Exploding Tank in Utila

[2airishuman]

Which shows the limitation of the PSI course, because that statement on Eddy Current testing is so untrue it’s laughable (not aimed at the messenger)

I had understood from your previous posts in other threads that it is your opinion that the Eddy Current testing performed by most dive shops and hydro facilities is so ineffective it's laughable, regardless of whether the container being tested is made of aluminum, steel, fiberglass, marine bronze, or white oak. I believe you made some specific criticisms of both the equipment used and the operator skill level.

I believe PSI's position is that use of eddy current testing equipment and procedures on tanks made of 6061, or of steel, leads to false positives without any useful safety benefit.

 

[Diving Dubai]

I had understood from your previous posts in other threads that it is your opinion that the Eddy Current testing performed by most dive shops and hydro facilities is so ineffective it's laughable, regardless of whether the container being tested is made of aluminum, steel, fiberglass, marine bronze, or white oak. I believe you made some specific criticisms of both the equipment used and the operator skill level.

I believe PSI's position is that use of eddy current testing equipment and procedures on tanks made of 6061, or of steel, leads to false positives without any useful safety benefit.

Well remembered. And you're right. That is precisely my opinion of the Eddy current procedure used for dive cylinders, the equipment and training.

To properly train an operator requires 2 weeks (PSI is what 2 hrs?) They are "trained" to repeat a procedure on a specific component, of a specific material, and if they get a defect indication to confirm it visually, because they do not have the training to be able to interpret the to signal from the equipment.

So yes, in the way it's used you might as well not bother with Eddy current testing apart from on 6351 cylinders and only then because its pretty hard to miss a significant defect

In the real world the opposite would happen, and that is a defect identified visually would be put forward for eddy current (or UT or Penetrate etc etc) to confirm or dismiss the indication.

With Visual, its very subjective - especially with corrosion. There are tools you can use to measure depth of corrosion but they're not available (nor affordable) to dive shops. You can say by Ultrasonic (UT) measure and quantify to a good degree of accuracy corrosion depth and area, but again outside the scope, skills and affordability of a dive centre

 

[Diving Dubai]

...

 

[abnfrog]

you are trained to interrupt the data and its more than 2 hours we did a full day and remember we are only looking at alum tanks not other metals . ive been doing it for 10 years now

 

[Diving Dubai]

Thanks for that. I perform underwater Ultrasonic Testing of ships hulls for a good time. Although I am not a metallurgist, and am merely a lowly technician, I understand the importance of procedures as it comes to performing NDT, and the problems you can run into when using the wrong procedures or tools to test the wrong materials.

The Visual Plus machine is perfectly capable of detecting sustained load cracking defects in 6351 cylinders. As 6061 cylinders are not subject to sustained load cracking, the test is worthless unless a manufacturing defect is present in the cylinder, one that was missed by the manufacturer.

But thanks for man'splaining NDE to me.

This is a bit of a rabbit hole, but lets stat down it, if only to clarify and add to people's understanding.

Sustain load cracking, is basically an inter-granular defect. Simply put, a weakness of the grain boundaries. In this case a mild load as normally encountered by day to day use of the cylinder can cause a defect to initiate. Generally in the threads and neck area but not exclusively so.

With a change of the alloy composition that weakness was eradicated, and thus the stresses and loads the materiel under goes in its daily life aren't enough to trigger a defect (because the material is stronger). But those stresses and loads are still there.

As you correctly pointed out, a defect and be triggered by having a weakness caused by a material defect (inclusions in the raw material) or an issue during manufacture not picked up by QA - normally this would be chemical or heat treatment, which would require the natural ageing of the materiel to become apparent.

Back to Eddy current. With aluminium, differing alloys have no real impact on the Eddy current inspection (EFD). As far as EDF goes, as long as it doesn't have a surface coating(clad or anodized) its pretty much the same. People don't EFD 6026 because it's unlikely it will crack - you're generally looking for corrosion or mechanical damage. If however you did inspect it and see a crack indication you'd be unwise to dismiss it.

When you get to Steel and its alloys, as far as EFD is concerned its a whole different ball game, because different materials, alloys and heat treatments all have a major impact. So an EFD procedure for one isn't' transferable on another. On the other hand as you well know, UT is - as long as you adjust for the slight differences in sound velocity and get a back wall echo (on compression wave) you're good to go

I am fairly sure that different materials are subject to different failure modes when subject to different stresses, such as neutron flux or heat. Austenitic Stainless Steel is subject to chloride stress corrosion, where inconel is subject to chloride pitting corrosion. Hint. The failure modes are different, and the methods for testing are also different.

But this isn't in anyway the same or comparable as finding defects in aluminium - yes I've written procedures for the nuclear industry (I enjoyed them as they were the most challenging)

 

[falcon125]

I wonder if the rebar was the last thing the tank hit? It appears to have failed at the bottom.

 

[Jonny Wishbone]

Just think for a second,,, What are the odds of a guy probably carrying 2 tanks down a dock heading towards the boat he probably works on in some capacity or another for the dive operation and a boat with people in it pull up to a portion of the dock and just as the guy walks by,,,Kaboom!!! and kills the girl and takes off part of his leg. That tank could have blew any time after the fill. Maybe as it stood among numerous other tanks waiting to be loaded on a boat. Maybe after it was on the boat half way out to the dive site. Maybe while a diver strides off the boat with it strapped to his back. How about before or after he passed by the girl with enough distance that she was not harmed when it blew. Crazy huh?

 

[iain/hsm]

In that line of thought, maybe had it been only one cylinder he could have carried it over his shoulder next his head.

A small point I would like to add is that not all mill certified materials of either 6351 or 6061 are the same.
The assumption is that all 6351 alloy is susceptible to SLC and all 6061-T6 material is not, and while this may be valid the assumption remains with divers and scuba test houses that all material regardless of origin is that all CofC certification of conformity and material certs are an assurance of the same identical materials regardless of the various worldwide raw material suppliers.
Granted while most divers would not have access to the bar stock origin certification data or know the meaning of the numbers stamped on the base of each individual cylinder extrusion or the relivence of the various raw material suppliers or even the country of origin.
Yet all aluminium cylinders that have exploded to date (and not due to heat or corrosion failure) have all been cylinders that used raw material billet that originated in China.

 

[Wookie]

Because of the problems with any American alloys, if you're going to build an Incat vessel in America, the aluminum is shipped from Australia. Likewise, if you put aluminum repair plates in an American vessel, the USCG requires that you trace the batch number back to the producer and keep the plate history recorded for all time.

So the point that one alloy is susceptible and one isn't is rubbish is a good point. And that's why we have QA.

 

[Diving Dubai]

you are trained to interrupt the data and its more than 2 hours we did a full day and remember we are only looking at alum tanks not other metals . ive been doing it for 10 years now

Thank you for the correction of the length of training of a full day.

I am not for one minute suggesting that you don't apply anything other than the highest level of care and diligence when carrying out your inspections and that you apply the full extent of your training and experience when you do so.

Unfortunately the industry has been sold snake oil

To give you an example. The American Society of NDT (ASNT) approved courses are 2 days for Visual (VT and 4 days for Eddy Current (ET) this only the classroom, you are required to complete elearning before hand for each.

How does your 1 full day of ET sound now?

The followign disclamer is on the fist page of the Visual plus manual. (Their emphasis, my bolding)
VisualPlus 3 is an eddy current inspection tool to be used in conjunction with visual cylinder inspection. Eddy current inspection IS NOT A SUBSTITUTE for visual inspection by a properly trained inspector. EACH CYLINDER MUST BE VISUALLY INSPECTED BY A PROPERLY TRAINED INSPECTOR AFTER EDDY CURRENT INSPECTION. We recommend our OpticalPlus product to assist in visual cylinder inspection.

As to interpretation of defects, - there is nothing to interpret. You get a spike on the data trace if it runs over a longitudinal defect. It can't find transverse ones (say in thread root). Even their data traces are misleading and exaggerated. You certainly can't decide if it's a natural crack, mechanical damage or corrosion.

This is a screen shot of a crack in a hole (calibration standard) using a proper handheld ET tester. It's telling you the type of defect (top image) the clock position in the hole (bottom trace) and teh width of the defect. As you can see this is using a rotating probe running at 2679rpm (no the 150rpm that visual test think is advanced and fast).

I don't mean to discourage you I just wanted to point out the disservice given to people be the claims made of this kit by the manufacturer.

The most important rule in NDT is not the amount of defects you find, but the amount you miss