[the.drizzle]

Well, that's the simplified version of what I'm trying to do anyway...

Basically, I'm attempting to model heat transfer into a liquid held in a compound material vessel, surrounded by air. As a starting point, I have worked through the chtMultiRegionSimpleFoam tutorial provided here, which is good enough for a first step except it clearly won't help me model natural convection in any way, nor does it seem to allow for incompressible fluids.

Thus, my only question at this point is which solver I should be looking at in order to model this situation? That is, stagnant liquid in a vessel, surrounded by a solid wall which is in turn surrounded by (hot) air, and I want to look at the heat transfer to the (cold) liquid. Getting some rough steady state results akin to those demonstrated in the above mentioned report would be plenty good enough for the moment; perhaps I can trick chtMultiRegionSimpleFoam into using an incompressible fluid?

Thanks in advance any suggestions.

[zhengzh5]

Originally Posted by the.drizzle

Well, that's the simplified version of what I'm trying to do anyway...

Basically, I'm attempting to model heat transfer into a liquid held in a compound material vessel, surrounded by air. As a starting point, I have worked through the chtMultiRegionSimpleFoam tutorial provided here, which is good enough for a first step except it clearly won't help me model natural convection in any way, nor does it seem to allow for incompressible fluids.

Thus, my only question at this point is which solver I should be looking at in order to model this situation? That is, stagnant liquid in a vessel, surrounded by a solid wall which is in turn surrounded by (hot) air, and I want to look at the heat transfer to the (cold) liquid. Getting some rough steady state results akin to those demonstrated in the above mentioned report would be plenty good enough for the moment; perhaps I can trick chtMultiRegionSimpleFoam into using an incompressible fluid?

Thanks in advance any suggestions.

why do you think chtMultiRegionFoam doesn't handle incompressible fluid? As far as I know, you can always use rhoConst as the equation of the state and specify the constant fluid density inside thermophysicalProperties.

I think chtMultiRegionFoam/chtMultiRegionSimpleFoam is exactly what you need in this case.

[the.drizzle]

I do believe you may be correct, but my confusion came from reading the document I linked in my original post. Thus, as a first step what I'm attempting to do is simply translate the the 1-D example provided on the "Getting started with chtMultiRegionSimpleFoam - planeWall2D" OF wiki page.

It is not going terribly well...

Basically, I have used the multiRegionHeater example provided in the tutorial as a roadmap and attempted to set up a simulation here with cold water flowing in the top region, hot air flowing in the lower region, and a homogeneous wall separating them--much like the planeWall2D case provided on the wiki page.

I am very much at a garbage in = garbage out result at the moment, primarily as I'm having trouble figuring out the fvScheme / fvSolutions options for this solver, and have employed a lot of copying from the tutorial in the attached case file.

Any help shedding some light on how I should set this up properly (and I'm assuming my IC's/BC's are off as well) would be greatly appreciated. Not so much in terms of making this work in some sense, but more pointing me to some enlightening documentation so I can understand everything better.

***EDIT***

I'll leave the above comment there for now for continuity, but it would seem the bulk of my problem was due to having nonsensical thermophysicalProperties dictionaries. Aspects of this simulation now performs as one would expect it to with the IC's / BC's as provided on the wiki page example. I've defined the case as having hot air flowing underneath, cool water flowing above, and an aluminium wall between the flows, with the flows driven by pressure.

So the next thing is to adjust the BC's / IC's to simulate something more sensible, and properly understanding the fvScheme / fvSolutions options for this solver. Any help still greatly appreciated.

Thanks!

[er_trebbia]

Hello the.drizzle

I'm having some problems too, about chtMultiRegionSimpleFoam

My simulation is a steady state heat transfer in a jet impingement problem. So I have a slot jet of air impinging a heated cylinder, (I have to mesh also the width of the cylinder, because of the non-negligible circumferential heat flux)

As far as I know, this solver doesn't work anyway in incompressible cases, (I'm trying to implement something...)

About fvSchemes and fvSolution files, I have two files that are working for me, I haven't liquid domains in my geometry, but you can easily figure out how to change them I guess. Let me know.

[the.drizzle]

Hi there,

My problem stems partially from the fact that I'm using chtMultiRegionFoam instead of chrMultiRegionSimpleFoam, as it seems to handle transients.

Anyway, after a lot of going bananas trying to figure out why things were failing so bad, I stepped away from this for a couple days and came back a couple hours ago to see the embarrassingly simple solution staring right at me. Sort of a forest for the trees sort of thing, and not including the empty patches in a couple boundary conditions. My bad.

So, the attached case seems to work fine, although it is rather time consuming and I have not fixed up the decomposeParDict file to run this distributed, but that's easy to deal with if anyone is interested.

The case simulates air flowing between flowing water, separated by a huge slab of aluminium. I think the BC's I have defined may be a bit absurd from a physical point of view, but the simulation yields the expected results--air gets cooler, water stays the same, localized cooling on the surface of the aluminium near the wall. What I should have done was made the air and water the same temp, and defined the aluminium as a heater instead, that would have given some more enlightening outcomes I think.

Oh, and because it occurred to me in my "ah-ha" moment that having a symmetry plane top and bottom with gravity pointing down made no sense in this set up, I simply changed it so gravity points to the left now instead, with both fluids moving from left to right.

I hope somebody finds this useful, it's been a very good learning experience for me, even if the boundary conditions are still a bit naff.

[the.drizzle]

Hm, actually this still isn't quite correct as the liquid is doing pretty much nothing...

Close though, close.

***EDIT***

Nope, nevermind--it is in fact working as one would expect it to--the lack of seeing any impact on the water side of things is due to the scale of the simulation and the velocity of the water as well.

That, and there is a typo in system/topWater/changeDictionayDict that doesn't really matter, a } is missing. Currently it's

    cellToRegion
    {
        boundaryField
        {
            topWater_top
            {
                type        symmetryPlane;
            }
        }

            frontAndBack
            {
                type            empty;
            }

    }

which of course messes things up, but since this item isn't actually used in this simulation it doesn't break things.