Lattice Boltzmann & Grid Refinement: A Study of the Vocal Fold
Ғылым және технология
The video shows the results of a study on voice generation and the vocal fold. The results were obtained by a simulation based on the lattice Boltzmann method (DNS, Re 1000, D3Q19, TRT). The simulation, which was performed on SuperMUC* using the HPC software framework waLBerla [dx.doi.org/10.1145/2503210.250...], makes use of grid refinement in order to drastically reduce the overall workload and memory requirements.
*supercomputer located at LRZ in Munich (one of the ten fastest supercomputers worldwide - as of February 2014)
Пікірлер: 7
hello， it is very nice work.Can you tell me how to generate the "gridDensityFunction" in grid refinement?
@immd10
4 жыл бұрын
If you mean how to obtain the fluid density information, it is the zero-th order moment of the particle distribution functions (PDFs) and thus the same as in the case without refinement. For other information, please have a look at the respective paper: doi.org/10.1137/15M1035240
Very nice video! I am currently working on Palabos (Lattice Boltzmann) for the simulation of a turbulent jet. I am investigating which boundary conditions should I use. Can you help me with this?
@immd10
4 жыл бұрын
We are using simple bounce back boundary conditions at the slit and at the domain borders in this simulation.
@nachogarcia5641
4 жыл бұрын
@@immd10 thank you very much!! For using that BC, the domain must be much larger than the Diameter (or extent) of the inlet, right?
@immd10
4 жыл бұрын
@@nachogarcia5641 In our simulation, the domain is indeed clearly larger than the extent of the inlet. However, the simple bounce back boundary condition does generally not depend on the size of the domain. Excellent information on the basics of the LBM (including boundary conditions) is available in literature, for example in the LBM book by Timm Krüger et al.
@nachogarcia5641
4 жыл бұрын
@@immd10 Thank you very much!!! You have been very helpful!