Looking for SP6498 or AA-6351 Bad Alloy Cylinders --- Must be cracked or ruptured

[richerso]

OK. You guys have found some things that I missed.... btw, I have been searching the research databases; but there is a lot of data to go through... Also, please be careful, let's not get general cylinder failures conflates with the topic that I am interested in..... Having said that, I do have one technical comment and one editorial comment here.

First, the editorial.... I love the "valve conversion" job the guy did from metric M25 to 3/4 NPSM... Classic.

Technical observation: If I can steal these two photos from the scubaengineer.com website,

Image on the left shows what I would expect a properly heat treated cylinder to do. Note that the rupture initiates, in this case, at approximately the mid point location of the body of the cylinder; although the initiation point for this discussion is not relevant. The crack then propagates some distance before self arresting. Note that the cylinder has some outward facing plastic deformation as if it has suffered an aneurysm. The plastic deformation is very indicative of a ductile material. This is a classic ductile failure.

Image on the right shows what I would expect from a brittle failure. Note that there does not appear to be any plastic deformation in the cylinder (just as you don't see bent shards of broken glass if you ever dropped a piece of glass). Further there is more than one fracture line; and as should be clear, the fracture line did not self arrest. The fracture just continued to propagate until pieces of the cylinder were flying. Classic brittle failure.

I still need a material sample. Metallurgical and metallographic analysis will definitively show either a proper heat treatment or an improper heat treatment.

 

[Luis H]

We have two issues here. One is the material condition. The other is the load causing the failure. They affect each other, but they are not the same thing. Even an extremely brittle cylinder will not fail if it is not loaded. The load and the material reaction are related, but are two subjects.

Yes, I have seen failures (cracks) due to residual fabrication stresses (in other structures), but I am sure that is not what we are seeing here.

I totally appreciate your interest in getting material condition data. I do have one original 1972 cylinder with a crack, but I am not willing to part with it due to its vintage collector value.

My observations are that crack propagation theory is traditionally associated with cyclic loading. But the number of cycles in a scuba cylinder are miniscule (even in a rental fleet where they are filled daily). There has to be a load to cause a crack and I don’t consider a few hundred or even a few thousand cycles (at most) a real cyclic loading. It is definitely not a reversing load.

It has also been documented that some of the failed cylinders have extremely low cycles (like just a few cycles) before failure. We are really talking about a static load.

I would agree that I have not really heard about SLC (sustain load cracking), in other structures (in my decades as a mechanical and structural engineer). But we learn new stuff everyday, if we keep an open mind. Crack propagation theory is an evolving subject… I have seen substantial new knowledge in the past few decades…

I am looking forward to your material condition findings, but there are two sides to the equation.

Note: (If I remember correctly) I read a while back that the alloy is known to have some issues. It certainly appears to be, not as ductile as it should have been. I thought some of the known issues had to do with the alloy composition, not heat treatment.

Good luck