Dive Cylinder Explodes - Sydney

[Wookie]

There may be a potential issue that remains unaddressed with the the procedures of the hydro test and then stamping of the cylinder, with the possible damage induced by the stamping. I was told that my local shop had one of those cylinders develop a crack after hydro and eddy current testing, when they stamped it. Scared the hell out of them. They said the only way to know is to stamp it first. Then see if it passes hydro and eddy tests. I'm not an expert, just sharing.

You are spreading lies and innuendo. You are right, not an expert. The failure mode of the 6351 cylinder is sustained load cracking, not brittle fracture. That means the crack forms and propagates under a sustained load instead of a sharp blow. A sustained load like putting 3,000 lbs per square inch inside the cylinder, specifically in the neck, not by placing a hydro stamp on the shoulder of the cylinder.

I'm going to sound like a dick and although I don't mean it that way, it's hard to type and convey this softly. You might consider studying and reading what is said on a subject instead of spouting what you heard down at the dive shop. This is the second thread of yours I've read. The first dealt with rebreathers, where you touted the wonders of the APOC. Now you're telling us that 6351 alloy scuba cylinders are bombs, just waiting for a sharp blow to blow up and take our legs off.

Neither is remotely correct.

 

[Diving Dubai]

If I may enter the thread.

I should declare that I'm not nor never have been involved with scuba tanks, Don't know the full history without spending time reading detailed papers (rather than forum posts) But am an ex materials scientist and Course Director of NDT inspections within aerospace, Nationally and Internationally.

@Wookie is correct

The Sustained load cracking is an intergranular defect generally accepted to be cause by high pressure on the material over prolonged periods. The presence of lead as well as bismouth at the grain boundaries can accelerate this.

6351 Alloy had the most defects, even with the lead content reduced. To be clear while most defect have been found in the neck, there have been many in the shoulder as well as others found in the main cylinder area.

The defect is slow growing with a period of 8 or more years before it becomes critical.

Eddy current testing has been put up there as a way of detecting the defect. Unfortunately the equipment used is the equivalent of a "My first Eddy current tester", and I'm assuming the operator training is limited to a day or 2 just on that technique and nothing else? (Stand to be corrected)

What makes the inspection worthless, is that is a defect is found, the area is cleaned and inspected again. if a defect continues to be located then you confirm by visual aids to prevent false positives. That's a quote (roughly) from the operator guide.

Eddy Current Flaw detection (EFD) is an order of magnitude more sensitive that visual, it need no visual conformation if a correct technique is applied with proper equipment and a fully trained operator. EFD is designed to detect surface and subsurface defects (Find me a sub surface defect with the naked eye)

The fact that The EFD process and (I Presume the visual) was for the neck only when it's obvious that a defect can propagate anywhere in the material just scares me (as a materials scientist)

Given that a Ally tank is apporx 1/2 thick in the thinnest areas, I wouldn't' even used EDF - possibly Ultrasonic inspection (I've devised procedures to find minute defects in thick Ally components) but of course the investment in equipment and training doesn't make it worth it.

Visual inspection is open to interpretation as well as significantly under human influence (training, experience, the state of the person inspecting that day) Ideally in any inspection you should be able to miss a defect on 3 different inspection periods before it goes critical.

Stamping the cylinder (given the wall thickness) isn't going to have an effect unless you're king kong with a super size sledge hammer.

All that said, there is no proof that (yet) that's its a 6351 cylinder. Again as a professional I abhor the practice of "cave fills" and the like as you are eating in to the safety margin calculated for a new cylinder with zero defects. The safety margin has a tolerance to allow for material defects and the like. So if you continually stress a cylinder that was at the lower part of its tolerance bad things can happen.

I know people cite the number of fills vs the number of fails, but that citation doesn't help the person filling a tank when an accident happens by telling then it was a 1:1,000,000 incident.

 

[BRT]

All in all a cylinder blowing up next to you is one of the low probability ways to die while diving. It is not a high probability thing like the danger of a moose coming through the windshield while driving or the plane being crashed by terrorists while flying.

 

[cerich]

As far as I know, every dive shop stopped filling them. However, it seems to me that a lot still use them as their course/hire tanks. The only old aluminium tanks I see at dive sites are very obviously dive shop owned ones.

If I may enter the thread.

I should declare that I'm not nor never have been involved with scuba tanks, Don't know the full history without spending time reading detailed papers (rather than forum posts) But am an ex materials scientist and Course Director of NDT inspections within aerospace, Nationally and Internationally.

@Wookie is correct

The Sustained load cracking is an intergranular defect generally accepted to be cause by high pressure on the material over prolonged periods. The presence of lead as well as bismouth at the grain boundaries can accelerate this.

6351 Alloy had the most defects, even with the lead content reduced. To be clear while most defect have been found in the neck, there have been many in the shoulder as well as others found in the main cylinder area.

The defect is slow growing with a period of 8 or more years before it becomes critical.

Eddy current testing has been put up there as a way of detecting the defect. Unfortunately the equipment used is the equivalent of a "My first Eddy current tester", and I'm assuming the operator training is limited to a day or 2 just on that technique and nothing else? (Stand to be corrected)

What makes the inspection worthless, is that is a defect is found, the area is cleaned and inspected again. if a defect continues to be located then you confirm by visual aids to prevent false positives. That's a quote (roughly) from the operator guide.

Eddy Current Flaw detection (EFD) is an order of magnitude more sensitive that visual, it need no visual conformation if a correct technique is applied with proper equipment and a fully trained operator. EFD is designed to detect surface and subsurface defects (Find me a sub surface defect with the naked eye)

The fact that The EFD process and (I Presume the visual) was for the neck only when it's obvious that a defect can propagate anywhere in the material just scares me (as a materials scientist)

Given that a Ally tank is apporx 1/2 thick in the thinnest areas, I wouldn't' even used EDF - possibly Ultrasonic inspection (I've devised procedures to find minute defects in thick Ally components) but of course the investment in equipment and training doesn't make it worth it.

Visual inspection is open to interpretation as well as significantly under human influence (training, experience, the state of the person inspecting that day) Ideally in any inspection you should be able to miss a defect on 3 different inspection periods before it goes critical.

Stamping the cylinder (given the wall thickness) isn't going to have an effect unless you're king kong with a super size sledge hammer.

All that said, there is no proof that (yet) that's its a 6351 cylinder. Again as a professional I abhor the practice of "cave fills" and the like as you are eating in to the safety margin calculated for a new cylinder with zero defects. The safety margin has a tolerance to allow for material defects and the like. So if you continually stress a cylinder that was at the lower part of its tolerance bad things can happen.

I know people cite the number of fills vs the number of fails, but that citation doesn't help the person filling a tank when an accident happens by telling then it was a 1:1,000,000 incident.

just a FYI, nobody sane "cave fills" any alm cylinder.

great post by the way!

 

[Storker]

nobody sane "cave fills" any alm cylinder.

So, that US practice of grossly violating regulated safety margins is only for steel tanks, not for Al tanks?

I'm fine with a 10% overfill to my (steel) tanks since the gas in the tank heats up while filling, so the actual tank pressure will be below service pressure once the tank cools off. More than that? I'll pass, thanks. I figure there's a good reason for those limits, and I won't second-guess those reasons because I'm not a material scientist.

 

[cerich]

So, that US practice of grossly violating regulated safety margins is only for steel tanks, not for Al tanks?

I'm fine with a 10% overfill to my (steel) tanks since the gas in the tank heats up while filling, so the actual tank pressure will be below service pressure once the tank cools off. More than that? I'll pass, thanks. I figure there's a good reason for those limits, and I won't second-guess those reasons because I'm not a material scientist.

Yes...it is steel tanks only.

 

[Shotmaster]

You are spreading lies and innuendo. You are right, not an expert. The failure mode of the 6351 cylinder is sustained load cracking, not brittle fracture. That means the crack forms and propagates under a sustained load instead of a sharp blow. A sustained load like putting 3,000 lbs per square inch inside the cylinder, specifically in the neck, not by placing a hydro stamp on the shoulder of the cylinder.

I'm going to sound like a dick and although I don't mean it that way, it's hard to type and convey this softly. You might consider studying and reading what is said on a subject instead of spouting what you heard down at the dive shop. This is the second thread of yours I've read. The first dealt with rebreathers, where you touted the wonders of the APOC. Now you're telling us that 6351 alloy scuba cylinders are bombs, just waiting for a sharp blow to blow up and take our legs off.

Neither is remotely correct.

Whoa! I simply shared what was relayed to me while I was at my LDS who does hydros, and one of the few shops that will hydro and fill the older aluminum tanks. I know the owner and he told me what he had experienced that day and was saying that he was seriously re looking whether they might stop working on these cylinders. He has been doing this for decades and he said the tank passed hydro and eddy tests, but developed a crack during fill. He said it is a wake up call for him. I possess no expertise on this subject. My friend said he has been comfortable working with these tanks and understood the failure mode and this was the first time he had seen this happen. So don't shoot the messenger!

To address your flaming me about my posts on the rebreather thread, I have been looking at KISS for a few years. The short comings I see for all of them are the O2 sensors, which were discussed. I have looked at everything from DIY Kiss units to the most expensive off the shelf units. I had recently, like two days before the thread started found the APOC during my comparisons of the eCCR units on the market and liked what they advertised on their website. Which I posted. I have since learned that the owner has some serious issues and has not delivered these to the market either after a lot of fanfare. I had no idea it was vapor ware. My apologies.

 

[Diving Dubai]

Whoa! I simply shared what was relayed to me while I was at my LDS who does hydros, and one of the few shops that will hydro and fill the older aluminum tanks. I know the owner and he told me what he had experienced that day and was saying that he was seriously re looking whether they might stop working on these cylinders. He has been doing this for decades and he said the tank passed hydro and eddy tests, but developed a crack during fill. He said it is a wake up call for him. I possess no expertise on this subject. My friend said he has been comfortable working with these tanks and understood the failure mode and this was the first time he had seen this happen. So don't shoot the messenger!

It's a worry that your friend doesn't understand things.

If the tank sprung a leak on fill, post eddy current test. I'd stop using that place because they'd failed to spot the defect of the eddy current test and the visual inspection ( assuming it wasn't a cylinder wall defect)

These defects don't just suddenly happen, they become critical only because they haven't been found on the regular inspections.

These old tanks if properly inspected should be okay, the fact is that the industry doesn't have the expertise and is unwilling to make the investment in testing procedures. It's cheaper and easier to get the individual to replace their tanks

 

[wetb4igetinthewater]

You can't fix negligence either. I bought a 40 cu ft cylinder less than a year ago that I was to use for technical diving as a 100% O2 bottle. The owner of the shop put O2 in it without O2 cleaning it apparently and not conducting a VIP either. I discover this because I'm decided to have this cylinder O2 re-cleaned at another shop and that shop owner notified me of the issues and told me I was lucky it didn't ignite.

 

[BoltSnap]

You can't fix negligence either. I bought a 40 cu ft cylinder less than a year ago that I was to use for technical diving as a 100% O2 bottle. The owner of the shop put O2 in it without O2 cleaning it apparently and not conducting a VIP either. I discover this because I'm decided to have this cylinder O2 re-cleaned at another shop and that shop owner notified me of the issues and told me I was lucky it didn't ignite.

Here in Libya, commercial spearfisherman who spearfish on scuba in depth of 50 - 60 meters use 100% O2 for deco in bottles that were NEVER O2 cleaned at all! These tanks are very old and were never prepared for O2 service at all. No incidents so far with these tanks.