Printable View
I posted this question a few weeks ago but got no repsonses, maybe I rambled on a bit too much.
Can anyone point me towards a formula for estimating the heat transfer coefficient from a condensing R134a vapour inside a tube to the tube internal surface?
Happy New Year!
James.
Well my first Google search for " heat transfer coefficient from a condensing R134a vapour"
found this, start on page 16..
http://www.intechopen.com/source/pdf..._mass_flux.pdf
or this...
http://www.ijest.info/docs/IJEST09-01-02-03.pdf
there's lots out there.
Thanks Brian.
My question relates to a horizontal smooth tube condenser carrying R134a, immersed in seawater, cooled by natural convection.
After extensive Googling I found a paper by Cavallini ('Condensation in Horizontal Smooth Tubes: A new heat transfer model for heat exchanger design') which gives a model for the local heat transfer coefficient in terms of the vapour quality and the liquid phase heat transfer coefficient which I assume can be calculated using the Dittus Boelter equation.
I'm thus modelling my condenser in 3 parts:
1 - the superheated vapour part (Dittus-Boelter)
2 - the condensing part (Cavallini)
3 - the subcooling part (Dittus Boelter again).
For the calculation I will divide the condensing section of the condenser pipe into lots of short segments.
I will write down a system of equations describing the heat transfer coefficient from the vapour to the tube inner wall, the wall thermal resistance, and the convective thermal resistance from the tube outer wall to the seawater, then solve for each tube segment to give me a heat flow rate which will allow me to calculate the quality of the vapour exiting that segment and entering the next one. Start at the first segment (quality = 1) and iterate until all the vapour has condensed. This should give the required length for the condensing part of the condenser.
Is this a sensible approach? Any obvious mistakes? Anybody have experience of this model?
I've not been able to find a worked example, however my Googling ability isn't up to Brians standard!
James.
Long time since I worked with heat transfer correlations but the problem is not as easy as it seems.
Dittus Boelter does not account for heat transfer that by radiation that may be important with high inlet superheat, so you may underestimate.
The flow turns quickly into mist flow und unstable annular where if I remember no correlations were good.
Cavallini may be good for stable annular flow but you need long tube runs for it to develop. In the moment you hit the first curve it turns into nothing of the sort.
Then you get all kinds of wavy flows which again I remember no good correlations. I don’t think Dittus Boelter will predict good in this situation but when you get a tube full of liquid refrigerant and the amount of heat transfer here is very low because of the low temperature differentials.
I really don’t remember but I’d say that with these correlations you will explain 1/3 of the heat transfer a condenser can do.
If you want to dimension your coil, It will work because you will overestimate the surface this way, but for an economic solution you have to run tests and find experimental correlations that most (serious) manufacturers do but keep as a secrets to themselves.
The ones you get in Internet like Brian_UK mentions do not apply exactly but may give you more insight on the problem.
Cheers Aramis,
I was coming to the conclusion that a rough initial guess followed by some thorough experimentation is going to be the most appropriate approach.
I've built what I hope is an oversized condenser which I will run under worse case conditions, reducing its effective surface area in steps and logging its performance.
Thanks for your help,
James.
TKS, but I wasn’t much of a help!
But I’m confident it will work if your guess is not too off, for the same reason you don’t hear of users suing condenser’s salesmen.
You’re probably considering differentials less or about 10°C (18°F) but with seawater you could go 4 times higher and your system would still work in safe (though not optimal) conditions.
Besides the expected live of this system is less than others.
The only problem you will run into is the cost of the materials, though you may use cheaper stuff that would't normally last in seawater, for your tests.
Cheers to you too.