[kimotbwb]

Hallo everyone,

i want to set up a case in CFX 14.0 including the simulation of cavitating flows around a hydrofoil.
Cavitation around hydrofoils is already explained in the CFX tutorials, but i want to perform my own studies, looking further in the physical behavior. So for example, i am interested in running a transient calculation in order to better reproduce the physical cavitation flow pattern.

Am i right, that the tutorial is only performed by steady state ?
Why didnt they use the unsteady solver?

I am not encouraged by only running a steady state calculation including cavitation, because especially at high Reynolds flows, the flow pattern in cavitation regions becomes highly transient, and flow is convected downstream at high Re.

I just wanted to know if everybody agrees with the following proceeding, setting up a cavitation calculation in CFX:

1. steady state result for incompressible flow. Using the res-file of this case as an initial guess for later transient calculation.

- outlet pressure of the flow domain is set high (e.g. 100.000 Pa @ Re=1.3e5; isothermal vapor pressure of water @ 298 K is about 3500 Pa)
- the high outlet pressure ensures, that the cavitation number wont drop as far towards zero that cavitation could accur.
- using a homogeneous mixture model including liquid water and water vapor @ 298 K while heat transfer is set isothermal
- Fluid pair model mass transfer is set to "none"
- interphase transfer ist set off
- as a fluid pair model of liquid water and water vapor is used, the boundary conditions for the volume fraction at the inlet ist switched to 1 (only liquid water) an 0 (no water vapor) (alternatively no liquid pair model is used, an only liquid water filles the domain. But i think this is the same procedure as pointing out the liquid pair model at setting the volume fractions to 1 and 0)
- no slip wall at the wing profile and SST turbulence model
- low physical timescale of 1.0e-3 s and convergence criteria RES_MAX=1.0e-5

Ones the simulation case in 1. is done, the result is used as an initial start value for the unsteady run with cavitation:

2. compr. unsteady flow

- settinsg of case 1. are basicly used
- static outlet pressure remains at 100.000 Pa
- cavitation model (Rayleigh Plesset) is switched on and saturation pressure of 3500 Pa is selected
- using a low time step of 1e-05 s
- using second order backwards time scheme and high resolution advection

I suppose that the outlet pressure should be decreased consequently to provoce cavitation. Even if cavitation model is turned on, cavitation should only appear in regions where the static pressure in the flow field falls below the saturation pressure. Nevertheless cavitation is also provoced by a few number of other influences....

Does anybody agree with my procedure?

I would be very thankful for further help.

Best regards!