SimScale CAE Forum

Anyone looking for a small freelance multiphase CFD project?

#1

Hi guys, I’m not sure if this is the correct place for this posting. If not, please point me to where I should post instead.

There is a lot of interest in atmospheric water generation these days, and I have a specific in seeing the multi-phase CFD analysis for something simple like this: https://grabcad.com/library/sun-pumped-ground-cooled-condenser In particular, I would like to know whether such a system can even function in a purely passive fashion (as indicated in the design) or whether some major tweaks would need to be made for it to even function. In particular, I’m most concerned that the boundary layer friction in the coils would pretty much prevent it from working at all, hence I’d like to simulate it first :wink:

I’m happy to see that there is a fantastic cloud-based system like SimScale out in the market now, as it makes things much more accessible to non-experts! I have not done any form of CAD or CFD in many years, so I’m really looking for a mentor or expert who can help get me started by setting up a first-pass model and letting me tinker with it from there.

I don’t have a huge budget as this is a personal project at the moment, but I would be happy to pay someone for assistance. Is there perhaps someone in the community who might be willing to take this project on?

Thanks for your interest!

#2

Hi @ericbsb,

Thanks for the post and welcome to the forum :smile: So you are looking to see whether this boundary layer friction can be overcome, interesting concept!

I’ll throw this one to the CFD guys to see if they have any thoughts/other referrals (@dheiny, @Ali_Arafat, @gholami, @pfernandez, @akosior, @oscarcorripio, @dylan, @TobiasHolzmann )

Best,
Anna

#3

Thanks @AnnaFless Also, I should mention to the CFD guys that I realize it may be a tall order to have the bottom tank venturi suctioned up to the top tank. If it makes the initial calculations easier (ie: system more feasible), I would be fine skipping that step and just letting the water pool in the bottom tank. I’m mostly concerned that even without the water pumping we might not be able to get over the boundary layer friction within the tubes.

Something just isn’t passing my smell test here and I’d like to figure it out before I go try to dig a hole in my backyard and test it, lol! Happy to entertain anyone else’s thoughts on the subject too!

#4

Hi @ericbsb,

I am not an expert but did you consider the NPSH (net positive suction head) which includesthe parameter for friction losses. But since you mentioned venturi I am not sure if you are using a pump now. Interested in the discussion! Your project sounds pretty cool :slight_smile:

#5

@jousefm Thank for your response and interest. Sorry for the confusion with the venturi comment. I was not referring to an actively driven pump, just merely the passive effect due to airflow. The original diagram had a small line going from the buffer tank up to the above ground tank, and my thinking is that just might be too much for the system to do passively, so I figured that I would leave that part out for now.

As for the NPSH, are you referring to the NPSH that would be available at the outlet of the upper hose once (or if) the system is in steady state?

#6

Something like that yes. Just threw it in the discussion :slight_smile: The first things that came into my mind were NPSH and Bernoulli :smiley: Where NPSH often refers to cavitation effects. Did you do any calculations by Hand and see what Magnitude of Pressure you would need for Venturi ?

#7

Dear ericbsb,

the project you are trying to do is not be an easy task as it would look like. First you need a model that can handle the problem in a good way. Due to the fact that you have air / water this would more or less lead to multiphase solvers (that is always expensive). Further more, as you can imagine, the water you could get due to the condensation is more or less - not a big amount, isn’t it (some droplets in a few minutes / hours?). I am sorry but I do not have any idea about the volume fluxes of air and water in that kind of geometries.

Further more you have a separation of humidity air to dry air that will form droplets in the pipes. You see that I can even bring more and more aspects to the simulation. I think the main challenge is to figure out the correct model but as far as I know, FOAM will not offer you a model that will directly handle the stuff you want.

Other topics would be - meshing etc.

What you need would be a solver that includes the enthalpy equation and hence a temperature depended solubility of water in air and a two phase model (water and air). The VOF method would be the wrong approach. Maybe some Euler-Euler approach. If the pipe diameter compared to the length is in a special ratio, I would neglect at the beginning.

In my opinion the critical thing is meshing here (it is definitely not a big deal to mesh something like that, but the amount of cells you get is important).

Hope I could help you a bit :wink:
Tobias Holzmann

#8

@TobiasHolzmann Thank you for your very thoughtful comments! I think that you bring up some really good points. Being a non-expert in all of this, one thought I had is to break the design down into different components and look at those components in a more “idealized” state.

It seems to me that the driving force as to whether this water generating concept produces any meaningful results would be to see what happens inside a coil which spans two compartments of different ambient parameters and assume that the compartments themselves are insulated quite well from one another. p% of the coil would be in one compartment and (100-p%) of the coil would be in the other. So rather than analyzing the full cycle, I’m wondering if we could just put together a very “simple” model of a coil since that is where all the magic happens. What do you think?

#9

@ericbsb … the way you want to analyze the stuff is very good. Splitting the problem into single simple one is often the best choice. I am thinking if you could use a cyclic boundary condition to simulate only a part of the coil, lets say 5% and let the fluid pass it 20 times. This is a common practice in turbine analyses. I never did it in such a way but it could work. But I have doubts due to the fact that the coil is going up on z direction and normally the patches should be similar. It would be definetly worth to try it. Otherwise you could simulate as you suggested.