Sealed N2 Cylinder High Pressure 1

[karlb]

If LN is poured or syphoned into a strong cylinder under ambient pressure so that it is full then the cylinder is sealed and left to warm to room temperature, will the contents all turn to gas at high pressure?

Does the high pressure force any of the N to remain in liquid form?

Can someone describe what I need for a filling system that would allow me to perform this experiment?

I don't have any experience with LN and certainly do not have a chemistry background so your feedback would be greatly appreciated.
All I know is that it would be performed within a safety containment box/area of some sort.

Thanks and Cheers,

karlb

 

[iainuts]

The critical pressure of LIN is about 475 psig, so anything above that won't be liquid. When it warms to ambient it will be gas at an extremely high pressure (somewhere between 20,000 and 50,000 psi as I recall). To do the experiment, I'd suggest using a 100,000 psi rated cylinder and finding some safe place for it to explode.

 

[karlb]

Thanks for your advice.

Do you have any thoughts on how one would set up a simple gravity fed system to fill the cylinder? I am unfamiliar with LN liquid withdrawal devices and LN circulating systems.
I was thinking of a high pressure 3 way shut off valve somewhere close to the cylinder.

 

[iainuts]

Can you explain why you want to do this? What's the point? I can't in clear conscience, tell you how to do something like this if you have no experience and no obvious reason to perform this 'experiment'.

 

[karlb]

Hello iainuts, thank you for your time. I concur with your hesitation.

My primary field of work is in the realm of explosive physics R&D and largely in turn, the development of protective structures, systems, protection of frontline and first responders, protective garments and methods of maximizing survivability of energetic events.

I'm unable to disclose much but basically I am looking into these experimental trials and methods of setup not only to develp protection but also examine the non-desirable real world results if combined with othher available technologies...some of which have not been considered thoroughly in the past.

I am at the strawman concept level but still need to establish a fairly workable design before bringing in more LN experienced lab techs to get involved in the process. Basically, to asses feasibility with available funds.

Hope you could help point me a little further..

Thanks and Cheers,
kb

 

[iainuts]

I guess it depends on how exacting you want this experiment to be. Do you simply want to put any old LIN into the cylinder and make it explode? Or do you want to put LIN into your cylinder at a very specific physical state and measure properties?

Note that LIN for an experiment like this would be most easily obtained from one of the industrial gas companies or possibly a smaller gas distributor. LIN comes in various sized dewars (vacuum jacketed containers) that have integrated pressure build (PB) coils, delivery valves, relief devices, pressure gages, etc... They have a valve from which you can take liquid, and liquid can be 'pushed' out by adding pressure to the tank using the PB coil. A PB coil simply takes some of the liquid from the bottom, warms it up by rinning it through a tube welded to the ID of the outer jacket, then dumping that gas into the vapor space. Dewars up to about 500 psi can be obtained. Here's a link to a PDF file that shows you a basic layout of a typical dewar:

http://www.cryodepot.com/CRYO CYL LP 180-230.pdf

You would use a dewar, pressurize it, and use the liquid withdrawl port to fill your container, then valve off your container. But note that depending on how you fill it, the physical state of the fluid will vary. So if you're trying to get some kind of consistent result, you need to be able to control the state of the fluid before you valve off your receiving container. Is that important to you? Or are you simply trying to make something explode?

 

[karlb]

This is important. There would be plenty of instrumentation used to obtain data and once a baseline was established, having a repeatable trial is ideal. We would like to minimize variables and have control over a duration trials with incremental revisions of our choosing.

Thanks.

 

[iainuts]

Without knowing a bit more about what you're trying to accomplish, it makes it hard to make any specific recommendations. The initial state* of the fluid in your vessel and many other factors such as vessel geometry and material, will all affect how the subsequent fluid states evolve. In other words, you can put 100% liquid, saturated at some specific pressure inside the vessel (with some degree of accuracy) but any subsequent state will depend on what state the liquid is in.

The initial state can be carefully controlled, but depending on what state you decide to control to; the subsequent pressure & temperature rise as it warms up will vary. For example, if you control to 0 psig and saturated liquid, the final state may be ~48,000 psi. But if you control the initial state to 100 psig and saturated liquid, the final state may only be ~25,000 psi. To make things even more complex, any subsequent fluid state is not only a function of the initial fluid state but it is also a function of your system geometry and how your vessel, piping and other components 'stretch' as pressure and temperature increases, so any stretch will decrease the final pressure.

You can make your experiment very repeatable for any given test rig by simply controlling the initial fluid state but a different test rig may could give completely different results. This is especially obvious if you were to look at how even exceedingly small changes in density at any given temperature can produce huge changes in pressure. I have a couple different databases for LIN and they give significant differences in pressure for any given initial liquid state when going to ambient temperature.

What I need to understand is what you're really trying to accomplish, or at least, what initial fluid state you're trying to fix.

*Note: When I talk about the fluid state, I'm refering to the physical state of the fluid which is determined by two fluid variables such as pressure & temperature, density and internal energy, entropy and quality, etc... From these two variables, the remaining variables are fixed and can be determined. In your case, I believe you're looking to fix the initial state by controlling pressure and quality. Subsequent states can be determined by measuring pressure and temperature - however ... depending on what database you use, other properties can vary considerably because properties such as density at extremely high pressure are not known with sufficient accuracy.

 

[karlb]

The vessel geometry and material would essentially be identical each time +/-0.05mm. Size would only be on the scale of 1 to 2 inch pipe and 6 to 8 inches length machined from Inconel 625 solid bar stock. Geometry may be an acceptable variable but I hadn't thought about these other many variables.

As I mentioned, strictly in concept stage but I'd suppose we would want to fill at a controlled initial pressure and initial temperature of a liquid.... Perhaps a circulating system until the desired temperature of the vessel is achieved? Then it would need to be isolated close to the vessel and removed from the system. Possibly to sit in a controlled temperature environment?
If vessel expansion is neglected, should the theoretical final density not be same or similar to when in liquid form?

What variations in "quality" could be expected?

Thanks for your input.

 

[iainuts]

As LIN flows from dewar to your cylinder, there will be some heat which can cause boiling of the LIN. The flow is typically 2 phase (liquid and gas). The gas could get trapped if there is no way to eliminate it, just as hydraulic systems must be bled to ensure no gas is trapped.

When designing your system, the first thing is to reduce the amount of gas generated in your flow stream. The second consideration is to eliminate pockets or dead spots where gas could accumulate.

To reduce gas generation, the system can be insulated, either with a foam or fiberglass insulation or using vacuum jacketed lines. Also, having a flow through your system for 5 minutes or so will help to bring the temperature of your system to equilibrium and reduce the amount of heat generated.

The second consideration, eliminating dead spots, is simply a matter of having a design that allows gas to be carried through the system and out to vent.

You can also reduce or eliminate the amount of gas formation by capturing the liquid in your container when it is slightly subcooled. If your dewar for example, is kept at 0 psig for an extended period of time (ie: a few days), the saturation pressure should be very close to 0 psig. Do this by keeping a valve cracked open so it is constantly blowing nitrogen out as heat enters. Pressurizing the dewar then to 20 psig to transfer the LIN to your cylinder would give you some subcooling. You could then flow through your cylinder and out to vent. By keeping your cylinder at 10 psig, what will happen is this:
1. There will be a transient in which the lines and cylinder cool down. All of the fluid coming out the vent will be gas.
2. The transient will gradually shift so that more and more liquid begins to come through to the vent.
3. The transient will eventually disappear as pure liquid comes out of the vent. Note that this depends on heat ingress to your LIN. If there is sufficient heat ingress, you may never get pure liquid out.
- All of the above assumes your cylinder is at 10 psig, the dewar is saturated at 0 psig and then pressurized to 20 psig.

Note that the final state of your fluid in the cylinder using the above procedure would be determined by the total amount of heat ingress and the thermodynamic processes that the LIN undergoes. Since the LIN is saturated at 0 psig to start, it will follow a line of constant entropy as you pressurize it to 20 psig. When flowing, it follows a line dictated by the first law - the flow is an isenthalpic expansion but with the addition of heat being added to the liquid. If you eliminate all heat, the state of the fluid in your cylinder will have followed an isentropic compression from 0 to 20 psig, and an isenthalpic expansion to 10 psig. The fluid is therefore subcooled, so it will be 100% liquid. Any heat flux will push that state towards the saturation point and can cause it to boil and enter the 2 phase region.

Note that variations in density of this final state should be small if done as described above. The difference in density between the liquid given the above description with no heat flux and the liquid density at the saturation pressure of 10 psig is less than 1%. But as pressures increase, this variation in density will also increase. For example, starting at 0 psig, pressurizing the dewar to 200 psig, then transfering to the cylinder at 100 psig will result in a density that is 17% higher than the density assuming the liquid was heated to the point of saturation inside the cylinder at 100 psig. So although your cylinder may be at 100 psig in this case, the final density could vary by 17% which directly impacts the pressure the cylinder will go to when warmed to ambient.

The lesson here is that to reduce the inaccuracy of the initial density, fluid conditions during your transfer have to be carefully controlled and pressure variations minimized. To improve on this, I might also suggest having your cylinder in a bath of LIN to eliminate heat transfer, or possibly even reverse it (ie: take heat OUT of the LIN). So there are things you can do to more carefully control the final density, but without knowing how accurately you want to control density, and what density you really intend to have, it's difficult to provide more specific instructions.

 

[karlb]

I feel I've taken a crash course in part of the subject matter. Thank you very much for your efforts!

Will update you if things progress.