Submarine stuff... Submersible one way release valve

[single]

I am working on an autonomous submarine for a senior capstone project at school. Basically, I made a large pressure vessel that houses our computer systems as well as our pneumatics system. It is a polycarbonate tube that is glued to two aluminum flanges with O-ring seals and outer plates that clamp down on the rings to create a seal.

This vessel does not experience very significant pressures as it only needs to be rated to about 15 feet of depth, and it has been tested to those limits so far multiple times.

The trouble I am having is with our pneumatics system. We have a torpedo launching mechanism that is driven by a small paintball gun tank full of CO2. The C02 enters into the pressure vessel where it is input into a manifold and distributed electronically when we want our torpedos to fire.

I am concerned about a leak in the manifold or any connections that might cause pressure to build inside of the pressure vessel, and could potentially cause it to fail. To that end, I have been researching one way release valves.

We have procured a couple of samples from Halkey Roberts, and the one that is most interesting to us is this 1.5psi release valve.

In an ideal world, this valve would release pressure inside the sub once a 1.5psi pressure differential was found, and hopefully release and re-seat without letting water come in and potentially compromise our electric systems.

My thoughts on testing this were to create a small PVC pressure vessel, and install this valve as well as a Schroeder valve on the PVC ends. I would then take it into something like a bathtub and let it release pressure for a while, then take it apart and check for leaks. IT would be really hard to mimic what this thing would be doing at depth, unless I just take my bike pump into the swimming pool I guess.

Am I missing anything? Any advice is appreciated.

-g

And here is a picture of the sub, click to enlarge...

[single]

And here it is swimming around. This is before we put real graphics on it and stuff...

[otto parts]

Auvgeek to the white paging phone, please.

[shroom]

^^haha, no shit

[single]

Does AUV in auvgeek stand for autonomous underwater vehicle? If so, I really dropped the ball on not consulting him sooner.

[shroom]

that is correct, he's your guy.

[auvgeek]

Yup, that's me. I'm really more on the controls/robotics/navigation side of things and I'm super busy at work right now, but I'll to try take a look at your problem sometime in the next few days and see if I can't offer something useful.

Shooting from the hip (and I'm really pretty tired right now, so this may be completely retarded): I wonder whether you even need to be concerned about this...if your pressure vessel is made correctly, it should be able to withstand a certain amount of pressure, regardless of whether that pressure is internal or external (right?). Are you really compressing enough CO2 to worry about this issue? Did anyone do some analysis when the PV was made? Another option is to have the computer logic trigger an alarm as soon as a certain pressure is reached inside the PV, after which the vehicle would surface quickly.

You guys lookin' to compete in the AUVSI competition?

[Norseman]

G, I have some exposure to very similar issues. This is important stuff! Here are a few thoughts to follow Z's above:

1. What if you have a complete failure of your solenoid or a rupture that dumps all the pressure from the CO2 flask very quickly? Can all components of your pressure hull withstand this worst-case internal pressure with a comfy safety factor? If so, maybe you don't need a relief valve. If not, will the pressure relief valve have a high enough flow rate to vent the pressure before reaching a critical stress? Even if you find a relief valve that cracks at some desirable pressure delta, you gotta make sure it can vent that pressure quickly enough to save your ass.

1.5. Do you have any way to monitor internal pressure? If you decide not to use a relief valve, and you do have an internal pneumatics leak, how would you tell whether you were bringing a bomb to the surface?

2. I'd venture a guess that most pressure relief valves wouldn't let more than a couple of drops of water past the seal, if any. What would a couple drops do inside your hull? If you placed the relief valve at the bottom of the hull, any water that might enter would just stay in a small puddle, right?

3. If you take your bike pump into the swimming pool... you won't be pumping air any more.


Last week of classes, yes? Let us know if you have more questions.

-S

[single]

Yup, that's me. I'm really more on the controls/robotics/navigation side of things and I'm super busy at work right now, but I'll to try take a look at your problem sometime in the next few days and see if I can't offer something useful.

Shooting from the hip (and I'm really pretty tired right now, so this may be completely retarded): I wonder whether you even need to be concerned about this...if your pressure vessel is made correctly, it should be able to withstand a certain amount of pressure, regardless of whether that pressure is internal or external (right?). Are you really compressing enough CO2 to worry about this issue? Did anyone do some analysis when the PV was made? Another option is to have the computer logic trigger an alarm as soon as a certain pressure is reached inside the PV, after which the vehicle would surface quickly.

You guys lookin' to compete in the AUVSI competition?

I will definitely look into the amount of pressure that the CO2 cartridge would input into the system. I think it might take us into a critical zone, as it is a pretty large tank. We did not do any analysis into the internal pressure of the PV, as when it comes down to it, we really didn't think it through thoroughly. I think we were also planning on non-pneumatic weapons at the time. The quick surfacing is a great idea though.

We are indeed headed to the AUVSI comp, the rest of my group is anyway. We are a student team sponsored by NAVSEA through Montana State.

I really appreciate the input, I also am going to send you a PM in a minute here.

1. What if you have a complete failure of your solenoid or a rupture that dumps all the pressure from the CO2 flask very quickly? Can all components of your pressure hull withstand this worst-case internal pressure with a comfy safety factor? If so, maybe you don't need a relief valve. If not, will the pressure relief valve have a high enough flow rate to vent the pressure before reaching a critical stress? Even if you find a relief valve that cracks at some desirable pressure delta, you gotta make sure it can vent that pressure quickly enough to save your ass.

Some really good points there.

1.5. Do you have any way to monitor internal pressure? If you decide not to use a relief valve, and you do have an internal pneumatics leak, how would you tell whether you were bringing a bomb to the surface?

I do believe we have an internal pressure sensor, I do not believe it is currently in implementation. Would be worth looking into for sure!

2. I'd venture a guess that most pressure relief valves wouldn't let more than a couple of drops of water past the seal, if any. What would a couple drops do inside your hull? If you placed the relief valve at the bottom of the hull, any water that might enter would just stay in a small puddle, right?

Yeah, and we could even catch that water potentially in a little trap. Good call.

3. If you take your bike pump into the swimming pool... you won't be pumping air any more.

Last week of classes, yes? Let us know if you have more questions.

-S

I could totally redeem myself here and say I was planning on taking one of those CO2 cartridge pumps into the water with me, but I wasnt.

Last week, almost done! I really appreciate the advice Scott! I will keep you posted with the goings on.

[single]

here is a video of our sub the first or second day of pool testing...

[jondrums]

At only 15ft depth, I'm guessing you could even use an umbrella valve. As mentioned before, it might leak a few drops, but so what.

[angrysasquatch]

Paintball background here...

The CO2 tank will give you about 800 psi. There are never really any massive failures, such that you need to worry about your relief valve not keeping up. Use steel braid hose, that stuff is really legit, especially with only 800psi. The quick couple plastic hose can sometimes fail, I'd avoid it if you can. After you've made all the pneumatic connections, throw the whole pneumatic part of it underwater if you can. If it's leaky, it will leak from the start, I've never seen a leak just start when something hasn't been dicked with.

IMO, don't worry too much about it re-seating, worst case it lets in a few mL of water, the electronics aren't directly on the floor anyways. Basically, in a worst case scenario of shit leaking, you have 0% chance of blowing it up, and maybe a 1% chance of something getting wetter than it should.

[single]

Thanks so much for the input, I will pass it on to my pneumatics guy.

Jon, we already have the valves, as the company that makes them hooked us up. We made a small, airtight vessel out of PVC fittings and did the release valve on one side with a Schroeder on the other. Put the release underwater and connected it to a pretty big air compressor. The thing can dump air like crazy, and didn't appear to have any problem seating. Didnt find any evidence if water getting in.

[Norseman]

Word. Sounds like you've got it handled.

When is your competition?

Is your NAVSEA sponsor out here at Keyport?

[single]

Competition is in July down in San Diego. I won't make it, but it looks like we have a pretty good representation. And yeah, we are sponsored by Navsea keyport, working with an MSU alum. I'm curious to see what happens to next years project with budget cuts and stuff.

How's the late season out west dude?

[auvgeek]

Keep the questions flowing if you guys need help. The large ornery biped seems to have solved this problem pretty well tho. Steel braided hose + testing underwater to see if there's a leak is an excellent idea.

Now I'm curious about your approach to navigation and thruster allocation logic...