How old is "too" old for aluminum cylinders?

[Tanks A Lot]

[...]Eddy test can't see into the crown, only the threaded area. It cracked on a fold inside the tank crown.[...]

This should be emphasized much more and most cylinder inspection course materials gloss over it completely.

I don't want to dismiss the usefulness of Eddy testing too much, but I had several cracks that the machine simply couldn't pick up, as they sat below the threaded area in the crown. No, I did not mistake something looking for a crack, as these either leaked or I cut the cylinder open later to confirm the find.

The whole situation with SLC on 6351-T6 aluminium has narrowed the perception of too many inspectors in my experience. Cylinders made from this material get a very thorough inspection for cracks in the threaded area. Eddy testing is reasonably well at picking these cracks up.

However, the crown area is mostly neglected, even on 6351-T6. Furthermore, this obsession with 6351-T6 and cracks has made many inspectors blind to cracks on 6061-T6 cylinders or in extreme cases steel cylinders. While these do not suffer SLC, they are indeed susceptible to SCC (Stress corrosion cracking), as are most alloys for that matter.
In a nutshell, the grain boundary on these alloys can be anodic. If an electrolyte like water is introduced, this anodic grain boundary can slowly corrode, leaving a small empty area behind. Tensile forces from manufacturing and the contained pressure can start to pull the alloy apart along the now empty, corroded, grain boundary.

While this obsession with 6351-T6 and the neck area is somewhat understandable, I wish inspectors were taught that it is by no means exclusive to this alloy, nor the neck area. True, this alloy and area is most susceptible to cracks, but it doesn't mean other alloys or areas are immune to it. They may not suffer SLC, but SCC is a real possibility and does happen. On very rare occasions cracks have formed in the sidewall as well as the base of cylinders. This is usually much rarer, as the leftover tensile forces from manufacturing are much small in this area than the neck.

 

[Wallowa]

OK....but by this logic all cylinders are suspect and potentially pose a risk...look, empirically how many years have thousands of bottles been used in diving industry? How many have had a wall or crown or thread failure? I appreciate an analytical approach but the sky is not falling. Yes, put water especially salt water in a bottle and pump it up thereby increasing the Pp of oxygen and oxidation occurs....but until I see numbers of failures VS number of tanks sold and in-service that have not failed this is in my estimation making a mountain out of a mole hill.

Visual, eddy current, hydro or any inspection protocol only denotes a potential for cylinder failure, but how many tanks have actually failed? By failure I mean they leak air or have a dramatic structural failure. Only dramatic structural failures I know were the tanks previously mentioned that were painted with designs and then baked at 350F degrees to set the paint thereby killing the elasticity of the bottles; as I remember that was in the early '80s or late '70s.

I promote diving safety, that was my occupation, but likewise after 6+ decades of diving I have never seen or have firsthand knowledge of a cylinder cracking/splitting with a loss of gas. We have become a risk averse society; fine, but diving cylinders are not the diving risk factor, rather diver judgment is. Take a deep breath and apply a rational evaluation of how much of an actual threat cylinder failures are.

Then go diving and have fun!

DSO

 

[Tanks A Lot]

[...]but by this logic all cylinders are suspect and potentially pose a risk[...]

In a sense I guess they are, otherwise what is the point of inspections?

look, empirically how many years have thousands of bottles been used in diving industry? How many have had a wall or crown or thread failure?

I'm by no means trying to fearmonger and exaggerate in saying that every single cylinder is a potential deathtrap. I'm absolutely with you that nearly 100% of cylinders are sound are in perfectly good shape. Unfortunately that doesn't mean anything for the individual that gets caught up in an accident with one that wasn't within the nearly 100%.

[...]Visual, eddy current, hydro or any inspection protocol only denotes a potential for cylinder failure, but how many tanks have actually failed?[...]

When working on one particular island off the east coast of Africa I know of at least 6 cylinders detonating within a 2 years period.

[...]I promote diving safety, that was my occupation, but likewise after 6+ decades of diving I have never seen or have firsthand knowledge of a cylinder cracking/splitting with a loss of gas. [...]

I can only assume that this must have to do with the region being the U.S. you are basing your knowledge off. The further I worked in developing countries, the more this is an issue. Some places even have own words for compressor operators losing limbs in accidents.

[...]Take a deep breath and apply a rational evaluation of how much of an actual threat cylinder failures are.[...]

I personally have met over a dozen people with lost limbs or lodged aluminium parts inside their body.


Below is for example a cracked 6061-T6 Catalina cylinder from 1996 cracked in 2021 or 2022, I'd have to check the records. This crack developed due to SCC.
The crack developed along the grain boundary after the cylinder experienced minor corrosion damage. No fold or valley was present.

VisualPlus clearly showing the crack at 11 o'clock.

OptiPlus magnification of the very same crack.

The outside shows no damage besides the usual scratches, making impact damage unlikely.

I have to admit that I rarely have seen steel cylinders crack, leaving hydrogen embrittlement aside. However I can remember one crack in a Chromium molybdenum cylinder in the thread area. To this day I'm not entirely sure how it developed. The Catalina above is by far not the only cracked 6061-T6 I had in my hands, AliTek, a now defunct/absorbed South African manufacturer had some terrible long necks riddled with folds and valleys. I assume there is a lot of tensile stress left from manufacturing. These cylinders cracked vastly more often than anything else I have seen, leaving 6351-T6 aside.

Personal anecdotes or lack thereof are a notoriously bad example to base ones opinions off. Not all cylinder failures get reported into an official database. I'd be surprised if any substantial amount of them actual do get reported at all. Your personal lack of firsthand knowledge of cylinder failure may be more of a testament to the effectiveness of modern inspection protocols in the west than a lack of cylinders potential to catastrophically fail. Poorer countries or regions are often not as lucky and get second hand cylinders that no one wants in the developed world anymore.
We often think of SCUBA mainly for our purpose of entertainment in fun dives. Looking at developing countries, this is far from the truth. One island I worked on, which had a sizeable amount of dive centres for its size (45+ centres), was outnumbered by their daily cylinders filled from the local fisherman by nearly 30%. This local group of fisherman was in no position to condemn a bad looking cylinder as this would mean that there is no food for the family on the table for that day. So even a dodgy cylinder gets hooked up to the compressor.

My gripe is with most inspectors focus on cracks in 6351-T6. Yes, these are more prone to cracking due to SLC which 6061-T6 doesn't suffer from. But again, this by no means should be taken as 6061-T6 being immune to cracks.

Again, I want to agree with your basic sentiment that cylinders are safe. It is in fact amazing how much abuse they take before giving in. But especially us as inspectors should be aware of some basic metallurgy and what this can mean for certain cylinders. I have read far too often that catastrophic cylinder failures are unheard of, but as a worldwide community we should strive to achieve the same safety levels globally and for everyone.
And because just buying new is often not an option, we should educate inspectors everywhere to spot certain things.

Some further interesting reads may be:

• J.W.H. Price, R.Ibrahim - Observations of crack growth in aluminium cylinders (1999)
• Francis C. Cichy, Hilbert Schenck, John J. McAniff - Corrosion of steel and aluminium cylinders (1978)
• M. BOBBY KANNAN, P. BALA SRINIVASAN - Stress corrosion cracking (SCC) of aluminium alloys (2011)
• Kohji MINOSHIMA, Tadashi SUGIYAMA and Kenjiro KOMAI - SCC In High-Strength Steels (1990)

 

[Wallowa]

First thanks for the effort and information seen in your post. I am always learning; could you describe "When working on one particular island off the east coast of Africa I know of at least 6 cylinders detonating within a 2 years period."......what is the common denominator that linked these 6 cylinders causing 'detonations'; I assume detonation refers to an explosive failure?

"AliTek, a now defunct/absorbed South African manufacturer had some terrible long necks riddled with folds and valleys."; not familiar with AliTek, do you mean they made [manufacturer] dive tanks and they were the bottles that failed?

"Poorer countries or regions are often not as lucky and get second hand cylinders that no one wants in the developed world anymore." can't speak to this, but do you mean the countries you are referring to get hand-me-down abused/damaged bottles? Never heard of defective tanks being sent anywhere let alone to poor nations." I can only assume that this must have to do with the region being the U.S. you are basing your knowledge off. The further I worked in developing countries, the more this is an issue. Some places even have own words for compressor operators losing limbs in accidents." Not discounting your observations but I have never heard of a compressor operator losing a limb or suffering any serious injury while filling bottles. Western countries VS the 'poor' countries? Lack of knowledge or ineptitude filling or handling a cylinder? I do not see the risk of a scuba tank failure being an issue of global locations, rather from abuse of cylinders from overfilling, allowing water/moisture into tanks or physical damage.

Lots on the table here to be certain; but bottom line is damn few scuba tanks have ever failed and even fewer caused any injuries to divers. At what point do any of us consider a threat to be insignificant and just move on with diving? Nothing is 100% but at what point does the failure rate of anything preclude using it or dictate a required or recommended protocol [VIP, Eddy, Magnaflux etc] that may or may not have any value?

I do appreciate you posting what you have and your views and opinions. I will close and stay off this thread with the premise that scuba tanks can fail; but very rarely do and no amount of 'inspections' will prevent that failure, nor is the risk of injury to a diver a significant possibility. Western saying: "The proof of the pudding is in the eating". Thousands if not millions of dives over decades and the incidents of injuries from scuba tank failures is very close to zero. Poor diver judgements are the threat not tanks.

Take Care,
DSO

 

[Tanks A Lot]

[...]what is the common denominator that linked these 6 cylinders causing 'detonations'; I assume detonation refers to an explosive failure?[...]

Most often one can see some form of mistreatment, usually from a terribly "maintained" compressor. It is rare to see a compressor used by fisherman in an African country that does not pump wet air. However some AliTeks and one steel I have seen had cracks which I can't attribute to stress corrosion cracking (SCC). I'm still not sure why they developed cracks.
There is usually a few things that go wrong when crack develop. You may have a cylinder with high tensile stress from the factory. Couple that with someone inserting a DIN477-6 compliant valve into an EN144-1 cylinder neck, which will start to spread the neck apart when torqued down heavily, introducing more tensile stress, some corrosion to induce SCC and so on.

Yes, by detonations I mean explosive failure.

[...] not familiar with AliTek, do you mean they made [manufacturer] dive tanks and they were the bottles that failed?[...]

AliTek was a South African cylinder manufactured, analog to Catalina or Luxfer. They belonged to Hulett Aluminium, which is now Hulamin, which is owned by Tongaat Hulett, for DiveTek. As far as I know they do not manufacture SCUBA cylinders anymore. I'm not a fan of their design.

[...]"Poorer countries or regions are often not as lucky and get second hand cylinders that no one wants in the developed world anymore." can't speak to this, but do you mean the countries you are referring to get hand-me-down abused/damaged bottles? Never heard of defective tanks being sent anywhere let alone to poor nations."[...]

This does not happen as a hand-me-down. It is usually someone from a poorer country flying into a western country and buying second hand cylinders. He of course buys with the motto "The cheaper the better", as that will increase his profit margins when transporting them back home. He does not intentionally buy defunct/bad cylinders, but of course cheaper cylinders increase the risk of getting a bad one. This is how it comes that a ton of 6351-T6 cylinders end up in poorer countries for example. They are unwanted in the west and therefore cheap to buy.

[...]Western countries VS the 'poor' countries? Lack of knowledge or ineptitude filling or handling a cylinder? I do not see the risk of a scuba tank failure being an issue of global locations, rather from abuse of cylinders from overfilling, allowing water/moisture into tanks or physical damage.[...]

What you say is exactly true and all of it contributes to the cylinders failing. But we should keep in mind that different locations globally have different means at their disposal to implement and enforce inspections schemes. Poorer countries have much more pressing things on their plate than thinking about pressure vessel inspection standards. And this lack can lead to exactly what you describe.

[...]Lots on the table here to be certain; but bottom line is damn few scuba tanks have ever failed and even fewer caused any injuries to divers. At what point do any of us consider a threat to be insignificant and just move on with diving? Nothing is 100% but at what point does the failure rate of anything preclude using it or dictate a required or recommended protocol [VIP, Eddy, Magnaflux etc] that may or may not have any value? Western saying: "The proof of the pudding is in the eating". Thousands if not millions of dives over decades and the incidents of injuries from scuba tank failures is very close to zero. Poor diver judgements are the threat not tanks.[...]

Absolutely, I'm right with you here. If I came across as saying that all cylinders pose some inherent and mystical danger, that was certainly not my intention. Especially for the diver itself the risks are absolutely minuscule if not absent, leaving freak scenarios with dropping cylinders from a considerable height or on the valve aside.

I absolutely do not want to say that cylinders are cracking left and right, that is certainly not the case. Most of the time when they do, there is a reason for it, which can be attributed to some form of abuse.

What I do want to say is that we as an industry should strive to minimize pressure vessel failure as much as we can. Eddy testing tools are a great tool, but because they are only stipulated for 6351-T6 cylinders, I have met all too many inspectors disregarding crack possibilities in other alloys or locations other than the thread. Eddy testing has led to the thinking in many inspectors that cracks are only a concern in 6351-T6 with its SLC issues. Too many inspectors have never even heard of SCC or when and why it happens.

We already have great safety standards in the west, but if a little more attention can make things safer, I see no downsides, except a couple of minutes of work for the inspector. If tangentially we can thereby raise the safety in poorer countries, its a win-win in my book.

 

[Tracy]

Aluminium cylinders are never too old to keep out of the tip

View attachment 854167

I have cut them up into many many things, but haven't made a door stop yet. I will have to make one this week.
Here is the Bow T-post on my boat.

 

[tursiops]

I will close and stay off this thread with the premise that scuba tanks can fail; but very rarely do and no amount of 'inspections' will prevent that failure, nor is the risk of injury to a diver a significant possibility.

You are way off base in saying this, and it is both illogical and dangerous to say.
"Very rarely" does not mean zero.
Since inspections DO catch cracked and failing tanks, of course taking those tanks out of play helps prevent them from failing and hurting someone.
Is one limb lost a "significant" possibility? Ten? One hundred?
You have stated this as a premise. You now have the choice of proving it, or withdrawing it. Note that even ONE failure or ONE failure prevented by an inspection disproves your premise.
Otherwise your premise is hogwash, using wors like rarely and significant.
Meaningless.

 

[BRT]

You are way off base in saying this, and it is both illogical and dangerous to say.
"Very rarely" does not mean zero.
Since inspections DO catch cracked and failing tanks, of course taking those tanks out of play helps prevent them from failing and hurting someone.
Is one limb lost a "significant" possibility? Ten? One hundred?
You have stated this as a premise. You now have the choice of proving it, or withdrawing it. Note that even ONE failure or ONE failure prevented by an inspection disproves your premise.
Otherwise your premise is hogwash, using wors like rarely and significant.
Meaningless.

Yes, If it saves just ONE life..........

 

[Gone for diving]

I have cut them up into many many things, but haven't made a door stop yet. I will have to make one this week.
Here is the Bow T-post on my boat.
View attachment 854504
View attachment 854505

Do you Eddie current, and look for stress cracking on that T-post. I don't recommend hydro testing it...

Look really cool.

 

[Still Kicking]

I have cut them up into many many things, but haven't made a door stop yet. I will have to make one this week.

I am out on Kauai and on my morning walk there's a neighbor with a mailbox that is a polished metal pipe .... it got me thinking about repurposing a steel 72.

Cut the top off, weld on a hinge and a latch, add a small dive flag mail signal...mount it on a post. How cool would that be? It's on the to-do list

Edit ... of course I'm not the first to think of this LOL