-80¦C refrigeration, single stage in -20¦C ambient?

[JMOwen]

Hi,

Thanks for taking the time read this post...

Does anyone know whether there are compressors and refrigerants available that would transfer ~200W at -80°C evaporating, -25°C condensing?

The unit would be located in a -20°C room, hence the low condensing temperature.

I'd like to avoid a cascade system if possible, 55°C delta accross the compressor does not seem too extreme based on my experience of developing R404A based chillers but I'm not aware of any commercailly available units that will provide this performance.

 

[MintJulep]

This is essentially the second stage of a cascade system, right?

 

[JMOwen]

Yes.

What I'm looking for is a compressor / refrigerant combination that would meet the -80 evap, -25 condensing condition. I've spoken with a few of the big manufacturers here in europe and they don't offer anything below -40 evap.

 

[MintJulep]

Have you considered liquid Nitrogen?

 

[tesfa]

I donot think so.

Is there any environment with -20 degree centigrade ambient tempreture.

I mean the manufactureres do not have the motive that refrigeration unit.

 

[tesfa]

I donot think so.

Is there any environment with -20 degree centigrade ambient tempreture.

I mean the manufactureres do not have the motive to produce that refrigeration unit.

 

[pennpoint]

JMOwen
Can you guarantee holding the -20C condition with the heat of rejection from the R-508B system.
"You have 200W +/-plus heat of compression."
I would think the system would be far more robust if you built a cascade system.
Most cascade system are built by non-commercial, industrial intities.
I do not understand your last statment re: 55C across the compressor and R-404A. How does that relate to -80 evap.

Regards

pennpoint

 

[JMOwen]

Thanks for all your responses:

To clarify a few points:

I can guarantee holding the -20 condition as the unit will be located in a room chilled by with multiple redundant -20 cooling systems.

The condenser could be air cooled by the -20 air resulting in a condensing temperature ~ -25°C.

The 55°C delta and R404A comment was meant as an example. The temperature difference is simular to a -80 evap / -25 condensing system although I appreciate that the pressures may be very different depending upon the refrigerant.

So I'm looking for a single stage system at -80 evap / -25 condensing. (the primary stage of a commercial cascade unit?)

Perhaps the question would be better put as follows,

Which compressors and refrigerants are typically used in small to medium sized cascade systems designed for -80°C cooling? What are the typical evaporating and condensing temperatures accross the cascade?


Thanks again!

 

[JMOwen]

MintJulep, LN2 would be my first choice but it this instance it is impossible to supply it (or any other fluid) to the process to area to be cooled.

A small local mechanical system is the only practical solution I think. Even considered peltiers (for about a nanosecond! - think 200W and ~80 dT across the junctions)

 

[pennpoint]

Ok! Maybe this could work if you keep the low temperature environment (-80F) totally isolated from the -20F environment.
Be sure of your heat load.
I alluded to using R-508B, which is preferred to R-13 or R-23. (Reference ASHREA Fundamentals Handbook)
You need to use synthetic oil (verify).
Check on you’re location regarding cost of R-508B. It’s not cheap $$$$.
When the system is off the balance pressure can approach 400psi account for that by providing ample expansion space in your system. Then again most expansion tanks are sized for 70-75 F ambient. You still will need an expansion tank and when? –20F or at 70F ?
A Scenario: What if your –20F system fails, your system will go out on high head pressure, and then (temperature/pressure) up to what ever the ambient temperature is at that time. If that’s beyond the relief valve pressure setting you will loose refrigerant.
I still think you should consider a cascade system as these can be more robust.
And not add load to existing systems.

Wishing You Well
pennpoint

 

[sterl]

200 Watts? for minus 85-C. ET? With the equipment in a min 20-deg. C. Space? There are plenty of fluids and lubricants that will do that, but the equipment engineering is going to be pricey...

The quick means would be to modify something made for a materials lab Neslab makes an "Ultralow-Temperature" to -90 C. but I can't tell you how much cooling it will make at min 90.

 

[IRstuff]

Here's something that will DEFINITELY get there:

http://www.qdrive.com/index.php?page_id=17

check out the 4.2kW input power required. Probably will be A LOT less if you're only going to 193K. You'll probably need to get the temperature setpoint option.

TTFN

 

[pennpoint]

IRstuff
You are right on, with the low temperature capabilities of these exotic types of single stage systems. I am familiar with them and overlooked them for sure. Small capacity at high cost but they could do the job as a single stage. An exotic cocktail of refrigerants are used to achieve these low temps.
With this or any ultra low refrigeration system the cooled space must be kept sealed of moisture intrusion. Using a positive pressure of dry nitrogen gas purging of the cooled space will ensure reduced frosting/icing of the coil or cold finger.
JMOwen, let us know which way you go wuth this, ok?

Best Regards
pennpoint

 

[JMOwen]

Thanks Sterl, IRstuff and Pennpoint for your latest posts.

I liked the cryocooler - looks like an interesting technology, but It would be impossible to integrate and certainly more expensive that a single stage or even cascade conventional mechanical system.

The area I am trying to cool is located on a robot arm in a resricted area into which human access is undesireable.

This is the main reason why I do not consider LN2 to be an appropriate solution as I cannot see a way of delivering the liquid from a remote vessel. Locating a vessel on the robot is not an option due to space and payload contraints and filling issues.

An ideal solution would be a small single stage / cascade unit mounted on the robot adjacent to the area to be chilled. I have had good experience of small ~200 to 300W danfoss condensing units for which I've designed the evaporators and fitted expansion valves to cool a variety of applications, however being R404a the evaporatng temp is limited to ~ -40°C.

Given the -20 ambient I'm wondering whether the second stage of a cascade system similar to that found in a -80 lab freezer would do the job. The interstage could be replaced by a air cooled condenser.

The simplest solution, it appears to me, would be to take one of the danfoss condensing units and replace the oil and refrigerant to achieve a lower evaporating temperature. I have spoken to danfoss about this - it is done but not by themselves and they will not endorse any modifications as the units are only approved for R404a.

I have no practical experience of cascade systems (yet)and would be looking for a partner to develop the unit

 

[pennpoint]

JMOwen

I am wondering if you could use vacuum jacketed hose to and from your point of use. Use LN2 and exhaust it out of the building. Or use flexible refrigeration hoses to and from your evaporator back to the compressor/s. Let the flex hose move with the robot.

View this site for vacuum jacketed hose

http://www.vbsflex.com/index.html

How do you achieve the -20 room temperature at this point?

Cascade refrigeration systems of your needs are fairly easy.


pennpoint

 

[JMOwen]

pennpoint,

I'm familiar with vacuum insulated hose and but it is not appropriate for this application because of the design of the robot and concerns over its longevity (it could be flexed up to 3 million times over its life). -20 will be acheived using multiple dx refrigeration systems. I'm unable to describe the application in much more detail because of confidentiality issues but would appreciate more information on the development / specification / supply of cascade systems.

The unit should be as small as possible (less than 0.3m high x 0.5m wide x 0.5m deep and idealy weigh less than 15kg.

Thanks again for your continued help

JMOwen

 

[pennpoint]

JMOwen
Where are you located?

pennpoint

 

[JMOwen]

I'm in the UK

 

[pennpoint]

JMOwen
Ok! that explains your reference to Danfoss.
I'd like to give it a go and help you but it's a little difficult through this forum.

Tell me what you'd like to do?

pennpoint

 

[mysterylectricity]

I don't think those cryocoolers use a mix of exotic refrigerants, they use medium-pressure helium (~300 psi) always in the gas phase. This is rapidly compressed and relaxed (60 hz) with about a %25 pressure swing. Some designs use a moving part in the cold head, but some don't (pulse tubes). Even the ones with moving parts are generally wear free, for some reason the displacer centers itself in its bore and although the tolerances are fairly tight, deos not contact the tube wall during operation. The cold head is designed with metal screens and free volume spaces arranged in such a way that at the operating frequency, a phase "mismatch" is introduced which I think is between pressure and flow, or perhaps pressure and volume. Although the device is strictly linear(classicly: new designs use U tubes and/or have ports at the top end that feed back to the bottom), somehow the gas is tricked into expanding in the far (cold) end and compressing/dumping heat into the near (hot) end almost continuosly. I don't think I've ever fully understood them, or if I did I've forgotten the physics. This technology is virtualy identical to that use in cryopumps for high vacuum. Anyone interested in experimenting with this technology should be able to purchase a used cryopump on e-bay for ~$250. However, whereas the Qdrive system seem to use voice coils and other "acoustic" engines mated directly to the pulse tubes, cryopumps use an external compressor. The compressor supplies helium, but this helium is not circulated through the cold head. Instead it is used to drive a "scotch yolk" mechanism attached to the cold head, which provides the mechanical "accoustic" energy. Nevertheless high purity helium is required: else water and ice may form in the scotch yolk. I've heard that traditional refrigerator compressors are used, the important thing being to follow the compressor with a large activated carbon filter bed to trap compressor oil. The traps must be periodically replaced (dunno how often) and they're quite a bit pricier than they should be.

-Jeff