ASME BPVC VIII-1 UG-22(g) Submarine testing vessel 1

[FoxRox]

A submarine is to be enclosed inside a larger pressure vessel. The space inside the vessel and outside the submarine is filled with water, and a hydrotest is performed to test the sub.

If the submarine implodes, what loads affect outer vessel? I am thinking of something occurring like in the video below.

https://www.youtube.com/shorts/xM6zdim0yk4

I believe this would fall under UG-22(g), but I do not know how to predict or quantify this effect. Even if cavitation does not occur, I think the violence of the implosion is worthy of consideration.

And beyond that, I would appreciate any other thoughts on the loadings and other design considerations for this submarine testing vessel.

FWIW, (before I get rebuked), I am anticipating the involvement of a third party in this design, as it is beyond my expertise. At this point this is just preliminary research.

 

[MJCronin]

Great post ... and very interesting question .. This question is worthy of the venerable "eng-tips" GRAYBEARDS !!!

I am making some wild conjectures and assumptions here and offer the following:

1) RELATIVE VOLUMES:
I would imagine that the ratio of volumes (Volume test vessel/Volume submarine) ... lets say Vtv/Vs would be a crucially important ratio for design. When Vtv/Vs is close to one, implosion loads would be maximum, when Vtv/Vs is very large, then implosion loads would be ~ zero ...

2) ASSUMED IMPLOSION TIME:
I would try to develop some number for this. While I am well aware that an IT of near zero give maximum loadings (conservative but unrealistic) isn't there more realistic data out there and won't it make calculated loads more realistic

3) SIMILARITY TO NEARBY NAVAL EXPLOSION LOADS
I believe that a starting point would be review of explosion loadings (and stress analysis techniques) by underwater ordinance exploding nearby a ship. I feel certain that there was some kind empirical work done in the 1950s-1960s about this dynamic impact load

4) FULL VACUUM DESIGN PLUS A MARGIN
Because there is a verified, well known methodology existing in ASME VIII for "full vacuum" design, for preliminary/budgetary grade estimes I would start there... Design the test chamber for ASME VIII "Full Vacuum" .... then double the wall thickness

Tell us more about the submersable vehicles that you envison ....

I believe that a unique "implosion loading" may have to be developed for each and every "test chamber/test sub" combinatiion

What do you think ?


MJCronin
Sr. Process Engineer

 

[DriveMeNuts]

A sudden implosion of the inner vessel will be an energy release, and therefore exert energy into the outer chamber, and not so much exert stress.
I'm not sure how you would handle that with static PV design philosophy.
Sure, you could just add margin, but how much? And if the material is very brittle, adding stress margin will achieve little.
I would consider having material with a high impact fracture energy. But how high? The glass bottle in the video shattered because it has a very low impact energy.

 

[LittleInch]

The pressure in the test chamber in the event of an implosion wilful immediately reduce to near zero with the void in the sub being much bigger than any water volume being pumped into the chamber.

I'm struggling to see how there could be anything which causes a pressure spike in the test vessel because of this void.

The energy available to implode the sub is just not there as soon as the sub starts to crack or break the energy just vanishes. You don't have a whole ocean. You have a few ccs.

Hopefully it's not carbon fibre (cough)....

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[FoxRox]

1) RELATIVE VOLUMES:

Yes, I had thought of this, and I agree with you that a larger Vtv/Vs would mitigate the effects of the implosion. Unfortunately, it would also be much more costly. For reference, I am looking at a sub close to eight feet in diameter, and the test pressure needs to be over 700 psi, so you can see how increasing the diameter of the test vessel will become cost-prohibitive very quickly. There are examples of test units like this online, and in many cases, the sub barely fits in the test vessel, so I assume that it can be designed that way.

2) ASSUMED IMPLOSION TIME:

I believe the implosion will occur on the order of a few milliseconds. We are not talking about Titanic depths here, but even low pressure implosions occur very quickly do to the nature of buckling. That being said, I suspect every millisecond counts when you are trying to avoid an asymptote. There are research papers and data available online regarding these implosions, but I do not yet know if the data will translate well for use with ASME loadings.

3) SIMILARITY TO NEARBY NAVAL EXPLOSION LOADS

Yes, this could be useful for the pressure wave created by an implosion, but I don't imagine that it would help with cavitation.

And on the topic of pressure waves, say that I am able to quantify a pressure wave for a given implosion. Does the vessel need to be designed for the the peak pressure of that wave, even if it only occurs for milliseconds, or could it reasonably be designed for a lower pressure, since the impulse of such a wave might be very small even if the peak is high?

4) FULL VACUUM DESIGN PLUS A MARGIN

Yes, I agree full vacuum should be one of the design parameters. I think a sub implosion would undoubtedly cause something close to full vacuum in the test vessel, even if only for a short time. Given the high positive pressure I need to rate this vessel for, adding full vacuum to the design will not likely drive more thickness than I already need, even if I double the UG-28 and UG-33 minimums.

Tell us more about the submersable vehicles that you envison ....

I gave a few details in the response to number one. I may be able to answer a few other questions, but I cannot get into specifics. The subs themselves are not something I will be involved in designing or building.

 

[r6155]

I assume that the Acoustic Emissions Examination is used during the test (see ASME V).

Regards

 

[FoxRox]

LittleInch, that is a good point. A full implosion may not even be possible. Once buckling begins, gauge pressure in the test vessel should go to zero. The pressure wave caused at the end of an implosion may never even occur. You have me questioning if this is even a problem at all...

But what all happens in that moment? Given the rapid nature of buckling, will the sudden depressurization cause cavitation like in the YouTube video I posted? Will the gauge pressure just go to zero, or will it spike downward into vacuum before stabilizing?

 

[Compositepro]

The cavitation effects shown in the video are due to inertia of the water. There has to be a void space in the bottle for the water to move to in order to create a cavitation void at the bottom of the bottle. Water moving into the the cavitation void then hits the bottom of the bottle and breaks it due to the sudden stop.

In the case of a submarine in a chamber of water, as soon as the submarine begins to buckle, the chamber pressure will drop to zero because water has little compressibility. So there should be little crushing occurring after initial buckling.

 

[JStephen]

A related video:

https://youtu.be/tzl93q4tUWM?si=cIJVDBaV_8y9APc2

 

[LittleInch]

That second video is not the same. The amount of air in the bottle is very small compared to the test being discussed where the air is >50% and probably closer to 80% of the total volume of the test chamber and hence 2 to 10 times the amount of water being pressurised.
If you look a the pressure trace even then it goes almost to zero before climbing back as the pressure pump keeps going.

I don't think in the OPs case there is any problem here, but design the vessel for full vacuum.

The issue is more that the vessel won't implode in the same way it would under the ocean, but once it cracks or deforms the only thing stopping it imploding is the flow rate of the pressurising pump. That flowrate is infinity in the ocean, but much much smaller in the test chamber.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[DriveMeNuts]

And on the topic of pressure waves, say that I am able to quantify a pressure wave for a given implosion. Does the vessel need to be designed for the the peak pressure of that wave, even if it only occurs for milliseconds, or could it reasonably be designed for a lower pressure, since the impulse of such a wave might be very small even if the peak is high?

I think neither, because this is an energy problem, not a stress/pressure problem.

It is no different to releasing the water valve at the completion of a conventional Hydro test of a pressure vessel.
If a vessel is full of water with no air gap and pressurised to full test pressure, think of what happens when the water valve is opened. If it is slowly released, then the vessel deflates gradually, producing a very gradual release of energy. If the valve is popped open in 1 nano second, then there will be a sudden energy shock to the vessel. I am not aware of any regulation relating to releasing the pressure valve slowly. I guess that the sudden energy shock will never be enough to do damage the vessel, unless it is made of glass or Carbon fibre that is.

The vessel in the OP is no different. Accept, instead of the valve being suddenly opened to release the energy, it is the failure of the vessel being tested which releases the energy.

I don't know how you would go about quantifying the energy release, and assessing if it is a concern. I doubt that it is. As a precaution, use materials tested to a high impact energy. Having a small air gap at the top of the test vessel, may act as an energy release damper (i.e. the same amount of energy will be released, just more gradually).

 

[r6155]

It sounds like someone has no experience in hydrostatic pressure test.
The pressure test must be monitored to avoid a destructive test. See Acoustic Emission Examination (ASME V).

Regards

 

[LittleInch]

There should be no air in the test vessel as this is stored energy which could result in some shock loading if the submarine collapse

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[zeusfaber]

If you were going to mess about with small air pockets, the place to put one would be inside the submarine with the rest of the hull flooded - just big enough to ensure that the pressure hull, if it's sound, sees the required pressure differential throughout the test but small enough to limit the scale of the collapse if the structure fails. Hard to do in practice, especially if the boat has already been fitted out.

 

[r6155]

ASME VIII Div 1 for a submarine?: Ridiculous
Waste time.

Regards

 

[LittleInch]

R6155,

It's the testing vessel which is ASME VIII...

Not the submarine.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[r6155]

Yes, thanks Littleinch.
I insist: it is ridiculous to consider a submarine destroyed during a pressure test because the concept of what the pressure test is for is lost.
ASME VIII Division. 1 is very far from being applied for the required pressures, not mentioned by (OP).

Regards

 

[LittleInch]

The proof test idea though is well established. This is what the Triton sub didn't do.

How else are they supposed to test the sub to destruction?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[r6155]

@ LittleInch
"How else are they supposed to test the sub to destruction?"

Why destruction?

Suppose ASME VIII Div 1 is used, do you think UG-99 or UG-101 applies?

UG-101 PROOF TESTS TO ESTABLISH MAXIMUM
ALLOWABLE WORKING PRESSURE

Regards

 

[DriveMeNuts]

It sounds like someone has no experience in hydrostatic pressure test.
The pressure test must be monitored to avoid a destructive test. See Acoustic Emission Examination (ASME V).

Acoustic Emission Examination has nothing to do with external pressure tests, and is not mandatory for the testing of the test vessel. I don't understand why you bought it up at all. Are you sure that you have experience in hydrostatic pressure testing?

I insist: it is ridiculous to consider a submarine destroyed during a pressure test because the concept of what the pressure test is for is lost.
ASME VIII Division. 1 is very far from being applied for the required pressures, not mentioned by (OP).

What do you think that the purpose of the external pressure test is for? Just for fun? The external pressure test is not a destructive proof test, however vessels do implode. And consider that the margin for buckling failure in an external pressure test is much slimmer than for plastic collapse in an internal pressure test. Div 2 (and App 46 of Div 1) has buckling design margin at 1.67xMAEP, while external pressure test must be at least 1.25xMAEP for a PVHO-1 submersible. The concept and reason why the external pressure test exists, is to 'verify' that the vessel is safe for human occupancy. If the vessel implodes, then someone potentially made a calculation error, or there was a fabrication anomaly. For vessels which are too large to go in a test chamber like the one in the OP, the next option was to lower the vessel into a flooded mine shaft, or even lower the vessel from a ship into deep ocean water, down to at least 1.25x maximum depth. When they pull the submersible up, they just hope it is still in one piece.

It sounds like someone has no experience in designing/testing submersibles.
Where a submersible is designed to ASME PVHO-1, it is mandatory that the corresponding ASME VIII Div 1 or Div 2 internal pressure testing requirements are directly applied to the submersible. For external pressure testing, Div 1 & 2 are irrelevant. PVHO-1 mandates that an external pressure test of 1.25xMAEP must be completed.
Div 1 vessels designed to "design by rules" can be very heavy and therefore difficult to design to be buoyant, but it can be done. Div 1 App 46, Div 2 and PD 5500 with more accurate rules and FEM analysis methods, makes designing a light buoyant submersible, much easier.