Reply to post: Re: Over tightened screws?

Satellite cannon starts shooting Doves, this time under control


Re: Over tightened screws?

"How the hell can something as basic as torque be screwed up on something so expensive and destined to be rather remote?"

It happens all the time in the aerospace industry. The mis-torqued situations I've seen seem to go like this:

1) A new aerospace gizmo is designed. The engineer(s) behind the relevant drawing carefully, painstakingly address common issues for the fastener: material compatibility with the socket, galling, finishes (dry film lubes or wet installs - careful of off-gassing), fastener type/size/specification, and so on. However, no one bothers specifying torque because they don't think it's necessary for all those finely-trained, extensively-experienced guys on the factory floor (or haven't personally experienced a torque problem, or expect Manufacturing to step in and write give decent input on the assembly procedure).

2) Some years later, enough of the gizmos have been in service that lack of torque control is causing problems. The drawing is updated, assembly procedure is expanded, and more paperwork is added so the tech doing the bolt-torqueing is now required to write down the torque applied to each of the 37 bolts he's responsible for. Torque problems vanish, customer signs off on the changes, everyone goes out for a beer.

3) Torque problems resume a year later, mostly occurring in older gizmos but also in some coming out of the factory. Drawings and instructions are verified as being correct; paperwork shows the techs are doing the correct job (or at least filling out the paperwork with the correct numbers); attempts by engineering to interrogate the responsible tech properly (empty room, one chair, bare bulb swinging overhead, etc) are blocked by union representatives, so engineers must content themselves with shop floor chats where the tech swears they're following the orders the engineers wrote.

3a) Finally, an intern assigned to sift gizmo data realizes that all the gizmos having trouble have either undergone recent maintenance in which 37 critical fasteners are removed and replaced, or they failed testing in the factory and had to be repaired, which also entails all 37 fasteners removed during the repair process.

3b) A quick check of the depot maintenance process shows that the maintenance instructions do mention torque wrenches, but a site visit reveals the customer's techs prefer their own socket wrenches and were using the printed maintenance instructions to prop up a wobbly table. Meanwhile, the factory's own repair procedure was never updated to reference torque for the fasteners when the drawings and assembly procedure were fixed a year ago. The repair instructions are updated, the customer signs off on the changes, and everybody goes out for a beer.

4) 6 months later torque problems resume, this time caught on the factory floor. It turns out the fastener vendor changed the passivation process on the stainless fasteners to something more eco-friendly after being threatened with plant closure by the local regulatory body. The vendor's hasty study and certificates of conformance claim the fasteners perform just like the last ones, so Procurement didn't notify Engineering of the change. However, now the torque characteristics are different despite the vendor's claim. A review board is held, drawings updated, instructions (all of them) updated, a representative visits the customer at their maintenance depot, and all will work well for at least the next 3 years.

Of course, if you're only building a single gizmo and installing it on a space station where maintenance is infrequent and difficult, then that learning curve is not possible.

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Biting the hand that feeds IT © 1998–2022