids have been generated such that the airfoil closes at chord = 1 with a sharp trailing edge.
![naca 0012 airfoil naca 0012 airfoil](https://media.cheggcdn.com/media/832/83221353-7446-4290-a5b6-3c235d84e1bd/phpxEPS2s.png)
Among the available options, I have opted for the Family II of structured grids, from size 225 x 64 to size 1793 x 513.
#Naca 0012 airfoil verification
NASA makes available several grids used in their verification and validation studies.
![naca 0012 airfoil naca 0012 airfoil](https://media.cheggcdn.com/media%2Ff98%2Ff9888eb3-7ca8-4539-99e0-209a2452195b%2Fphp9PAlzy.png)
#Naca 0012 airfoil download
Here we will give only a brief outline of the case setup for more details you can download the finest mesh case at the end of this post, here. 2.4 Various profiles along different lines.2.1 Convergence of residuals and variables of interest.1.4 The Spalart-Allmaras turbulence model.