strange behavior in interPhaseChangeFoam
Hi,all:
I'm doing a cavitation simulation around the hydrofoil recently with interPhaseChangeFoam. However, I found quite strange results from this solver. Description of my model: A hydrofoil fixed in a channel , the free stream velocity (inlet velocity is U = 5.33m/s), the cavitation number is defined as : a = (p_out - p_sat)/(0.5*rho*U^2)=1.25, as in interPhaseChangeFoam, the reference pressure is set to 0 Pa by default, the pressure in the calculation is the absolute pressure. Therefore, p_out = 20055.6 Pa according to the definition of the cavitation number. The mesh used is about 20000 elements, with yPlus ranging from 37 to 267, which is suitable to use the wall function. First of all, I ran a case (3D, 0.5 million cells) in simpleFoam to see the pressure distribution under noncavitating condition, here is the comparison with experiment. http://www.cfd-online.com/Forums/mem...cavitation.jpg The result is in acceptable agreement with exp |
As 3D simulation is time consuming, I study 2D case in interPhaseChangeFoam.
Still, I'd like to see if the noncavitating case agrees with the exp. So I set the following boundary conditions:(the estimation of k and epsilon can be found http://openfoamwiki.net/index.php/TurboECPGgi2D and http://support.esi-cfd.com/esi-users/turb_parameters/) alpha1: Code:
INLET Code:
INLET Code:
INLET Code:
INLET Code:
INLET Code:
INLET |
the transportProperties:
Code:
phaseChange on; Code:
ddtSchemes Code:
solvers The resulting Cp at x/l = 0.1 is about -2, which is much lower than the exp, indicating a much lower pressure than exp. While the calculation is at least stable. I used to think the time scheme may affect the result, so I change it to the second order backward/ Crank , however, the two schemes used can easily lead to overflow> I have got stuck on this problem for days:( Is there anyone who can tell me what's wrong? Any advice is appreciated.:) Best regards Xianbei |
Hallo to everyone in the Forum.
Huang, have you figured out how to improve your model to avoid an over estimation of the cavitating phenomena? I have, more or less, the same set up you use and I obtain a cavity longer than the experimental. Thank you in advance Tom |
Quote:
The reference is here: http://www.tfd.chalmers.se/~hani/kur...ChangeFoam.pdf The major modification is: from Code:
max(p - pSat(), p0_)/max(p - pSat(), 0.01*pSat()), Code:
max(p - pSat(),p0_)/max(p - pSat(), 0.001*mag(pSat())), Maybe this method can improve your results. Xianbei |
Differences of InterPhaseChangeFoam between OF 2.3 and OF 4.1
2 Attachment(s)
Dear All,
I'm having similar problems in the case of simple cavitation estimation using interPhaseChangeFoam. My test case is the well known NACA66 hydrofoil by Shen and Dimotakis (Shen YT and Dimotakis PE. The influence of surface cavitation on hydrodynamic force. In: Proceedings of the 22nd ATTC, St. Johns, NL, Canada, 8–11 August 1989, pp.44–53.). I already simulated this hydrofoil using OpenFOAM 2.3 with quite satisfactory results. When I moved to OpenFOAM 4.1, I re-ran some old calculations to check if everything behaves similarly. Unfortunately this is not the case. As you can see from the figures below, using OF 4.1 leads to significantly longer cavity bubbles (longer than OF 2.3 and in turn, to the experiments). Calculations have been carried with exactly the same setup: same mesh, same numerical schemes (except for the new conventions in fvSchemes of OF 4), same numerics (exactly the same fvSolution file). Calculations were initialized with the same non-cavitating calculations in order to provide a realistic estimation of the pressure around the hydrofoil and avoid/limit the initial transient. Do you have any explanation circa these differences? Many thanks, Stefano |
All times are GMT -4. The time now is 00:23. |