I tried few times, but I am not satisfied with the results. Can you present me some tutorials?
Can you simulate in CFD flow around propeller with high precision?
I still do not find streamlines satisfying results. ......
Any models?
Can you define, what "high precision" means to you? Your posted image shows already a flow prediction which would be in my opinion of quite satisfactory precision, because it resolves the tip vortex on the propeller blades over 4-5 revolutions (without letting it dissipate due to a too coarse mesh). So what you are striving to achieve? What result you are looking for?
Regards,
Dr. Th. Frank.
Dear Ionut Cristian Scurtu ,
There are several reasons to have good and satisfying results of numerical methods. To have a good insight into your simulation, please use low velocity with laminar flow, then you can continue with turbulence flow with the k-e model.
Nazaruddin Sinaga : from my perspective not a very suitable advice for a high-quality flow field prediction.
By definition the flow around a propeller is a flow field with strong curvature of streamlines. A k-epsilon model is usually not very suitable for those predictions. k-epsilon model produces a much too high level of eddy viscosity leading to very strong turbulent dissipation in the flow field. So e.g. it is not very possible to resolve the tip-vortex flow, which one can observe in the included image. The minimum turbulence modeling approach would consist in:
a) high local mesh resolution, where it is to be expected that the blade-tip vortex is propagating.
b) I would further recommend, to include the space where this blade-tip vortex is propagating, into the rotor domain which moves together with the propeller and to place the rotor-stator interface well located outward of this location in space, i.e. not to disturb the propagation of the blade-tip vortex with crossing over the rotor-stator interface.
c) Use lower dissipation SST turbulence model (SST= Shear Stress Transport) with curvature correction model/term included.
d) If mesh resolution and computational resources are sufficient, try to apply either SAS-SST or SBES (Stress-blended eddy simulation). Both are higher accuracy hybrid scale-resolving simulation methods (SRS methods). More expensive in terms of computational ressources, but if the mesh resolution is adequate than these methods are more accurate than 2-equation turbulence models as well.
For further advice I'm looking forward to a more concrete description from Ionut Cristian Scurtu on what he is specifically looking for...
Best regards,
Dr. Th. Frank.
- Badges
- Science topic
- Similar topics
- Computer Communications (Networks)
- Simulators