Aqualung Legend LX First Stage Failure at depth

[rsingler]

I don't want to sound alarmist. It's fascinating engineering. The number of incidents is very small.
As suggested in post #135, it's relatively simple to protect yourself and benefit from the closure device.
1) Rinse your equipment after every dive trip
2) Ensure that when you attach your first stage you screw in the yoke knob until the Crown is fully depressed (1 1/4 turns)
3) Confirm that there is no obstruction to flow by watching the SPG during a vigorous inhale.
With those simple maneuvers, you should experience no problems whatsoever.

Our search is for the elusive etiology of this rare but scary problem. But I think the necessary preventative measures are straightforward.

Dive safe!

 

[Diving Dubai]

if tank pressure gets by one of the two o-rings around the Shutter Valve, it looks to me like it would percolate right out the seam between the Crown and the yoke retainer, which has no oring

You are right-I mean, that was a test and you passed! A closer look at the drawings shows an escape path (also a saltwater intrusion path) past the springs and between the moving parts.

THis is one of the things I like about the DIN version (which accept is a different animal). The shutter valve is held on with a hex screw which can become loose over time (the shutter crown can bind if there is a litter residual pressure) It only needs 1 quarter turn. The gas path leaks by your routes giving a clear indication when you apply pressure that something's not right. Happen to me many times, I always have a hex key in my dive kit

 

[Diving Dubai]

As I was mentioning on the phone: I think you are in the right track, but to calculate your pneumatic area, you need to subtract the area of the shuttle valve.

For the calculations @rsingler

The internal dia or the shutter crown is 12.5mm and the dia of the Shutter valve 9mm (I'll let you do the math)

The water pressure argument isn't compelling for me, because if it's going to act on the front face it's going to also act on the rear of the shutter crown, by virtue of it filling the void with the spring

If we assume that water can get behind the shutter crown then some corrosion may happen at the rear of the shutter crown - perhaps enough to prevent the shutter crown fully retracting

We know that statistically speaking this is a pretty faultless design. It performs without issue on many 100's perhaps 1000's of dives daily around the world. I do know some companies use them in their rental stock. At least on liveaboard company I know use them on their fleet (so washed only at the end of the trip)

But clearly there have been a tiny amount of incidents

So my contributing factors are/could be (at the moment)

Poor or incorrect servicing
Poor user maintenance
Human error

Possibly the flaw is that the design is too well made and that the tolerances are quite small allowing little latitude for error. Just a thought

But playing devil's advocate could a similar fault occur on a reg without ACD? That is if you take the ACD out of the equation is there another possibility that we're ignoring.

I'll leave you all to ponder that while I'm off diving for 3 days....

 

[Luis H]

So my contributing factors are/could be (at the moment)

Poor or incorrect servicing
Poor user maintenance
Human error


.

Those are very likely contributing factors and removing those factor could very well prevent the undesirable outcome, but I am interested in understanding the mechanism of failure or the “root cause”.

I am not an accident investigator, but I have been involved in my fair share of incidents/ accidents investigations and “root cause analysis”.

Those factors would just be part of your “fish bone diagram” in the root cause analysis.


Note: Anyone interested can just Google “root cause analysis” or “fishbone diagram” and will know exactly what I am talking about. It is just a simple methodology to try to determine cause and effect (there is nothing magical here).

Another factor to put in the “fish bone diagram” is the observations (from the reports) that tank pressure affect the gas flowing and closing and in the DAN report it was a repeatable effect.

That implies that there is a pneumatic effect directly involved with the tank pressure and it the description doesn’t imply that it is linear as the tank pressure changes, but more like an on/ off effect.

To me this rules out a clogged filter or other linear obstruction and points towards a mechanical state change in a opening / closing device.


I don’t want to jump to conclusions, but I think you can see some more cause and effect.

Again I am just adding stuff to my “fish bone diagram”.


Like Sherlock Holmes would say: You have to consider all the possibilities and then rule out all the impossibilities. Once you rule out the impossibilities, whatever is left, however improbable, has to be the solution.

For the calculations @rsingler

The internal dia or the shutter crown is 12.5mm and the dia of the Shutter valve 9mm (I'll let you do the math)

The water pressure argument isn't compelling for me, because if it's going to act on the front face it's going to also act on the rear of the shutter crown, by virtue of it filling the void with the spring

View attachment 483845

If we assume that water can get behind the shutter crown then some corrosion may happen at the rear of the shutter crown - perhaps enough to prevent the shutter crown fully retracting

...

I think that you misunderstood the water pressure point that @rsingler was making. He was talking about the water getting behind shutter valve crown and applying pressure in the same direction as the spring force.


The shutter crown will act like a piston (with pneumatic, hydraulic, mechanical contact forces and spring forces acting on it), but I am having a little bit trouble doing an accurate balancing forces analysis of it.

 

[lexvil]

I started reading this from the beginning again, I was wondering if any change was noticed before failure, there wasn’t. If someone had said something along the line of “it started getting really hard to breath and than stopped” I would feel better about using one but if I had one I would convert it to DIN (threat of a class action could possibly get AL to do this w/o charge??) and become religious about a pony.

I know the risks of diving and choose to accept them but I don’t like to compound them if I have options, same reason I have a perfectly good aluminum 80 in my shop for rigging and adjusting webbing, it passed hydro and the eddy current but after reading of the sudden effect damage if the old alloy tank decided to let go, the risk of damage in my shop was enough for me to self condem this tank.

I have also given up on a popular regulator brand (SB popular anyway) due to some issue that bothered me, there are simply to many rugged reliable regulators available to be bothered by little problems let alone big ones.

 

[Couv]

Sherlock Holmes? Occam's razor? Bah! It's time to find the problem and use Herman 's Band Saw! One of you guys need to sacrifice an ACD and let @herman create a working cutaway that we can closely examine. In fact, to rid ourselves of the problem all such devices should be cut in half.

 

[rsingler]

Okay. Have a Legend ACD coming from eBay. More to follow next week, after I service it and tune it up.
After our experiments, I'll have a serviced, used Legend LX in perfect condition ready for anyone interested. No, @couv, I'm not going to saw it in half.

 

[JackD342]

Hmmmm…. I'll add a little insight, but try to avoid judgement. I am an Authorized Aqualung dealer, and I do the service. I actually prefer having the ACD on any unit I rent to folks flying out on trips, so I lose less sleep over what they are doing with my rental gear when out of my sight. (A customer that floods a rental 1st stage is VERY unlikely to let you know.) I also dive a Legend LX, so yes I use the feature myself.

First, I can't tell from your post whether or not that first service was done by an Authorized (and trained, and receives updates, and has OEM service kits) service center. There are more than a few places out there that take on more than they should, and send out less than they should. I'll take in a brand I don't do, but it goes out to another (authorized) shop rather than "figuring it out" myself.

I will admit the ACD is a little tricky to service. It is not a reverse thread, but it is a reverse tighten - as I turn the wrench counter-clockwise, it pulls the part into the retainer from the back side. That can throw folks off. There was also a torque spec revision 7 years ago, and a recall notice last year for a very recent production run that was assembled with the old/wrong torque spec. So it has had a couple of issues. Here is the recent ACD bulletin:
http://www.aqualung.com/us/pdfs/Consumer_Safety_Notice_ACD_AQA.pdf

So would I mind if they redesigned this with a higher and less precise torque requirement? Sure. But I also have a pretty high confidence that if serviced and assembled correctly, following each step in order and using a torque wrench rather than "an experienced touch," that there should be no trouble.

I also suspect that if a different dust cap is used, and overtightened to the point that it might bind by compression and turn the yoke shutter valve, you could theoretically loosen the part by accident. I had one strip the threads during disassembly, and the only non-standard element was an old, flat, very hard dust cap. (The dust cap is used during disassembly)

 

[JackD342]

Adding a bit more after reading a LOT more of the other posts, and really thinking through the mechanics of the valve:

First, the "closed" shutter valve is unusual in that it is a metal to metal seal, no o-rings involved. The rubber is sealing the assembled parts, not the air flow through the closed ACD valve. Now I am thinking that a loosened shutter (thus raised higher than intended when the spring loaded retainer is compressed), might just be able to seal against the tank valve face, metal to metal. This could be even easier to accomplish on a convertible valve, where the metal face of the valve opening insert is barely recessed. The top of the shutter is a little wider than the 8mm hex opening on the convertible I just checked, so there is uninterrupted surface to seal. And... I am thinking those valves are pretty prevalent on liveaboard cylinders.

So that's my root cause theory of the issue, at least for now! I'll let someone else think through the physics of how the air would flow with higher cylinder pressures, and then seal closed at lower pressures.

I have a couple of shop ACDs to service soon, so I may just try that theory out as I assemble them, before I torque them down.

 

[JackD342]

Adding a bit more after reading a LOT more of the other posts, and really thinking through the mechanics of the valve:

First, the "closed" shutter valve is unusual in that it is a metal to metal seal, no o-rings involved. The rubber is sealing the assembled parts, not the air flow through the closed ACD valve. Now I am thinking that a loosened shutter (thus raised higher than intended when the spring loaded retainer is compressed), might just be able to seal against the tank valve face, metal to metal. This could be even easier to accomplish on a convertible valve, where the metal face of the valve opening insert is barely recessed. The top of the shutter is a little wider than the 8mm hex opening on the convertible I just checked, so there is uninterrupted surface to seal. And... I am thinking those valves are pretty prevalent on liveaboard cylinders.

So that's my root cause theory of the issue, at least for now! I'll let someone else think through the physics of how the air would flow with higher cylinder pressures, and then seal closed at lower pressures.

I have a couple of shop ACDs to service soon, so I may just try that theory out as I assemble them, before I torque them down.

Okay, I did a little experimenting with a partially assembled ACD - just the yoke, the shutter valve, the retainer, and the connector. I was able to block airflow on a Thermo convertible valve with 200 PSI in it. It wouldn't block flow on a full Sherwood convertible, a full Sherwood 5000, a full Thermo Deluxe, or a 1,200 PSI Sherwood 5000. (I checked and the contours of most standard valves are such that the shutter doesn't have to be much higher for the metal to metal contact than for a yoke/DIN insert, so I tried those too.)

I am not inclined to keep experimenting with different valves and different pressures, but I have at least confirmed one scenario where a loose shutter valve could completely restrict the air supply with regulator valve to tank valve contact, so that may very well be the basic issue. And at the very least, I can see it restricting flow in many yoke valves when loose, even if not completely blocking it.

So if you dive an ACD and are concerned, feel free to ask your service center enough questions to feel comfortable that the ACD was reinstalled correctly. And I would also highly suggest NOT changing your dust cover to something other than OEM, just in case that can grab and twist the shutter under the right conditions.