Temperature transfer to Stainless Steel Surface? 2

[bbnc1]

I tried reading all the threads to see if there is already a topic on this but couldn't find any. I have an application where one end of SS container is attached to Liquid Nitrogen source and the other is open-end to accommodate for air expansion when LN2 is filling the enclosed container. My question is what will the surface temperature of the SS container be if not insulated (i.e. -300F?) and the LN2 is filled. Does the thickness of the container matters? What kind of welding is needed to make sure the cold stress doesn't cause leak? Will there be condensation inside the container? I am not in this field, so I don't where to start the research. Please help. Thanks!

 

[IRstuff]

Why in the world would you use a non-insulated container for LN2?

I suggest that you get some help before someone gets hurt:

http://www.cryogenicsociety.org

TTFN

Eng-Tips Policies FAQ731-376

 

[bbnc1]

This is a hypothetical example similar to the application that I am reseaching without revealing too much. I know this is dangerous just like a stove top would be. You can guess that the surface is use as a cooling mechanism, and no human will touch it (wasn't bvious). Asking questions on this website is trying to get the data I needed from experience people before I attempt anything "stupid". If you know that answer, please help me with the data, so I can design the application appropriately and safely. For example, if the answer is that the surface will reach
the temperature of the LN2, then I know that I will need to add "layers" of material to bring the temperature to what I would need. Anyway, appreciate any data anyone can provide.

 

[IRstuff]

You should do the heat flow equation for your material.

For the simple case of LN2 on one side of a 1-inch thick stainless steel and ambient air on the other, I get about 80 K for the temperature of the steel.

TTFN

Eng-Tips Policies FAQ731-376

 

[iainuts]

My question is what will the surface temperature of the SS container be if not insulated (i.e. -300F?) and the LN2 is filled.

How accurately do you need this? Generally, the thermal conductivity of stainless is so high in comparison to the thermal convective heat transfer, that the steel is generally considered at liquid temperature. That's not true of course. To do this properly you need to consider the convective heat transfer between LN2 and stainless, the thermal conductive heat transfer across the stainless surface and then the convective heat transfer from the surface to atmosphere just as IRstuff mentions.

This analysis can be shortened a bit by assuming the convective heat transfer between the boiling LN2 and stainless wall is infinitely high (ie: no dT). You can then base the convective heat transfer of air on a shortcut provided in "Cryogenic Systems" by Randall Barron. For this, he gives a typical value of 0.25 Btu/ft2-s for LN2, LO2, LAr. This assumes relativly still air. For the thermal conductivity of stainless, I'd suggest checking with NIST, here:

http://cryogenics.nist.gov/NewFiles/material_properties.html

Note at LN2 temp, thermal conductivity is roughly cut in half.

What kind of welding is needed to make sure the cold stress doesn't cause leak?

Generally, welding is not considered a problem area. Thermal contraction of the steel may produce stresses, but it's very hard to say much without a drawing. In general, a conventional tank that just sits on the ground will not be adversely affected in any way by the temperature. 304 is ductile even at this temperature, so there's much less of a problem than with other steels. You've got the best steel available for this application.

Will there be condensation inside the container?

A little bit inside, mostly outside. The outside of the container will be covered with frost, and that frost can build, creating a large ice ball around it. Note that as it does, it will change the thermal conductivity, and in general it adds a very poor layer of 'insulation'. If the vessel is completely open at the top, you'll get a little inside as frost falls off the cold portion of the tank above the LN2 and into the liquid. If there's a small opening in the top of your tank, or a back pressure regulator to maintain a blanket of nitrogen, that will provide a nitrogen blanket above the liquid and prevent any kind of condensation inside the tank.

 

[bbnc1]

Thanks so much IRstuff and iainuts. The info you provided is very helpful. I'll check out the links and do more detailed research. The parameter I was most interested in was the surface temperature of the SS filled container (w/ pressure vent for boiling of LN2 during filling), And it looks like SS has good thermal conductivity.

Last question: Once the determine the feasibility of the application and want to proceed, do you think a welding company can build this customize container or go to a cryogenic specialist company to get it done?

 

[iainuts]

Hi bbnc1. If you're in the US and the container is going to be pressurized above 15 psig, it has to be an ASME code stamped vessel. If it's an open container, such as could be fabricated from a length of pipe and a single pipe cap, anyone can fabricate it. There may be a bit of a gray line between the two, if you have any kind of open hole or vent on the tank, you may need to consider who makes it more carefully.

Please also note that the boiling LN2 is an effective asphixiant, so operating this indoors is a big problem. Precipitation outdoors makes this problematic too. Finding a good home for this tank is an important safety issue.

 

[bbnc1]

The application is indoor and I am aware of the 700/1 expansion, so the gas will be vented outdoor. You mention precipitation outdoors is also problematic, so I am stuck again . I thought about putting a collector on the vented side, but then I will have do deal with pressure difference between the source and the collector; thus, open vent was the simplest design I could think off. I'll try to figure out how to attach a diagram.

 

[IRstuff]

I guess no one has asked why you need a custom container, when you can buy a commercial insulated LN2 container:

http://www.cryofab.com/html/cl_clpb.htm

We've probably got a dozen of the 80 liter jobs standing around in the lab. Since they're insulated and sealed, there's no worries about asphyxiation. They're directly filled from our external LN2 tank.

TTFN

Eng-Tips Policies FAQ731-376

 

[bbnc1]

IRstuff, Thanks for the link to the containers. I have a diagram drawn, but don't know how to attach to this thread (I searched for instructions but not avail). Anyway, I'll try to describe it better. The LN2 source is connected to a pipe (say 1"), this 1" pipe is then attached to another odd shape pipe (rectangular in this case), and the other side of the rectangular pipe is then connected back to a 1" pipe to feed outdoor or to a LN2 container. The rectangular "pipe" will be used to cool a "warm" agent attached to the surface. Here is the process: When the valve is turned on to fill the pipes, LN2 will boil as it traverse through the 1" pipe,then rectangular pipe, then through the output 1" pipe. The gas volume spike will force air out of the open end as the LN2 begin to fill all the pipes. Once filled, the LN2 will flow out of output 1" pipe into a container. If I figure out how to attach the drawing, I'll do that. Thanks for your help.

 

[iainuts]

Hi bbnc,
Am I correct then to assume the 'container' is just a receptacle for LN2 that has gone through the system and is essentially being discarded? Is the container only there to protect the surroundings from the spray of liquid nitrogen? If so, it sounds like a perfectly valid method to me, go ahead and make an open ended container. There's no real concern about water getting into it, just put a drain valve on the bottom if there's a chance of it filling with rain.

The only real concern is a safety issue of who might be walking by when this operation is being performed. You're in the best position to evaluate that hazard though.

 

[bbnc1]

Hi iainuts,

Your assumptions are almost correct. The "container" in this case has little volume since I am only interested in the heat transfer (cold transfer)and not as containment of LN2. Because I am interested in super-cooling the rectangular box, LN2 will have to flow constantly to make contact with the inner surfaces of the box so that the external surface will be close to LN2 temperature. The excess flow will be discarded. Think of it as a thin box laying on its side with pipe going into 'in' port and pipe go hook to out port. The LN2 flows into the 'in' pipe, fills the very thin box then flow out on the 'out' pipe. Thin box because I don't want to waste LN2. The out port is open ended. Is your opinion still the same that you think it is doable? You mentioned precipitation outdoor is problematic, but I am not sure why (asphixiation?).

 

[IRstuff]

Your application is still unclear. Why aren't you usnig a simply pour-fill Dewar? Your LN2 flow doesn't really seem to be doing much, other than wasting a lot of LN2.

Or better yet, why not use a electromechanical Stirling cooler and eliminate the waste altogether?

Perhaps you should explain the overall problem, rather than the solution to your immediate perceived problem. What you've described doesn't sound like anything anyone else involved in LN2 usage does.

TTFN

Eng-Tips Policies FAQ731-376

 

[bbnc1]

IRstuff, the high level requirement is that I need a cooling source size 1 ft x 2 ft of about -300F +/- 30F, so this eliminated all the cooler I am familiar with. My application is not in crygenic area, thus, seem odd to you, but because of LN2, I seek for help in this forum to make sure I achieve the -300F surface temperature but do it safely of course. The 'container' in this case is 1'x 2'x ?' where ?'= enough LN2 to achieve the temperature I need. Maybe I don't need an open end but a pressure relief valve during the boiling when LN2 entering the 'container'. Again, thanks for all the inputs you provided.

 

[IRstuff]

despite your protestation, -300ºF is a cryogenic temperature

Again, you fail to describe your requirements, i.e., what is your heat load?

This is a 200W cooler:

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

The only addition you need to add is a heat spreader

TTFN

Eng-Tips Policies FAQ731-376

 

[iainuts]

Is your opinion still the same that you think it is doable?

I see what you're doing now - and yes I think it's very doable.

You mentioned precipitation outdoor is problematic, but I am not sure why (asphixiation?).

Nevermind. I had some other idea in mind when you explained your system. I think what you have is perfectly ok.

Regarding the idea of using a cryocooler, I see a few problems with that. Cryocoolers don't typically have the ability to get anything too cold that isn't well insulated. The ones I've seen are used inside a VJ box or container of some sort to minimize heat transfer from the enviroment. In comparison, the use of LN2 pouring through a pipe or rectangular box will provide orders of magnitude more refrigeration than a cryocooler ever could. Besides which, cryocoolers are relativly expensive compared to renting a LN2 dewar and just using the product you need from it.

 

[bbnc1]

I really appreciate both IRstuff and iainuts for your inputs. Believe me when I say that as soon as I reveal what I use it for, it would be very obvious since there has been prior art in this area. Anyway, the solution has to be cheap and of course, safe.

iainuts: If I take out the open end and replace it with a low pressure valve to "bleed" the air expansion during initial condition (boiling), would it still be safe? Safe in this case means the 'container' not blowing up on someone face. I can have a welder make me a 1/2" thick SS 'container' and do some trial run. I can calculate the pressure generated based on volume, but conceptually, does it make sense to you?

 

[iainuts]

If I take out the open end and replace it with a low pressure valve to "bleed" the air expansion during initial condition (boiling), would it still be safe? … I can calculate the pressure generated based on volume, but conceptually, does it make sense to you?

Any portion of any system that can trap a cryogen must be evaluated for pressure created assuming worst case trapped fluid in that portion of the system. In other words, if you put a valve that can be shut or even pinched down significantly, you need to make the assumption that someone will leave the valve wide open, maximizing the amount of liquid nitrogen that may enter the system, and then shut the valve.

If you put a valve on the outlet of this system, you must design it for the worst case scenario regardless of who may or may not use the system. In this case, a thermal relief valve will be warranted and required. The relief valve must pass enough fluid to prevent the system from over pressurizing given the blocked in case. Since you don't even have a fluid system pressure, I'd suggest you not pursue using a valve on the outlet. Nevertheless, if there is some unforseen reason you need a valve on the outlet, you must size a relief device for the system that can pass enough product such that if someone were to block this system in, the thermal relief valve could handle the flow generated.

A second consideration is that a valve on the outlet of your container probably won't add any additional benefit. I'm assuming you can control the amount of LN2 into this container from a valve that blocks off the line feeding the container. You could go around in circles trying to determine the most efficient pressure and flow to operate this system at, but it seems to me you're using it for a lab experiment and not a large scale production system. In the case of a lab experiment, I'd suggest just going with an opening on the outlet of this container, and forgetting about putting a valve on it unless there is some very strong driving reason for it.

 

[IRstuff]

I suggest that you study the construction of commercially available pour-fill Dewars:

http://www.sbfp.com/Pour_Fill_Dewar.htm

http://www.triadcryosolutions.com/pour_fill.html

http://cryoindustries.com/mtd201b.pdf

You may need to fully review your heat load requirements. Boiling LN2 has terrible thermal conductivity

TTFN

Eng-Tips Policies FAQ731-376

 

[bbnc1]

iainuts, IRstuff: You have given me great help and advice. I will do further reading with the links provided and understand full consequences before I set out on the prototype. Great forum!