[LeoRed93]

Hi everyone, I'm new to the forum, I hope I'm posting in the right section.

I'm trying to simulate the 2D release of superheated water out of an orifice, wich generates a supersonic vapour jet. I'm having some troubles determining the model to use: I've tried with the multiphase VOF model, with the HRM to simulate the flashing, but it uses a segrageted solver, so when the jet becames supersonic, it has some troubles solving the shock.

On the other hand, if I use the EMP model, or the Two-phase Equilibrium with the coupled solver, the simulation won't even start, because it cannot tolerate the interface between the two phases.

Any advice?

[JBeilke]

When you only have gases you have to use multicomponent-gas instead of VOF or EMP.

[LeoRed93]

Originally Posted by JBeilke

When you only have gases you have to use multicomponent-gas instead of VOF or EMP.

I don't have only gas, I have superheated liquid water (5Mpa, 537 K). The idea is to simulate the rupturing of the wall of a pressurized water tank, so the simulation starts in the moment in wich the water enters in contact with the atmospheric pressure through the hole in the wall. The external space is conveniently saturater with water vapour, to avoid having too many species.

[JBeilke]

I still don't understand the complete setup.

- do you include any phase change?
- do you treat the superheated water compressible?
- what about the equation of state for the mixture?
- do you expect a sharp interface between the phases or something diffusive?

I tried some compresisble multiphase simulations with ccm+ (and other codes) and the results are far from good. There are some people around, developing codes in this direction. Anyway, this is a real hot topic :-)


This might work when there is no phase change:

https://unicfd.ru/en/simulations/liq...merical-model/

[LeoRed93]

The setup is as follows:
I have a tank (a rectangular domain), connected to the external space (a semi-circular domain) through a hole (a smaller rectangular domain that connects the two domains). The tank is filled with superheated water, the external space is filled with vapour in standard conditions; the initial condition sees the free surface of the superheated water in direct contact with the external vapour through the hole (hence the use of the VOF model). When the simulation starts, the superheated water "enters in contact" with the lower pressure, so it flash boils (phase change, so the HRM model), generating a vapour jet in the semi-circular domain.
The water is considered compressible, so for both phases I'm using the IAPWS-IF97 properties table.

If it were merely a phase change problem, there would be several suitable models; the problem is that the jet becames supersonic in less than 0.1 ms, so the whole simulation needs to be able to resolve a moving shock, but the VOF with the HRM uses a segregated solver, wich sadly is not very good at it.