I recently finished an MSc project on cavitation over a 3d hydrofoil. While running a DES simulation for cavitation on NACA 0015 in Star-ccm+, I observed vapours near the outlet developing after the bubble diminished from the trailing edge. The pic attached shows the volume fraction of vapour at 100ms. The vapours downstream start developing at 90ms, would be great if you could shed some light as I am still an amateur, could this be flashing or cavitation inception in the vortices developed?. The conditions are at σ = 1, inlet velocity 6m/s and outlet pressure 20.9 kPa.
Thank you.
It looks very much like cavitation forming inside the vortex cores. Did you evaluate the pressure field to check whether it is below the local vapour pressure? What puzzles me is that the vapour does not seem to be located inside of the wake and that no cavitation is present on the airfoil.
Is this a transient phenomenon or does the vapour persist in steady state? In the first case it may be connected to the starting vortex of the airfoil.
thank you for the reply Michael. Unfortunately the pressure field scenes were not saved by star-ccm+ as the software was not auto refreshing while saving. I have attached the initial cavitation seen on the airfoil, its mostly attached/sheet type shedding from the trailing edge. If you zoom in you will see a sheet of vapour. The downstream vapours are transient as you can see in the attachments, by 0.29 s they have increased and moved closer to the airfoil.
the first pictures look like vortex shedding from the hydrofoil.The large zone of vapour moving upstream is difficult to explain if no other data are available. It may be a result of boundary conditions or numerical errors / bugs in the code. If the main flow was accelerated somehow (e.g. by boundary conditions outside the view), there might also be cavitation forming like in a nozzle.
Thank you Michael, I will keep it to this for now and rerun the simulation for more data. I am attaching the pressure coefficient curve along the chord and link with scenes from the hydrofoil surface.
https://youtu.be/KjbP-jpxh9A
the pressure coefficient alone is not enough for evaluation, you need at least to indicate the vapour pressure in the diagram
Please see : Water Vapor Condensation in forced convection flow over an airfoil" Aerospace Science and Technology 12(2008) 150-158. Is it something similar? Alternatively please specify fully the case
I have come across some research on this phenomenon which is termed as vortex group cavitation, "...a special type of secondary cavities,
developed downstream an upstream-moving collapse cavity
and termed as “vortex group cavitation” by Bark et al [1],
appears to be observable in the simulation at this condition. The
existence of this type of cavity has been reconfirmed in a recent
experiment in the SSPA’s cavitation tunnel." https://www.researchgate.net/publication/280152928_A_modified_SST_k-_turbulence_model_to_predict_the_steady_and_unsteady_sheet_cavitation_on_2D_and_3D_hydrofoils
https://www.researchgate.net/publication/280230940_On_some_physics_to_consider_in_numerical_simulation_of_erosive_cavitation
Article A modified SST k-? turbulence model to predict the steady an...
Article On some physics to consider in numerical simulation of erosi...
I'm certain this is a numerical artefact. Cavitation can of course be initiated in a vortex, e.g. created in a shear layer following a cavity collapse as we discuss in the papers you list. But you wouldn't expect that kind of flows that far behind in the wake of that foil.
I'm not familiar with the mass transfer model implemented in StarCCM+ nor what outlet conditions or numerics you've applied, but that's where I would start looking to resolve the problem.
I simulated the NACA0015 some years ago, probably similar conditions to what you're doing here, where I also discuss the secondary cavitation of the type you mention.
Good luck!
Article Capturing Secondary Cavitation -- A Step Towards Numerical A...
Thank you very much for your prompt reply. I will be running the simulations again soon and try resolving the issue. A similar case is noticed for the steady RANS-omega simulation at σ = 1.6, but this time the vapours are closer to the foil. This scene is captured around 20 time steps (0.001s). i don't have access to high power computers right now so did not run the simulation further.
The structure close to the trailing edge could be physical, but the rest are spurious. Looks a bit like it could be in proximity to a mesh coarsening if you have a refinement wedge around the foil?
- Badges
- Science topic