I want to model two phase flow of hot compressible gas (H2 at 2000K) and water ( 300K) using Euler- Euler approach. The geometry is similar to that of a Tee Junction. Hot gas stream from one inlet flows through the channel and cold water impinges the gas jet stream at the center of the channel. Which model should I use in FLUENt, VOF with evaporation or Euler- Euler (Boiling model)? It is a high mach number problem. Do you think it possible to model such a problem in FLUENT?
Depends on what exactly you plan to simulate. Is it important to capture the interface between two fluids? If not, then perhaps VOF won't be necessary. It is again true if you plan to have H2 and water streams with different inlet velocity because in VOF, you can only have a mixture velocity. What phase velocity you are going to use? If Mach no. of the system
The interface is important to capture. Mach no is higher than 0.3. H2 and water are flowing from two different inlets. I may use mass flow rate as boundary condition. I am trying to model the evaporation of liquid and at what place does it take place.
1) Yes, it is possible with the fluent.
2) In Euler-Euler approach both phases are considered as interpenetrating continua; It does not allow to track interface. However, VOF methodology allows to track the interface. Consequently, you should go for VOF model.
3) You can use VOF model for compressible flow.
1) Yes, it is possible with the fluent.
2) In Euler-Euler approach both phases are considered as interpenetrating continua; It does not allow to track interface. However, VOF methodology allows to track the interface. Consequently, you should go for VOF model.
3) You can use VOF model for compressible flow.
Try the evaporation-condensation model to use as a mass transfer mechanism between H2 and water stream. You would need to use species transport model to simulate evolution of water vapour. Also in pressure based solver in VOF framework, only one phase can be considered compressible which should be H2 in this case.
@ Sumer @ Subhasish, Yes, am using using VOF approach and trying to run a cold-flow simulation for the two phases. I am facing some issues:
1) I have two inlets for two phases ( H2 and water)and both of them intersect at TEE joint. The velocities of both the inlets are very different h2 inlet is at 1m/s and water inlet is at 0.001 mm/s.
2) From my initial simulation what am observing is the whole domain is filled with H2 and water just does not flow in. There is a mix of both volume fraction in the water inlet.
3) I am not sure what is wrong in my case. Since i presume, that water is incompressible so it can't get pushed over by gas.
That always happens in numerical simulation when you have one stream flow few order of magnitude higher than the other. You can try this: first patch a certain part of your system with water and then inject H2 at a slower flow rate comparable to the water stream. Time step size is critical for VOF model stability, so make sure you keep Courant no. < 0.25.
I tried that approach of gradually increasing the gas flow rate, but that does not seem to work for me. At lower flow rates, i see water in the domain, but as soon as i increase it to the desired rate, water just goes back near inlet.
I am using steady state solver to start with, so that i can get some quick solutions. For keeping Courant no < 0.25, the timestep for my simulation is around 1e-8, which is way too small. Due to this reason i switched to steady state solver.
Any other technique that i can employ? Due to different in magnitude of the two phase, am kinda stuck in the problem.
no multiphase simulation should be done in steady state since inherently these systems are unsteady. Not sure what is going wrong there. You have two velocity inlets - and you provide velocity magnitude and phase volume fractions equal to 1 to specify pure components at each inlet and then you have one pressure outlet where you set the system pressure and back flow volume fraction of water equal to 0.0? Is that what you are setting?
Yes, thats my system settings. Multiphase flow can be solved using steady state solver if you are not interested in the initial transient state. Since my problem geometry is that of T junction, am interested in the gas- liquid interface at the intersection. So at each of the inlets there should be only pure gas and pure water since the volume fraction is one. But that is not what am observing, I see gas flowing into the water stream.
Check the material properties: Diffusivity of A-->B and B-->A. There seems mass mixing.
Can you mention all the boundary and initial conditions? Also, thermo-physical properties.
I have no mass transfer model between these two phases. I am using just the basic two phase flow model of water and hydrogen. The problem is the difference in orders of magnitude of velocity.
May be you are not viewing converged/steady state data. Can you mention 'all' the boundary and initial conditions? Or, recheck?
I think you start with VOF model. See the results that it makes some sense
I apologize for responding for so along. In all my previous models, i did not activate surface tension model between hydrogen and air. I changed my initialization procedure and seem to get some reasonable results.
Could you share how you changed the initialization procedure, please?
I am trying to simulate an air bubble injection from a mass flow inlet at the bottom of a half-filled tank, using Eulerian model with compressible air. I didn't activate surface tension model. I see there could be only one compressible gas in Eulerian model. However, I don't know how to set up the initial procedure. Thank you!
- Badges
- Science topic