Morphing Air Conditioner into Autocascade System

[quintus]

Sticky well worthed! i mean there is so much good info in this thread!!

[mytekcontrols]

Chris thanks for making this a sticky. I just wish I had more time to put in the updates. As for the AC-3 Prototype, it made it down to -141C with a predominately hydrocarbon based charge (and a little bit of R14). Was able to use the stock AB oil that came with the compressor, and no coalescing oil filter. The R600/R170 combination as the 1st 2 refrigerants in the mix, seemed to do just fine at pulling the oil out at the phase separators before the system had a chance to freeze it up.

Performance was good, although I did not add any heat as in a test load, I did see rapid cooling from a warm start condition. Went from a +25C Cold Head temperature down to -125C in 30 minutes, a total decrease in temperature of -150C. Took another 25 minutes to reach -140C, and leveled off at -141C in another 10 minutes.

This concludes my tests on the AC-3 Prototype for now. What's next? On to the new AC-4 design. What is that? Do the same in an even smaller package. Shooting for meeting a footprint requirement of 9" x 9" at not much more than 12" in height (although this dimension has some fudge factor).

Here's a pic of pretty much where I left off on the AC-3 Prototype:

[yngndrw]

That's awesome Michael. I see you've gone with the water cooled condenser to reduce space requirements.

Have you used plate heat exchangers in the stack this time too ?

Also sorry if this has been asked before, but what is the overall purpose for these prototypes - What will they be used for ? (Or are they just for playing with ?)

[mytekcontrols]

yngndrw -- No plates being used in the HXC Stack, although I did do some fiddling around in this area, and failed because of my lack of experience with the strict orientation requirements (vertical is best).

As for what are these prototypes to be used for: My ultimate goal is to produce a product that serves the need of water vapor cryopumping in the Hard Disk manufacturing arena. Several of the specialized coating machines designed for this specific application use a series of very small low heat load cold heads, one per coating cell. Currently they use very large power hungry (10 HP) Polycold autocascades to do this, with one Polycold feeding 5 cold heads in series. My proposal is to create a small enough unit with a cold head directly exiting the side, that each Hard Disk coating cell will have its own cryopump. Since some applications have the cold heads spaced at 9" apart, the unit's footprint will also need to meet this requirement (very small indeed).

So the AC-3 unit was an early prototype aimed at this application, but one that had an unacceptable footprint. Essentially it helped me prove out some concepts such as providing a hot gas defrost without the need for a coalescing oil filter, develop a mostly hydrocarbon based mixed refrigerant charge, and establishing the minimum number of HXC stages required to meet my temperature requirements. It also showed me some alternative construction techniques.

[yngndrw]

Well you're well on your way to meeting your goal and it looks great so far. Good luck with the project.

[LeandroJVarini]

man you developed an incredible project! congratulations

[quintus]

Mytekcontrols, any update man???
I'm bored with all the normal cascades around here
The autocascade is so much beter than a normal cascade i feel sorrow for myself for having a normal cascade
....
..
.

[wdrzal]

Mytekcontrols, any update man???
I'm bored with all the normal cascades around here
The autocascade is so much beter than a normal cascade i feel sorrow for myself for having a normal cascade
....
..
.

Mytekcontrols did a outstanding job on this auto cascade project,not only building but in the support he gave for auto-C's advancement to the Xs community. . I think this build thread was the first to truely autocascade all the gases.

But Mytek will even tell you it was for a low wattage ,50 watts iirc. and he has a lifetime in experience behind him.

Going from 50 watts to 300watts is going to be diffucult especially for a DIY builder.

Many others who tried auto-c's failed or only got them to partly autocascade. I don't recommend one at all unless you well versed & experienced and have a very complete shop full of tools,most never even get mentioned on XS.

These are for the very experienced, I warned severial people in pm to stay away from auto-c's and they lost their a$$ and reputation when they couldn't deliver.They are much more technicaly diffucult than a classic interstaged cascade.

Mytekcontrols if you disagree with my opinion please feel free to point out where I,m wrong.

I hate to see guys attempt a project and in the end have one that doesn't do the job. Now if you want to start with myteks design and progress it to a higher wattage ,great,but be prepared for one step forward and two steps back. But if you have the desire & funds,by all means give it a try.

[mytekcontrols]

Yep I agree with you Walt. And yes the last design (AC-2) based on the oblong coiled HXC's worked very well, and actually out did my next design (AC-3) which had a higher CFM compressor.

Although I learned a lot from the AC-3 design, I ended up abandoning it, and have since been making plans for the next version AC-4 (albeit I must admit much slower than I originally thought).

I also agree with Walt on taking what I did with the AC-2 design, and scaling it up for 300 watts might be the way to go for your guys end application. Of course winding the HXC's in an oblong fashion would not be necessary, if tube-in-tube coaxial HXC's were to still be used. A scaled up design would also benefit from flat plate HXC's to keep the package size reasonable.

But be prepared to spend quite a bit of time and money undertaking such a project.

As for refrigerant charge, I think the following components would be appropriate, as based on my AC-3 tests.

-130C Charge (or better with increased Argon)

• R600 (Nbutane)
• R170 (ethane)
• R14
• Argon

-80C Charge

• R600 (Nbutane)
• R1150 (ethylene)

Keep in mind that there is a tremendous difference in required compressor mass flow requirements and number of stages required to go from -80C versus -130C or colder. I would say that the AC-2 design, which was based on a 6500 btu compressor, would stand a pretty good chance at coming close to -80C operation at up to 200 watts, assuming that the last stage was removed (eliminate Cascade Condenser #3, its associated phase separator, and increase flow in the final cap tube proportionally).

However if you are shooting for a comfortable -140C @300 watts, be prepared to bump up the compressor to something approaching 4 horsepower, and requiring 3 phase power.

[AC_Hacker]

mytekcontrols,

What method did you follow when you made your tube-in-tube HX?

Do you have a way to estimate heat exchange?

Also, I tried to make one and I ran into trouble trying to insert the smaller tube into the larger... I ended straightening the tubes out, making the insertion and re-coiling.

Surely there's a better way.

Thanks,

-AC_Hacker

[mytekcontrols]

I ran into trouble trying to insert the smaller tube into the larger... I ended straightening the tubes out, making the insertion and re-coiling. Surely there's a better way.

Well other than using a prefabricated HXC such as a flat plate type, the DIY tube-in-tube coaxial is going to be a bit of a challenge to make. However with that said, there is at least one trick to make this process much easier. And that is to pay particular attention to keeping the tubing very straight when rolling it out (prior to doing the insertion of one tube into the other).

So my method involves:

1. Begin by anchoring one end of tubing, using either a heavy weight, or someone standing on it.
2. Take your time, and roll the tubing out very straight.
3. If the outer and inner pieces have a curve, then line them up together, prior to stuffing one into the other.
4. Finally, make absolutely sure that the outer tubing has been reamed out on both ends (using a tapered reamer), back to it's original ID (tubing cutters tend to crimp down the tubing). Making sure to do this step, will allow the inner tube to slide much easier into the outer one.

I can't stress this enough: KEEP THE TUBING STRAIGHT

Another trick, although much more involved, is to securely anchor one end of the tubing (clamp it to something very solid and immovable), and attach a hydraulic puller to the other end to both pull and slightly stretch the tubing. This will make the tubing extremely straight as a result (good method to implement for full production, just be absolutely sure that the ends of the tubing are clamped really good before stretching).

I hope that helps, and good luck on your project AC_Hacker

[n00b 0f l337]

Hey Michael! (Popped you an email a while ago)
I also found as a cheap method, is to use a square steel tube and a steel angle. By rolling out a tube against the inner curve of the angle, you can press it down flat with the square to really get it nice and straight. That helped me roll 40' of 3/16" into a 3/8" tube with minimal discomfort.

[mytekcontrols]

By rolling out a tube against the inner curve of the angle, you can press it down flat with the square to really get it nice and straight. That helped me roll 40' of 3/16" into a 3/8" tube with minimal discomfort.

Interesting technique. Kinda reminds me of something that a guy I work with showed me. In his case, he would lay the tubing on a very smooth cement floor, and then stand on the tubing while sliding his feet sideways across it back and forth. The end result was amazingly straight tubing, almost as if it had started out life as rigid tubing.

[n00b 0f l337]

Mind dropping me a message? I'm getting a bounce back from your old email.

[mytekcontrols]

I was looking through my test notes the other day using hydrocarbons on my AutoC and came across what I think was my most promising blend. This was tested in my last design iteration of the AC-3 unit, which was essentially a 4 stage AutoC (Aux Condenser, 3 Cascade Condensers, 3 Phase Separators), having a 6500 BTU/hr Rotary Compressor, and a 3"x9" expansion tank (with head pressure activated Buffer Valve set to 350 psig).

Charging was done via tank transfer, where an independent tank (5lb propane cylinder) is evacuated, and then the liquid is first added while weighing the tank with an electronic scale. This is then followed by adding the gases one-by-one based on their additive pressure, starting with the initial vapor pressure of the liquid alone (method was previously described in an earlier post to this thread).

CHARGE
R600 (Nbutane): 2.5 ounces (10-15 psig vapor pressure)
R1150 (ethylene): 90 psi
R14: 25 psi
Argon: 30 psi

Tank B.P. = 155-160 psig
Unit balance pressure (B.P.) after charged from tank and allowed to "soak-in" for 12 hours = 210-215 psig

Note: This unit had a suction cut-off hand valve, which allowed isolating the compressor from the heat exchanger stack during the charging process, thus allowing the compressor to "suck-in" and fully evacuate the charging tank, thereby transferring the entire charge from the tank into the unit. Also important; the liquid butane was first allowed to charge into the discharge circuit only by inverting the charging tank, and letting the gases push it into the unit (only the discharge side of a connected gauge set/charging manifold was open). Afterwards the connected charging tank and unit were allowed to balance out for several minutes, then the manifold's discharge valve was closed, the manifold's suction valve was opened, the unit's suction cut-off hand valve was closed, and the compressor was started and allowed to run until the suction gauge no longer dropped in pressure.

With this charge I saw a very rapid cool-down (+20C to -125C in 30 minutes @30 watt static load), and an ultimate temperature of -140C within 2 hours. Only buffered 4 times within the first 15 minutes. For higher loads I would recommend reducing the Argon and boosting the R14.

5 lb Propane cylinder used as Charge Transfer Tank (with suitable adapter)

[n00b 0f l337]

So it works
Now to get a decent way to access butane instead of those damn pop top cans.
Any load figures yet?

[Sgrios]

Michael, what hardware and software did you use to monitor/data log all of your thermal couples on the AC-2 project?

It looks like a Measurement Computing, but I wasn't sure.

Regards

[n00b 0f l337]

And have you considered Methane to replace R14/Argon?

[mytekcontrols]

So it works

To be clear to all that may read this... These results are from nearly a year ago, and pertain to the 3rd prototype in the "AC" series units, a project that I have for the most part abandoned.

Any load figures yet?

No I never did set-up a way to actually load the evaporator on this unit, so the test results shown were just based on an estimated static heat load of approximately 25-30 watts (1/4"id x 2' Feed & Return flex lines w/1.5" thick insulation, solid machined copper vacuum insulated cold head --- see images below).

And have you considered Methane to replace R14/Argon?

Yes and no. But I would first have to get my hands on some, which in my case means spending some $$$. And on my next prototype (AC-4) I plan to go back to a non-flammable charge based predominately on HFC's.

Michael, what hardware and software did you use to monitor/data log all of your thermal couples on the AC-2 project?
It looks like a Measurement Computing, but I wasn't sure.

Sgrios you are correct on the hardware, and semi correct on the software. My brother wrote a chart recorder application in Delphi using the Measurement Computing supplied drivers, which was far superior to the TracerDAQ program that comes with the USB-TC hardware.

Can't beat it when it comes to a fairly inexpensive way to accurately monitor 8 thermocouples down into the cryogenic range.

Images of AC-3 unit's flex lines, cold head, and unit connection:

[Sgrios]

Sgrios you are correct on the hardware, and semi correct on the software. My brother wrote a chart recorder application in Delphi using the Measurement Computing supplied drivers, which was far superior to the TracerDAQ program that comes with the USB-TC hardware.

Can't beat it when it comes to a fairly inexpensive way to accurately monitor 8 thermocouples down into the cryogenic range.

Thank you very much, Michael.

[kang-China]

Ask a question:
1. The Ar critical temp is -122.4℃,how is about the R14 evaporation temperature
2.minor details Ar evaporation temperature

[johnksss]

To be clear to all that may read this... These results are from nearly a year ago, and pertain to the 3rd prototype in the "AC" series units, a project that I have for the most part abandoned.


No I never did set-up a way to actually load the evaporator on this unit, so the test results shown were just based on an estimated static heat load of approximately 25-30 watts (1/4"id x 2' Feed & Return flex lines w/1.5" thick insulation, solid machined copper vacuum insulated cold head --- see images below).


Yes and no. But I would first have to get my hands on some, which in my case means spending some $$$. And on my next prototype (AC-4) I plan to go back to a non-flammable charge based predominately on HFC's.


Sgrios you are correct on the hardware, and semi correct on the software. My brother wrote a chart recorder application in Delphi using the Measurement Computing supplied drivers, which was far superior to the TracerDAQ program that comes with the USB-TC hardware.

Can't beat it when it comes to a fairly inexpensive way to accurately monitor 8 thermocouples down into the cryogenic range.

Images of AC-3 unit's flex lines, cold head, and unit connection:

where did you get those flexlines at?

[mytekcontrols]

kang-China...

Ask a question:
1. The Ar critical temp is -122.4℃,how is about the R14 evaporation temperature
2.minor details Ar evaporation temperature

Yes the critical point for Argon (-122.4℃) is fairly close to the boiling point of R14 (-128℃), but I fail to see what your actual question is. It really seems that you are simply making a statement of facts.

But if I were to make a guess at what you want to know, I would think the question would be "How can the Argon be of any use in the AutoC, when the physical properties of such (critical temperature and boiling point) would certainly suggest that a condensate could not be formed under the pressures and temperatures that are present".

To try to answer this question, I will present you with a theory. The theory consists of two different aspects.

Aspect #1: The presence of Argon gas creates a partial pressure situation in the evaporator, which allows the R14 to boil at a lower temperature than if it were doing so in a pure R14 gas environment. This is similar to how the hydrogen in a propane heated absorption refrigerator works, exerting a partial pressure on the liquid Ammonia, and allowing it to evaporate.

Aspect #2: The condensed R14 absorbs some of the Argon gas, thereby creating a version of itself that has a lower boiling point that exists somewhere between pure Argon or pure R14.

I hope that answers the question.

johnksss...

where did you get those flexlines at?

Hosecraft USA

[kang-China]

in the AutoC, your AC-2 the refrigeration is
R123

R22 what's the evaporation temperature to be used condensation R170

R170 what's the evaporation temperature to be used condensation R14

R14 what's the evaporation temperature to be used condensation Ar

Argon what's the evaporation temperature to be used in the in the evaporator .if you will use the evaporator temperature is -150 or -160

In your AC-2 or AC-3 if the Ar is not enough ,the compressor suction is ?

in the end evaporator ,what's the refrigeration R14 or Ar to we used cool

[mytekcontrols]

kang-China...

in the AutoC, your AC-2 the refrigeration is R123

R22 what's the evaporation temperature to be used condensation R170

R170 what's the evaporation temperature to be used condensation R14

R14 what's the evaporation temperature to be used condensation Ar

Argon what's the evaporation temperature to be used in the in the evaporator .if you will use the evaporator temperature is -150 or -160

In your AC-2 or AC-3 if the Ar is not enough ,the compressor suction is ?

in the end evaporator ,what's the refrigeration R14 or Ar to we used cool

kang-China in order to answer your extensive list of questions, I would direct you to look at the individual refrigerant pressure/temperature chart for each one of the refrigerants in your inquiry. And then compare this to the running suction pressure of the AC-2 for evaporative data, and of course the running discharge pressure for condensation data.

Edit: And of course this would relate to the temperature data profile that I supplied in one of the earlier AC-2 posts.

Of course this will only give you an approximate temperature as it relates to each one of the refrigerants in question, since the blended refrigerant charge will act entirely different than if a single refrigerant were present. Basically there are just too many variables to calculate when going beyond a binary mixture. And I must admit that trying to calculate the exact nature of what is going on within an AutoC is also outside of my expertise (and I dare say, most anyone else).

Some of the things to keep in mind is that when the refigerants become sub-cooled, meaning to drop below the temperature point required for a 100% saturated liquid at a given pressure, other neighboring refrigerant gases will more readily go into solution with the sub-cooled liquid. This creates what on the surface appears to be an azeotropic blend, although in most cases it is zeotropic in nature (exhibiting a temperature glide when evaporated).

Trying to pin a number on any of this is simply out of the question for most people. However it is not impossible, since I did meet a man that could do this, but unfortunately he was never willing to share his methods, or more specifically his computer modeling program.

So I create and tune the charge for an AutoC based on my intuition gained from years of working with AutoC's, since I lack a specific tool to do this otherwise.

Edit: Since this involves extensive pattern recognition, it is not something easily taught to others who lack extensive AutoC experience.