Hope the focus be made on the lattice Boltzmann method if possible. for this method the use of large Eddy simulation (LES) is justifed by the direct calculation of the strain rate tensor Sij=(ui,j+uj,i)/2 at each node without finite differencing !
Even when we know that there is no 3D eddies and that the flow the flow is laminar in its core but turbulent between fringes ? Means that when we simulate the flow using RANS we find a laminar one with smoothed and contineous isolines, can we here use the SGS models ? see enclosed file. Is this a reason to use LES model with lattice boltzmann method ?
Thank you for your clarifications Daniel,
First, in the file I use LES in Figure 1 and RANS in the second.
I do not defend the use of "LES" but I try to find a connection saying that even if I use the "LES" I get good results in comparison with the experimental and simulation, at least for this case at high temperatures and velocities jets.
You can , but you will get good results even with simpler models like the more sophisticated 2-equation models.Please see Argyropoulos and Markatos ,@Recent advances in Turbulence modelling' Applied Mathematical Modelling 2015
I don't find in the indicated paper *, something speaking clearly about LES for axisymmetric flow, means we can simulate only a half plane !
*/Argyropoulos and Markatos, in Applied Mathematical Modelling 2015.
There is nothing different practically. Theoretically there is since turbulence is always 3D
- Badges
- Science topic
- Similar topics
- Control Systems Engineering
- Filtering