Learning Forum

[hze5042]

Hi,

 

Water is coming from the top corner of my domain. The idea is that cool air moves from the bottom to the top while warm water moves from the top to the bottom. Water moves on a hot solid surface and will be vaporized.

Should I model water vapor, water liquid, and air as a 3-phase VOF sharp model or 2 phase with species transport (mixture of vapor and liquid water as one phase and air as the second phase).

Regarding the simulation time and faster convergence, is it better to define a layer of initial liquid water on the solid surface or just let the software handle the thickness of water on the surface from time=0 when it is released from the top surface?

What UDF should I use instead of the Lee model? Like the Hertz-Knudsen equation?

Any hints, information, similar problems, or tutorials are greatly appreciated

Thank you!

[Rob]

Have a look at the DPM & LWF models and then compare what they offer (and cost in terms of mesh & cpu) relative to VOF. 

With multiphase start with what you need to know, then what the flow regime is (droplets, free surface etc) and what hardware and time you have. Then decide which model is best suited for your needs. 

[hze5042]

Thank you for the reply.

Would you be able to answer my questions (stated in my first post) regarding VOF?

 

[Rob]

2 phases with species for the gas. But, that assumes you're going to use VOF. Cooling towers typically spray water to contact with the warm air. 

[hze5042]

How about evaporation on a corrugation plate?

Is 2-phase VOF (mixture of vapor water, and dry air as phase 1 and liquid water as phase 2) valid for evaporation on a corrugate dplate?

 

How about the UDF that I need to use? UDF for Hertz-Knudsen equation?

 

Thank you!

[Rob]

Look at how you'd resolve the film. VOF is viable, but can you afford to resolve the film? Is it a hot plate or hot gas? Two phase with water & air+vapour is fine. I can't comment too far on UDFs and the like, but check what the built in models are designed for: some are for binary systems so won't see the partial pressure effects. 

[hze5042]

So, let me give you a better picture of my new problem (please forget the cooling tower and my previous questions).

I want to cool down warm water using cold air while water is moving down the sheets and air is going upward

I have multiple vertical corrugated sheets. Water is coming at a velocity of X m/s from the top of the sheets while cool air goes upward at a velocity of Y m/s.  Warm water is losing heat due to convection heat transfer and evaporation (no boiling happens). I want to model this physics.

I am thinking of 2 phases (liquid water and water vapor+air). I am unsure how to model evaporation and if I need to consider a few mm of water at time=0 on the sheets (for a shorter simulation time) and expect the volume of water and its border on the sheets to be automatically handled by the software, or only consider inlet water at the top of the sheets with no water thickness at time=0 on the sheets.

 

[Rob]

Assuming the plate is cold or adiabatic. Have a look at the wall film models if the film is "thin" relative to the domain. VOF will work here, but as you'll want 5-10 cells across the film depth the cell count may be quite high. 

You then have water (liquid, possibly droplets) and a gas phase of air & water vapour. Depending on the approach used you then account for mass transfer using the model options. Until you decide on the model I can't suggest the mass transfer as VOF, EWF and LWF use different mechanisms. 

Choosing the correct multiphase model depends on what you need to find out from the model. 

[hze5042]

I am quite unclear what model I need to choose. I want to know the temperature of the outlet water and how much energy is released during the water's cooling. Do you think VOF is still the right move, or do I need to look for LWF? Thank you!

[Rob]

Yes....  

Do you expect the film to dry out, and do you want (and are you able) to resolve the film thickness? Is the plate cold/hot/adiabatic? 

[hze5042]

I do not know. I want the software to show me whether, with water at a certain X m/s at the top surface of the sheet and air at Y m/s from the bottom, the film will dry out or not. 

I guess I do not exactly know the answer to that (I have enough hardware capacity tho), but can you please elaborate on the model selection, like if I want to resolve and if I do not want?

The sheets are at room temperature first, but they will exchange heat with the surrounding air.   Also, I expect conduction from water to the sheet from one side, but also heat loss from the other side of the sheet due to convection and evaporation. 

[Rob]

OK, so for the wall you'll have gas flow on both sides, so it's internal to the domain? Otherwise you need to define some external condition, which it seems you may not know. 

To resolve the film VOF is good, but you need a high cell count, roughly 5-10 cells across the film and an aspect ratio that's suitable (potentially near 1 where the film tries to dry out). That's doable in 2d, but I'd not want too large a domain. You're then evaporating water to vapoour into an air stream so the default phase change models aren't ideal as partial pressure matters. VOF would be added using a velocity boundary, pressure on the outlet plus some "fun" making sure gas buoyancy is considered. 

With EWF (or LWF) you add the film via either a wall setting or droplets. They don't require anywhere near the same level of mesh resolution, can dry out and the various phase change models are different (and covered in the manual). It's much cheaper but you won't see a temperature gradient across the film. That will need 3d because of the way the model works, but that can be one cell thick. 

I can't tell you which approach to take as both are equally useful, and I'm not supposed to use experience and non-public knowledge on the Forum. The above is all public domain and/or in the manual or various marketing presentations.