Poor material control causes helicopter crash

It's had to believe that this error was not picked up.
A chain of events that started with keeping stock rod sorted by size rather than material lead to the fatal loss of a helicopter main blade.

https://www.tsb.gc.ca/eng/rapports-reports/aviation/2021/a21w0045/a21w0045.html

in a PDF
https://www.tsb.gc.ca/eng/rapports-reports/aviation/2021/a21w0045/a21w0045.pdf

Unsurpsingly the company is no longer in business. Loss of life, employment and income for a error that was easily detected and should have been by a mandatory test.

Robert G8RPI.

There's a hint in the report that the selection of the wrong metal may have been more than a storage issue. The company hadn't ordered enough H11 steel to make 350 pins, there was only enough to make 320. Did someone decide 316 Stainless was 'good enough'?

More startling to me was the way the manufacturer (a subcontractor) dropped the ball when the pins came back from heat treatment, another subcontractor. The heat treat companies spot tests had detected that some pins weren't all properly hardened, so they tested them all, told FMC there was a problem, marked the failures with a yellow stripe, and returned the good and bad pins in separate boxes.

Unfortunately no-one told FMC's overworked QA Inspector there was a problem, he didn't read the paperwork, and then - because he was overloaded - the hardness tests were done by unqualified production staff whilst he did the dimensional checks.

It's very common to find accidents are the result of several small errors. Having a Slack Alice or persons contemptuous of H&S on the team doesn't help, but it's extremely difficult for good people following a routine to stay alert. And it's daft for management to expect unqualified staff will realise they're doing the job badly!

The pins could be made on a mini-lathe...

Dave

Edited By SillyOldDuffer on 27/10/2022 09:23:13

That report was an interesting read.

Shows the lengths proper manufacturing concerns go to to get things right and how errors can stack up to defeat the system. I guess the folk involved in setting the procedures didn't envisage that completely wrong material would have been used.

Dangers of modern automated and semi automated procedures set-up by people who know I suppose.

The "nobody would be that stupid" or "any person with experience would spot that" assumptions. When setting up such procedure its often hard to realse that the person doing the work doesn't have your experience and knowledge. Maybe more, may be less.

The one thing that did surprise me was that the pins that failed testing at the heat treatment firm were simply paint paint marked rather than destructively cut. I'd have expected an angle grinder cut or similar as the heat treating firm would surely have been aware that this material was seriously out of specification. I'd have thought that getting wrong material was something that happened often enough for the heat treaters to have robust procedures.

Clive

Surprised?

I am reading my way through old MEs, and repeatedly the use of Silfos in making boilers comes up. Why is this any different?

I would also mention the foolish method of making the boiler with just a seam joint between barrel and throatplate, not fully flanged. Again and again this is declared as a safe method of making a joint, possibly 8" long and at 100psi, daft.

Thank you for posting that - it made very interesting reading.

I was in Production Control management of an aircraft parts company in the early '80s. We had a number of issues with materials and process control, mercifully none as serious as this one. Problems leach into every aspect of storage and manufacture - for example, the laborious investigation necessary to establish whether a specific material called for on a drawing already existed in inventory (prior to introduction of computerised systems) led to new local part nos. being taken out to avoid the work of finding an existing one. Whilst implementing the computer system, I found that one type of cable was being held under 11 separate identities in stock, and nobody in engineering or inspection was aware of more than a couple of them. And there were scores of similar examples.

As Model Engineers, our materials selection and control issues are at kindergarten level compared to those in serious industry.

A while ago, the cockpit window on one aircraft failed due to someone deciding that BA screws were a good enough fit.

Regards Ian.

I was told years ago that in helicopter maintenance, if a part would have say, 6000 hours use in a fixed winger, it would be replaced at only 2000 hours on a rotary winger, to ensure safety. One third normal usage. That was when I worked for a flight simulator company.

Not quite

It was a BA BAC 1-11, G-BJRT. A number of fastners were 8-32 UNF instead of the correct 10-32 UNC. (4 of the 10-32 screws were not to the latest standard (slightly short).The windscreen was fitted by a manager (still holding a valid engineering licence) because the staff were on strike. He got the screws from a dispenser in a dark hangar and the were mixed in the container.
Additionally the deformed thread locking captive nuts were tight enough to allow the 8-32 screws to pull up but then slip. The manager was using a torque wrench but was using a plain 1/4" socket to hold the screwdriver bit and had to hold it in place. this hid the screw and he thought the thread jumping was the torque wrench clck.

The pilot was sucked out of the window when it blew out. He was trapped by his legs and survived. The cabin crew held his legs but they thought he was dead and were worried about him hitting the tail if he was released.

https://assets.publishing.service.gov.uk/media/5422faa7e5274a131400078d/1-1992_G-BJRT.pdf

The process did consider the use of wrong material. It required testing of samples of the stock materal for correct composition etc. Unfortunatly the subcontractor was not aerospace approved so did not have batch control. The olny sent two samples for test assuming that all the bar stock was from the same lot.
Bell have some responsibility because their QA audits did not identify the lack of proper stock and batch control.
Ever pin was serialised and required to hardness tested as well.

At least with proper traceability they will be able to find all the suspect components and replace them. The paperwork is expensive, but you can see why it's needed.

I ended up some time ago in a 'product support' role working for a firm making aircraft seats and galleys, we were extremely regulated and it beggars belief that someone used a non certified supplier for flight critical parts, a right PIA as regards paperwork but you can see why it's needed. There were conflicting business goals trying to force engineering to accept non certified parts to get goods out the door and earn cash but that had to be resisted, one clueless buyer said to me 'it's only glue' when he couldn't source the certified adhesive as if anything would do.

Tony

Have a look for a book (freely available in pdf form) called 'Truth, lies and O-rings' for the ultimate expression of this.

I haven't looked yet but am assuming it is the space shuttle disaster?

Tony

Remind me never to send anything to the Foster Heat Treatment Company Ltd LLC Inc. The idea is something that would be good to write into a procedure for critical parts, but in the absence of that procedure no business would do this on their own initiative to a customer's parts.

Edited By MW on 27/10/2022 16:56:34

I dread to think how many people have died over the years because of a failed O ring.

Many years ago I witnessed Production deliberately putting the wrong mill cert numbers on production batch cards to get the product made and out of the door. The material used was the correct material for the item but because the mill cert numbers were wrong it meant that customer did not have the traceability of product that he required in his contract. Fortunately there was no threat to life and limb so no nasty consequences should the items fail, Company did not stay in business for much longer mainly due to other appalling practices. Dave W

DC31k

I'd always understood that when dealing with aerospace components it was a general rule to positively mark any sub-standard (or fraudulent) components discovered in a manner than prevented them being inadvertently used should they re-enter the supply chain. Obviously the parts have to be passed back to the purchaser so and investigation into what went wrong can be mounted.

Presumably the heat treatment people would have known these were aerospace parts and would have dealt with them appropriately. Bottom line is the results indicated that something had gone seriously wrong and that there was no way the stainless steel parts could have been up for any job that the correct, tool steel ones, needed to be specified for.

Robert

I don't think the procedures actually covered wrong material. Sub standard or incorrectly treated material certainly. But flat out drawing the wrong bar from stock no.

Something of a "no one could be that stupid" issue. However stored, drawing the right material for a job is so fundamental to any metal working shops business that it almost has to be assumed that the firm will have ways of getting it right. Ultimately you can't specify everything.

Seems that a more robust approach would have been for the bar supplier to have permanently marked the relevant type, batch and certificate number on the bars. Then FMC could have been required to use only marked material from the same batch or at least having equivalent certificates. Given that 8 bars f new materils were insufficient to make the number of parts ordered its an odds on bet that FMC had a bar in stock which they planned to draw out and use to make up the numbers. Objectively fairly shady but something that could have been gotten away with if the right bar were pulled.

Bottom line is that FMC corporate culture wasn't really up to aerospace demancds.

Clive

Edited By Clive Foster on 27/10/2022 19:27:29

Duncan, correct, because the pins were serialised and batch marked they had located all the pins from that batch within two days and replaced them or grounded the aircraft. This was long before they knew what the issue was.

Clive DC31k, The heat treatment company did all the right things. They contacted FMC right away, separated the suspect parts, marked each one and marke the container as "Hold"
They did not and should not have destroyed the pins as they were not their property and doing so may might have hidden evidence. The requirement to carry out a material composition test on the raw material was intended to detect wrong material so the process did cover this case. However FMC did not apply this properly.

Tony Pratt 1, There is no need for a sub-tier supplier to a airframe (or equipment) manufacturer to hold a aerospace authority approval. They just have to be approved by the manufacturer. In fact there is no standalone approval for subcontractors in any modern approval system I know of (EASA, CAA, FAA, TCA) It is just too much red tape. How you you get Intel for example to get multiple national aerospace approvals for a CPU market that is probably less than they supply as free samples to PC makers.

Typically the OEM will require a subcontractor to hold a modern third party audited quality system e.g. AS9100, but they don't have to. I personally am currently an approved supplier to two aerospace companies and I have no formal quality system. Yes, there are aircraft flying around right now with parts made in my spare bedroom. Infact I made two assemblies this evening.

Robert G8RPI.

Edited By Robert Atkinson 2 on 27/10/2022 21:45:20

As part of a set of wax die moulds for RR I was the bench person involved in finishing them for delivery. The largest produced a cone like piece about 12 inch diamiter and 8 inch long with a section of about 1/4 inch thickness the tool was a cube shape with a cone machined in one end and a side plate with a smaller cone to make the part thickness. The side plate was held in place by 4 swing bolts 1/2 inch UNC and about 6 inch long with nuts made to fit. The swing bolts were made from high tensile steel I think EN 16 T but the apprentice making them used wrong thread firstly and helped himself to some EN 1A to remake them without telling anyone. The result was the die opened under pressure with molten wax spraying out the first time it was used. The operator was slightly burnt and there was an investigation when the steel was found to be incorrect but the faulty parts made with wrong thread were found.

David