Two Stage HX Concept

[[XC] gomeler]

I think I have a viable alternative to enormous tube-in-tube or plate HXs but I want to run it past you guys just to answer a few minor questions and have it critiqued. The design is basically a 3/4" diameter copper pipe and roughly 12 inches long, capped on both ends w/ 3/4" end caps. Inside this pipe I will place 6 1/4" copper pipes that are 10" long that will be hammered nearly flat. All 6 will be merged into larger tubes on each end and these large tubes will then form the inlet/outlet for the high side. On the bottom of the 3/4" pipe the captube for the low-pressure refrigerant will enter into the hx and the suction side will be placed at the top of the pipe, drilled through the top end-cap. The idea here is that the hammered 1/4" pipe will provide maximum surface area will occupying the minimum amount of volume, therefore maximizing the cooling potential of this HX. Using 6 hammered pipes is just a wild guess, figure I'll see how much space they occupy and adjust from there.

My only concern will be oil return, if I am using an oil-carrying low-pressure refrigerant(r290) is it unreasonable to expect it to haul the oil vertically 12 inches when it turns into a gaseous state or should I lay this HX flat and let the velocity of the refrigerant and the suction suck it horizontally? Pictures below show the concept. Feel free to provide constructive criticism but please don't tell me to use a plate-HX, I already realize it would be ideal, I just want to test out a different way of doing the same thing. I figure something like this will be cheaper to fab than a tube-in-tube HX while maintaining a smaller volume with something comparable to a plate hx.

[n00b 0f l337]

Your refrigerant will generally, like most things, attempt the path of least resistance for the most part. I think you'll get an uneven flow and possible liquid floodback

[[XC] gomeler]

I understand the path of least resistance part but I figured 12 inches of vertical height to be a pretty good form of resistance to prevent any liquid floodback. Care to be more specific?

edit: Just realize the colors might be slightly confusing. Blue = low-pressure refrigerant ie r290. Red = high-pressure refrigerant ie r744.

[DetroitAC]

What you have there is a variation of what's called a "shell and tube" heat exchanger. If you'll look up a shell and tube heat exchanger, you'll see they usually have header plates, one at each end for a single pass of tubes or a single header plate with a baffle to split it into two sections with the tubes bent into "hairpins". They are easier to manufacture than what you're proposing, but not as easy as brazed plate exchangers.

http://en.wikipedia.org/wiki/Shell_and_tube_heat_exchanger

gomeler;2085383'] if I am using an oil-carrying low-pressure refrigerant(r290) is it unreasonable to expect it to haul the oil vertically 12 inches when it turns into a gaseous state or should I lay this HX flat and let the velocity of the refrigerant and the suction suck it horizontally?

oil-carrying has not much meaning once the refrigerant has turned to a vapor. Basically oils are miscible in varying degrees in different refrigerants. There isn't any such thing with oil and vapor (that I'm aware of). The oil is pushed along by the shearing force exerted on it by the rapidly flowing vapor (or gravity of course). Picture a puddle of oil in the middle of a table, and you blowing on it to move it. I'd lay that shell almost horizontal with a slight downward pitch.

You should also look at the concept of baffles inside a tube and shell exchanger. They are plates that force the shell side fluid to take a longer more tortuous path with more turbulence.

[[XC] gomeler]

Interesting read but I don't know if I can bend the copper with hairpin loops w/o compromising the integrity of the copper, will have to stick to single-pass with lots of conduits :) Baffles and a horizontal orientation it is, now I need to think of a way to construct the baffled. Think I might try to make small coils that ring the inside wall of the 3/4" pipe so that the liquid refrigerant is forced to flow over these "hot" pipes and hope it'll flash to a vapor rather than just streaming through and barely vaporizing. Thanks for the information, tomorrow I'll be grabbing the few remaining parts I need and building this thing.

[n00b 0f l337]

Have you considered at each end putting a 1/2" pipe or the like, drilling 1/4" holes into it and regrouping them all into that pipe? Seems that would be pretty easy.

[[XC] gomeler]

That was the idea, have them all plug into a manifold of sorts while inside the HX to simplify the inlets/outlets and weakspots in the casing.

[n00b 0f l337]

Go for it then I say, but use caps and such and pressure test it before you use it in a cascade. Shells can = bombs.

[epion2985]

I think it already has been done and is a variation of a tube in shell:
http://www.usui.co.jp/images/i_products/spiral/spiral_l.jpg

[[XC] gomeler]

Ooohhhh how does one go about acquiring one of those. I thought I was being original/unique in my design, took me a few hours of solid thinking on how to emulate a plate HX with the parts I had, oh well still going to build it. Just got back from a 2 mile round-trip run to linde gas to exchange a CO2 cylinder, nothing like hauling a 20lb cylinder on your back, they did give me an alumnium cylinder in exhcange though which made it a lot better on the run back. Now time to run the ~6 miles round trip to Home Depot, I seriously need a car.

[Brettbeck]

@ epion2985 - WOW that thing is HUGE! And a smaller plate hx is probably more efficient!

@ [XC] gomeler - Good look with your build. Make sure to purge well when brazing it all up :)!

[epion2985]

@ epion2985 - WOW that thing is HUGE! And a smaller plate hx is probably more efficient!


Of coarse. But gomeler wants to be different :)

[Brettbeck]

Of coarse. But gomeler wants to be different :)

There's allways one eh :p: ?

There will of course be a specific use for them, but I can't see why you wouldn't just use a plate hx :).

[epion2985]

There's allways one eh :p: ?

There will of course be a specific use for them, but I can't see why you wouldn't just use a plate hx :).

I concur.

[wdrzal]

gomeler;2086788']Ooohhhh how does one go about acquiring one of those. I thought I was being original/unique in my design, took me a few hours of solid thinking on how to emulate a plate HX with the parts I had, oh well still going to build it. Just got back from a 2 mile round-trip run to linde gas to exchange a CO2 cylinder, nothing like hauling a 20lb cylinder on your back, they did give me an alumnium cylinder in exhcange though which made it a lot better on the run back. Now time to run the ~6 miles round trip to Home Depot, I seriously need a car.

This is what I keep telling you guys, instead of hours of "what ifs" go to the library and check out some good books on refrigeration & spend that time reading a few text books. you will learn the basic physics and the type of equipment & controls used. you will learn more in reading a few books than the info contained in all forums combined.

remember your NOT a designing engineer yet, but that is what many of you are doing, playing with pressure & flammable gases(chemical energy) can be very very dangerous unless you understand pressure & how pressure & temperature work and what it takes to contain that pressure safely .PSI is just not a number. start by memorizing the psi different sizes/thickness of copper tube can contain,the larger the tube the lower the pressure because of surface area. You should know what will happen before you flip the on switch or open the valve........Home made expansion tanks or receivers ,separators can be very dangerous because of size.

Lets all work safely, refrigeration is much more complex than most realize, there is no way anyone can teach everything you need to know in a few short paragraphs.

Walt

[hatemi]

I would never ever use plates or sheet copper to cap a pipe. I have made phase seps from 42mm pipe and I crimped the ends rather than braze a plate to them. I have made a tube in shell HXs with 42mm pipe but I used proper endcaps. The real maximum preasure in the HX is only the first stages static preasure but I wouldnt want to risk anything.

The best pipe in shell design I have used was 2x3m of 4.2mm OD pipe inside a 28mm pipe. I wrapped the 4.2mm pipes around 15mm pipe(crimped shut at the both ends to reduce the volume inside the HX) and put everything inside the 28mm pipe. That worked out real well and the HX was realy small. I used that to replace a HX in a Chilly1 cascade that UPS destroyed.

[epion2985]

While safety goes without saying, I find it amazing that every other person that comes on these forums thinks they will reinvent the industry or topple the best designs out there without any education and little understanding on what is actually at play here. And while in the infancy of the industry that may have cut it it most certainly will not cut it now :)

[[XC] gomeler]

I love how everytime someone tries something slightly different we are then told to stop trying to invent the wheel... I'm simply re-creating a wheel but instead of using steel I'm using aluminum, as an example. Yes, I could go out and pay ~$110 for a plate HX that'll be capable of transfering many kilowatts of heat or I could use my head, do a few calculations, and purchase the correct materials to do the job. I understand the basic gas laws, I had them hammered into my head in chem1 at my school, I understand that I shouldn't use aluminum foil to make a HX, and I also understand that materials have their breaking points. I also understand that if you overspec everything, use proper materials, use proper techniques, and double-check yourself on everything then things will work out. It's a good thing everyone doesn't stay inside the box otherwise we'd have a stagnant society.. In rant-less news I settled on 1" copper pipe with 1-1/8" copper end-caps for both my hx and phase sep, even if all else fails then my release valve will blow at 375 psi and I'll hear something to rival a nice b.o.v. Once again guys this isn't someone trying to re-invent the industry, this is someone using the right materials to create a product for his specific application. Safety first, as always. One other thing, did anyone notice the recent autocascade thread that dissapeared, it would be rather nice to have recent works to compare to, the 18 page autocascade thread is too much work to sift through for the third time to find the relevant data.

[wdrzal]

While the industry is changing, because of eliminating harmful refrigerants to our environment,its very unlikely someone without the basics is going to "discover" the new best thing in refrigeration. Read and study a little and your knowledge will get a big boost.

Those of you using co2 should be studying supercritical co2 used in a trans-critical co2 cycle.

I bet detroitAC is doing a lot of work in this area if he's still working in the designing engineer automobile field.

[DetroitAC]

I was doing much more CO2 work in 2002-2003, but I'm in more of a compressor position now. I did gain a good understanding of why the transcritical cycle is difficult to control, and why the compressor is relatively expensive and difficult (120 bar high side design pressure limit :toast: )