CryoBUG is Back

[mytekcontrols]

Any reason for SS over copper on the cap tube?

Has to do with getting smaller ID then what is offered in copper. When you drop down below 0.026" ID, the only option I could find was SST as used for chromatography. Lab Express Management CO has quite a few selections to choose from, with the smallest being 0.005" ID.

Judging by what I saw when I disassembled a CryoTiger condensing unit, I will need a very small cap tube to do what I am after.

CRYOTIGER.jpg

[mytekcontrols]

CryoBUG-156C.jpg

Charge...
R-600a (Iso-Butane)
R-170 (Ethane)
R-14 (Tetrafluoromethane)
Argon

Yesterday's test results...
CryoBUG_7-21-12_Test.png

[MasterOfTheReal]

Beautiful work.

[n00b 0f l337]

What're your operating pressures like?
Interesting temperatures across the stack. And well done.

[mytekcontrols]

What're your operating pressures like?
Interesting temperatures across the stack. And well done.

Hi Adam -- check out the diagram and you'll see them above the compressor. They were very comfortable.

Thanks MasterOfTheReal.

Using the Iso-Butane instead of N-Butane solved the temperature/pressure oscillation problem I was seeing when I'd cross the -140C threshold. Now nothing except perhaps the oil can freeze, even as far down as -160C.

Now that I know what it takes to do this, next step is to do it again, only this time make it pretty.

Interesting thing I noticed when adding R-170 (ethane) to the R-600a (iso-butane), is that the R-600a readily absorbs the R-170 staying at or near 73 psig vapor pressure. You definitely have to use a scale when charging in order to get consistent repeatable results. Initially thought I had made a mistake when charging the R-170 since the vapor pressure was way the heck lower as compared to my earlier R-600 charge. But when I was testing I added a good amount more, and after the unit had been shut off and allowed to warm up, the static balance pressure was nearly identical. R-600a acts like a sponge for the R-170.

[n00b 0f l337]

Interesting. I missed the pressures, VERY reasonable.

As always, nicely done. The r600 behaves as expected then.

[mytekcontrols]

The r600 behaves as expected then.

Well actually the R-600a. And yes it does as far as the freezing point is concerned.

I knew from working with it's cousin R-600 (N-Butane), that their would be a tendency for the R-170 to soak-in. But I just wasn't expecting it to be quite so drastic with the R-600a (Iso-Butane). Because of this, I had to add about 30% more R-170 in order for full connection to be made to the colder stages. Apparently quite a bit of the R-170 gets separated out in the phase separator with the R-600a, much more so then when I was using R-600. They really have an affinity for each other.

It does have a positive side though. My R-600 charge balance pressure was 235 psig, and my new R-600a balance pressure is 207 psig, even with extra R-170 in the charge. It kinda reminds me of what happens when you blend R-123 with R-22, where the vapor pressure of the R-22 drops. So even though the vapor pressure of R-600 = 16 psig, and the vapor pressure of the R-600a is higher at 31 psig, you will still see a lower blended vapor pressure with R-600a/R-170 versus R-600/R-170 (this assumes equal charge weights in both cases).

[runmc]

Very nice job!! You are the "King of Autocascde"

[mytekcontrols]

Very nice job!! You are the "King of Autocascde"

Thanks Ron

Although I still have a ways to go before I consider this a done deal.

And I would like to say without the ingenuity and suggestions from many here at Xtreme Systems, I doubt that I would have been anywhere near as successful in this current endeavor. I have learned so much from other peoples projects and their solutions. Before I came to these forums, the thought of using hydrocarbons for refrigerants always scared the ST out of me. But after watching how they were being used, and listening to the precautions of their use, I started to make the transition. And this is just one of many examples of what I learned by being a part of these forums.

[dejanbinladen]

man that looks great i love it, not too big just wright

[mytekcontrols]

man that looks great i love it, not too big just wright

Thanks

Here's a sneak preview of some of the panels that will make up the enclosure for the pre-production demonstration unit..

Front Panel
CryoBug_Front_Panel.png

Side Panels (The patterns on these panels will be cut clear through forming air vents)
CryoBUG_Side_Panel.png

Top Panel (The patterns on this panel will be cut clear through forming air vents)
CryoBug_Top_Panel.png

This is a 3D view of how the panels integrate to form an enclosure.

9_enclosure_housing_profiles.png

It'll be made from black anodized aluminum, with the artwork engraved and infilled with paint. And this will all be fabricated by Front Panel Express. Although I had to do the hard part, that being creation of the design utilizing their free software.

The overall size of the unit will be similar to a small PC. The cold head will be sticking out the back, and basically just plug into the customer's vacuum system.

It wont be cheap, but it should look extremely professional.

[Patrickclouds]

great looking design of the case.
what are the white points in the compressor or condenser? LED's?

[mytekcontrols]

great looking design of the case.
what are the white points in the compressor or condenser? LED's?

Thank you.

Yes those are fault LED's. So if the air cooling fan fails, or the condenser needs to be cleaned, then the LED within the condenser symbol will light up after unit shut down. Similarly if there is either a low or high pressure cut-out, then the LED within the compressor symbol will be lit. And when either fault condition occurs, the momentary power switch's bi-color LED will flash red (normally green when unit is running). All faults can be cleared by pushing the power switch.

Bulgin MP1002/28/D1
Bulgin_Switch.png

Some of those other small white points are blind tapped holes on the reverse side for mounting the controller board.

All the panel screws will be countersunk black oxide coated.

BTW here is a picture of the folding lift handle that goes on the top...

Rohde KK-18.120.84
Handle.png

Edit: I just received this handle today, and I got to say what a nice piece of engineering. It is built to last!

[plasmatique]

[mytekcontrols]

After reviewing all my previous test reports (and there are plenty), I noticed that on my first hydrocarbon refrigerant charge test that I got a very rapid cool down. And the report also showed that I had achieved -156°C, but only because the refrigerant R-600 (N-Butane) had begun to freeze, thereby artificially reducing the ID of the final cap tube (which lowered the evaporator suction pressure). Problem is that it didn't stop at the optimum flow rate, but continued until the evaporator started to warm up. What I really liked when looking at this report, was that the evaporator temperature dropped very quickly following start-up. And although just a few days ago I once again achieved -156°C without any freeze up problems, it took quite sometime to get down there (just over 2.5 hours).

So what this early hydrocarbon test also showed me, was that I had been using a much shorter final cap tube than what I was using a few days ago. Well if you haven't guessed it, I went back to this same cap tube today, in fact I shortened it a bit more and then recharged the system with slightly more Argon this time around (to make up for the inevitably higher suction pressure).

So how did the test go?

It was like gangbusters, hitting -100°C in about 19 minutes and -150°C in about 50 minutes. Check out the chart below...

CryoBUG_7-25-12_Test.png

Note: If you forgot what the TC locations were, refer to this diagram: CryoBUG V5 Piping & TC Locations

Final running pressures were 38/227 psig.

Final evaporator temperatures were, Evap-In -153°C, Evap-Out -151.5°C (very small spread). Not quite as cold, but certainly appears to have good load handling capacity based on the very fast cool down.

The small variation you see in the temperatures (jagged plots) are the result of my test room's window AC unit kicking on and off by it's thermostatic control (not very precise temperature control I'm afraid).

[Patrickclouds]

really impressive your work. and hadn't thought that an autocascade can cool down so fast.

[n00b 0f l337]

Interesting numbers. Printed them all out and looking at some of your older charts side by side, the hydrocarbons are better than I thought. Maybe its getting back to there temp range and such, because they're masses and densitys aren't that high which is counter intuitive at times. But it is amazing as always to see your results, and I'm sure we all thank you as always for your rather clear explanations and secrets.

[mytekcontrols]

I think what it comes down to as well, is that to get a very simple autocascade such as this one (meaning so few stages of phase separation) to work well, requires a fairly large gap between refrigerant boiling points. And a fairly warm boiling point for the one charged in as a liquid. Now unfortunately as we have transitioned from CFC's, to HCFC's, and now to HFC's the freezing point of the warmer boiling refrigerants has gotten progressively worse. This is detrimental to any autocascade, but especially one such as I have built. Because try as you like, no matter how good you think you've designed your phase separators, inevitably warmer boiling refrigerants will get through to colder stages. And as they do, warm freezing points are not a good thing.

In the CFC days I would have used R-114, R-13, R-14, and Argon. Out of these 4 choices the warmest freezing point is -181°C. And with these refrigerants I could have easily built an autocascade good for -160°C or better, assuming that I had some decent oil separation after the compressor.

Then we get into HCFC's and the choices would have been R-123, R-22, R23, R-14, and Argon. Now we have the warmest freezing point is -107°C for the R-123. Not too bad considering that with it's very warm boiling point it'll get super subcooled and separated out fairly well in the 1st phase separator. But then the R-22 freezes at -146°C which could be a potential problem if it weren't for the great bond that forms between it and the R-123. And then we get to the R-23 (it's an HFC, but unfortunately their are no HCFC alternatives for the former R-13) which freezes at -159°C, which kinda limits us to running no lower than this temperature anywhere else in the system. So our ceiling probably realistically becomes -155°C to insure that we don't freeze the R-23.

Finally all we have left of the once glorious Freons are the HFC's and our choices are R-236fa (or R-245fa), R-125, R-23, R-14, and Argon. In this group we pretty much get screwed because all the liquids freeze at <= -103°C, and of course we still got the R-23 problem.

So other than for air conditioning or sub -150°C autocascades, HFC's really kinda suck.

However all is not lost, since due to recent EPA SNAP initiatives we can now legally use hydrocarbon based refrigerants in the US, and of course they have been legal even longer in Europe. R-600a (Iso-Butane) kinda takes us back to the CFC days, giving us a refrigerant with a warm boiling point and a cold freezing point, making what I am doing possible. Still not sub -160°C performance, but none the less far better then what can be done with HFC's. Although the domestic limitation of 57 grams is a bit of a challenge, it is still doable (currently I am only using a combined R-600a/R-170 amount of 52 grams). And if your target application is industrial, then you are allowed up to 150 grams. So theoretically I could produce an industrial charge for CryoBUG that had no Freons what-so-ever by using Iso-Butane, Ethane (or Ethylene), Methane, and Argon. And of course there is Krypton, but it's very rare, and hence very expensive.

[n00b 0f l337]

57 grams isn't much though, so that's a tough bargain.
To break the -160C point though, does it then become better in cascade usage? Such that you can either use an autocascading stage one, or stage two, and therefore avoid the freezing point issue of these warmer refrigerants?

[mytekcontrols]

57 grams isn't much though, so that's a tough bargain.
To break the -160C point though, does it then become better in cascade usage? Such that you can either use an autocascading stage one, or stage two, and therefore avoid the freezing point issue of these warmer refrigerants?

It is possible to get below -160 with the available legal to use refrigerants, but probably limited to a standard cascade as you pointed out, or an autocascade with multiple stages of phase separation but requiring primarily HC refrigerants. Although with careful planning HFC's could be utilized in the standard cascade design, or perhaps as you also suggested, a hybrid of both technologies.

And if you don't mind a short compressor life, something like Polycold's CryoTiger which is based on the joule thomson cryostat can be made to work with a nonflammable mixture.

Edit: I always think it's very sad that CFC's got banned entirely. They were, and still are great refrigerants. I think when used in well designed hermetically sealed (minimum or no couplings) systems, and utilizing proper refrigerant recovery procedures, CFC's could still be safely used. There has been much debate concerning the Ozone hole over Antarctica and how CFC's have or have not contributed to this. The interesting thing from my perspective, is that the hole grows and shrinks each year based on the season, and has probably done so for as far back as has been recorded.

[wdrzal]

Congradulations Mytek I think this is going to be a winner for you !


I always thought autocascades where allways only partially autocascading and were forced. By that I mean the gases weren't quit the right ones and the performance was forced by high discharge pressures and low suction pressures to get to the cryogenic temps.

Your current project show pressures that a compressor could have a nice long life at.


Great Work

Walt

[mytekcontrols]

Congradulations Mytek I think this is going to be a winner for you !

Thanks Walt. I hope so too.

Your current project show pressures that a compressor could have a nice long life

Yeah my feeling exactly. Kinda funny in way, since I suspect the life will be even longer then the compressor's intended target application, that being air conditioning.

I am rather anxious to get a heat loadable evap on this thing so that I can see what the load curve will look like. I suspect top end will be 50 watts before I run into serious high pressure issues.

[mytekcontrols]

The case design is finished, but I am doing a very careful review of how everything really fits together. This is going to be very expensive ($1,350) to have fabricated, so I don't want any surprises. To this end I have been exporting the individual pieces as 2D DXF (CAD) files and doing checks in DeltaCAD and in OpenSCAD to see if there are any misalignment problems.

Alignment check in DeltaCAD...

CryoBUG Assy.png

Some 3D Views from OpenSCAD (colors are for better clarity, all parts will be black anodized)...

CryoBUG3D(1).png
CryoBUG3D(2).png
CryoBUG3D(3).png


The front panel will sport two pressure gauges and a temperature display/controller (temperature control aspect not used, and instead serves as a set point to let the customer know when the unit is down to temp).

The inner red box is where the insulated HX Stack will be housed. The gold fan panel directly above it will sandwich the air cooled condenser between it and the top panel which in the final form will have air slots (as well as air slots on the main enclosure's side panels). Note: There will be additional stand-offs tying the fan panel to the top panel.

On the back panel of the enclosure you can see where the cold head flange attachment will take place. And on the bottom (last view) you can see the 3 point attachment for the rotary compressor.

The overall size and shape of this unit will be very similar to a medium size PC tower case. I had even entertained the idea of using such a case, but there would have been many compromises to make in order to make this work. By custom designing my own case, it insures that I get the best possible fit.

Things are getting real close on taking this into reality.

[n00b 0f l337]

That price is a bit ridiculous, you can almost buy a laser cutter for that... Have you concerned just using 8020 like before?

[Charles Wirth]

Mytekcontrols,

Lets trade some services or gas, I can cut the case accurate to .001"

I can cut < 1/2" aluminum plate and softer, I have not run steel nor do I have the bits for.

I have 3D MultiCam router
Kongsburg XL router
Checkmate King 120 laser cutter (cost a bit more than $1350 )

With DXF files on the flat panels it could be done in no time including printing of the face in color.