Managing liquid nitrogen flow for fiber laser

I have a fiber laser that uses significant flow rates of nitrogen (as much as 2,550 CFH at 200 psi).

We are using a liquid nitrogen cylinder, feeding liquid into an evaporator, which then feeds nitrogen gas to the machine. Here's a photo of our setup:


I am finding that it managing the liquid nitrogen flow to to match the machine's demand is a roller coaster.

Let's say that a part program requires 200 psi of nitrogen gas. To reliably push the liquid out of the cylinder into the evaporator requires that I boost the cylinder headspace pressure to around 300 psi. This generally requires that I open the pressure builder circuit.

Since the liquid doesn't really flow (in a significant amount) into the vaporizer until the machine is actually consuming the gas, when I first get started, I have no choice but to try and run the program with an insufficient supply of vaporized N2.

The machine alarms out a few times (and pauses) due to low N2 pressure. Eventually, enough liquid has flowed into the vaporizer and enough time has elapsed to allow it to become gas, and then the machine runs reliably so long as it is used continuously.

However, if we then pause running the machine for a period of time (if the operator allows it to sit after finishing a sheet or during lunch break), then the nitrogen gas pressure builds up too high in the liquid cylinder (or in the vaporizer), because the pressure builder circuit is still on, without the machine using the gas up. So the cylinder or vaporizer starts venting the excess pressure.

When this happens, we run over and close the pressure builder circuit and rush to try and get the next sheet loaded so the machine can use this excess gas.

Now with the pressure builder circuit closed, when we are using the machine again, the pressure in the cylinder headspace eventually drops too low to push the liquid into the evaporator.

So you see, it is a roller coaster: First not having enough gas pressure, then having too much pressure, and then back down again. Not good.

Is this the way it is for everyone using liquid N2 cylinders with high-powered fiber lasers? Or is something wrong with my setup? Any suggestions for how to improve the situation?

Dumb question - could you throw an accumulator tank or two into the path so they could be valved to take up excess flow for short periods, then return it to the laser feed? Would take a few solenoid valves and some robust controls (don't want this failing!), but seems reasonable?

Milland said:

Dumb question - could you throw an accumulator tank or two into the path so they could be valved to take up excess flow for short periods, then return it to the laser feed? Would take a few solenoid valves and some robust controls (don't want this failing!), but seems reasonable?

Click to expand...

Yes, I could see the benefit in having an accumulator tank between the vaporizer and the machine. Possibly no fancy valves or controls needed, other than an in-only / one-way check valve at the entrance to the accumulator tank. I could see how this could smooth out peaks and valleys of demand, giving the vaporizer time to do its work, but still we'd end up with the same problem: If the machine sits long enough with the pressure builder circuit open, we'd still eventually hit the pressure limit in the tank or evaporator, and it would still vent. And then we'd still need to close the pressure builder circuit again, eventually leading to a drop in pressure. An accumulator tank would lengthen the amount of time until these things happen, but these problems would still happen.

Maybe a better way to handle the situation is to have some external control to feed pressurized N2 gas INTO the gas VENT port (not the emergency vent) of the liquid cylinder as needed to create the desired head pressure, instead of using the pressure builder circuit to create the head pressure. Obviously there would be a bit of a chicken-and-egg scenario on how to get that initial pressure, but I could imagine how the pressure could be automatically monitored and maintained using pressure sensors and solenoid valves.

But I have to think that I'm not the only one with this problem, and there is probably some off-the-shelf solution that allows you to run the machine without constantly watching the pressure gauge.

Heck, just put a accumulator tank downstream (accumulating the gas, not the liquid).
it will stay at 250 psi or so all the time, when demand comes on it will drop, size it (and set pressure) so it doesn't drop below 200 psi.

pMetal said:

Maybe a better way to handle the situation is to have some external control to feed pressurized N2 gas INTO the gas VENT port (not the emergency vent) of the liquid cylinder as needed to create the desired head pressure, instead of using the pressure builder circuit to create the head pressure.

Click to expand...

As I understand it, there's two flavors of "regular" LN2 Dewar tanks, a low pressure version of ~22psi max, and a high pressure of around 230psi max. I would triple-check which ones you get, then talk to a proper engineer at your supplier about whether it's safe to operate a HP tank as you suggest. You don't want a F-up this sort of thing, it gets expensive...

I think you're looking at the wrong side of this. You already have what you need from a nitrogen supply perspective (evaporator, pressure source for the tank, etc)

It sounds like you need a bit more finesse on the other side. For example, change your programming at the machine to turn on nitrogen without the laser running at start of cycle for 15 seconds or 30 seconds or whatever it takes to be reliable. If you did that, and then interlocked your pressure build enable logic to the signal feeding your nitrogen valve on the machine, then it seems like all your issues would go away. Do you not have the ability to turn on the shielding gas independent of the laser?

DDoug said:

Heck, just put a accumulator tank downstream (accumulating the gas, not the liquid).
it will stay at 250 psi or so all the time, when demand comes on it will drop, size it (and set pressure) so it doesn't drop below 200 psi.

Click to expand...

Thanks, but this doesn't solve the problem, because the only way to fill that accumulator tank is to be pushing liquid out of the liquid cylinder, and that requires turning on the pressure builder circuit, so we're still at the same point.

Milland said:

As I understand it, there's two flavors of "regular" LN2 Dewar tanks, a low pressure version of ~22psi max, and a high pressure of around 230psi max. I would triple-check which ones you get, then talk to a proper engineer at your supplier about whether it's safe to operate a HP tank as you suggest. You don't want a F-up this sort of thing, it gets expensive...

Click to expand...

We have the 500psi version. The liquid cylinder's max pressure is not the problem.

Comatose said:

I think you're looking at the wrong side of this. You already have what you need from a nitrogen supply perspective (evaporator, pressure source for the tank, etc)

It sounds like you need a bit more finesse on the other side. For example, change your programming at the machine to turn on nitrogen without the laser running at start of cycle for 15 seconds or 30 seconds or whatever it takes to be reliable. If you did that, and then interlocked your pressure build enable logic to the signal feeding your nitrogen valve on the machine, then it seems like all your issues would go away. Do you not have the ability to turn on the shielding gas independent of the laser?

Click to expand...

Well you have a good idea for just blowing nitrogen at the start of the cycle so that I can get to a reliable situation prior to running the part. I can do that easily.

However, the pressure-build enable situation that you described is not an actual thing. The pressure builder circuit is a built-in part of the liquid cylinder with a manual valve to open/close. It is not the sort of thing you would ever to mess with either to connect to your own solenoid valve, because the cylinders belong to the gas dealers.

And even if you could, turning on/off the pressure builder circuit is not an instant-results thing. It takes time to react either turning on or off. Once opened, and the liquid gets into it and starts to evaporate, it is going to keep evaporating and building pressure until it is all vaporized. That process is going to happen even after you close the valve, until it is all vaporized.

This is why I think that a solution will require getting pressure into the system without using the cylinder's built-in pressure-builder circuit.

A lot of work and parts to save very little when compared to switching to skids of gas. We have been down this road with o2 on plasmas here. Skids turn out to save money, a lot of money in the summer, even if they technically cost more per cubic foot of gas.

pMetal said:

Thanks, but this doesn't solve the problem, because the only way to fill that accumulator tank is to be pushing liquid out of the liquid cylinder, and that requires turning on the pressure builder circuit, so we're still at the same point.

Click to expand...

O.K. then install one of these to use the lower pressure:

Call them and explain what flows & pressures you need, I think the air ampilfier series would do just fine, just switch them to air driven, nitrogen pumping:

sounds like you need a larger Dewar, enough to supply all it needs but far enough away to get it to turn to gas. a tank between the evaporator might solve this.

I wonder if the OP uses enough Ni to justify getting a nitrogen generator. Not sure of the payback time, but it entirely bypasses needing to bring in LN2 or bottle pallets.

Milland said:

I wonder if the OP uses enough Ni to justify getting a nitrogen generator. Not sure of the payback time, but it entirely bypasses needing to bring in LN2 or bottle pallets.

Click to expand...

Oh, hell yes. If a generator can get up to whatever purity requirements he has then that should be a slam dunk.

Could you use a regular high-pressure gaseous nitrogen tank and a regulator to pressurize the headspace of the liquid dewar to the desired pressure?

jmkasunich said:

Could you use a regular high-pressure gaseous nitrogen tank and a regulator to pressurize the headspace of the liquid dewar to the desired pressure?

Click to expand...

Yes - I spoke with the liquid cylinder Manufacturer and they say that will work and will help matters. However there will still be a bit of a roller-coaster effect from any liquid that is in the hose and in the vaporizer. When you pause consumption of the gas, that liquid is going to continue to vaporize and will potentially cause the same overpressure situation.

But by using a high-pressure N2 gas tank to pressurize the headspace, you do at least get rid of a good portion of the problem.

I have ordered such a tank to come with my next delivery and I'll report on the results.

I didn't read the whole thread but the pressure builders on those dewars reliably do not work. We connect the gas port of a dewar that is up to pressure to the vent port of the feed cylinder, which pushes liquid into the vaporizer. We consistently operate at 300 psi on 350psi dewars with no issues until we are completely out. If we do not have a dewar up to pressure we use a high pressure bottle regulated to just under 350 psi.
We drain a dewar that size in about 5 hours.
We also have a 5/8" ID hose from the vaporizer to the laser. There are some caveats to keep things from freezing, but it's easy.

EDIT: If we are going very heavy on nitrogen (6+ dewars in one delivery) we can chain dewars in a sequence which doubles that gas draw support so we don't have to stop mid sheet to change bottles once one runs out of liquid.

Strostkovy said:

I didn't read the whole thread but the pressure builders on those dewars reliably do not work. We connect the gas port of a dewar that is up to pressure to the vent port of the feed cylinder, which pushes liquid into the vaporizer. We consistently operate at 300 psi on 350psi dewars with no issues until we are completely out. If we do not have a dewar up to pressure we use a high pressure bottle regulated to just under 350 psi.
We drain a dewar that size in about 5 hours.
We also have a 5/8" ID hose from the vaporizer to the laser. There are some caveats to keep things from freezing, but it's easy.

EDIT: If we are going very heavy on nitrogen (6+ dewars in one delivery) we can chain dewars in a sequence which doubles that gas draw support so we don't have to stop mid sheet to change bottles once one runs out of liquid.

Click to expand...

Sounds like a good plan. Using an up-to-pressure liquid cylinder to push the in-use liquid cylinder is a smart improvement over only using a high-pressure gas cylinder to do the pushing. I'm just waiting for some regulators and custom hoses to arrive and I'll be up and running with this setup.

Do you have a solution for knowing when the liquid is totally empty? My concern is that when pushing with a high-pressure gas cylinder, once the liquid runs out, it will train the high-pressure cylinder within just a few minutes if we don't realize quick enough.

pMetal said:

Sounds like a good plan. Using an up-to-pressure liquid cylinder to push the in-use liquid cylinder is a smart improvement over only using a high-pressure gas cylinder to do the pushing. I'm just waiting for some regulators and custom hoses to arrive and I'll be up and running with this setup.

Do you have a solution for knowing when the liquid is totally empty? My concern is that when pushing with a high-pressure gas cylinder, once the liquid runs out, it will train the high-pressure cylinder within just a few minutes if we don't realize quick enough.

Click to expand...

The liquid hose between the dewar and vaporizer is corrugated, and whistles when gas flows through.

The gas hose that is pushing the headspace of the other dewar will also get condensation very quickly once it starts supplying all of the gas to the laser. If you let it go too long things will get very cold and brittle, and even longer you will get liquid from the gas port into the hose. We use a rubber hose, so we have it sleeved in stainless in case it breaks. Our first hose developed a pinhole leak after a couple of years with many freeze thaw cycles, and I found that even at liquid nitrogen temperatures it is extremely strong and difficult to snap.
EDIT: We originally had a needle valve in the pusher hose to restrict the flow so that once the pusher started trying to supply gas directly to the laser the pressure would drop and the machine would alarm. We found in practice we would rather just let the sheet finish and then swap dewars. It doesn't matter if your pusher loses pressure because it will be pushed by the next dewar up once you change them.

I now have the system set up with my high-pressure N2 gas "pusher" tank. What an improvement!
I now have my liquid cylinder headspace at a consistent 300psi, which is giving me a consistent supply of liquid into the evaporator, which in turn is giving me a consistent supply of 200psi gas to the laser. I put a flow restrictor (0.029" orifice) after the regulator from the high-pressure N2 gas "pusher" tank, which is fast enough to replace the lost liquid volume, but more than slow enough to ensure that the machine alarms out when we run out of liquid.

Well actually I have to wait until I get to the bottom of this liquid cylinder to verify that this will work as expected, but I have high hopes!

I'm planning to get another custom hose so I can connect a fresh liquid cylinder to an almost-empty one, however right now I don't really have room to put two liquid cylinders next to each other, so for now I think I'll stick with the small high-pressure gas cylinder.

Thanks for all your help!