Torsion on a seam weld

[AnimusVox]

Hey all,

I have a project where there is a seam weld attaching two aluminum panels as shown.

I'm uneasy about this (especially the amount of torsion in this case) and recommended using fasteners to attach the two panels (which I've recieved from pushback on, 'we've done this on previous projects without an issue', and so on). How would you start to analyze this weld for capacity?

Thanks!

 

[GastonLagaffe]

What kind of loads will this be exposed to? Anything that'll cycle through that torsion more than once or twice a month and I'd be expecting it to fatigue out very quickly. The crack is essentially already initiated at the root and you'll never know how fast it's growing until it just snaps one day. Is this done by a robot or a welder? You've got 1/8" total thickness to weld on - that ain't much of a bead! If this is anything close to structural and this has been done before, you should have PQRs - look at what effective throat was reached and you can start getting an idea what static loads you can take. The notching effect won't matter too much in static loads since welded aluminum is usually pretty tough (I still wouldn't push it too much though).

Other ideas would include cutting one of the panels a little short so you can build up a small fillet weld. You'll end up melting off the longer side but at least you can be a bit more confident about how much weld is actually on there.

But really, sheet metal screws or rivets is what you want to go with here. If you're worried about water-tightness, there are formulas that'll give you the spacing required (3" + 3t is what the Canadian aluminum code uses). Once your weld cracks it won't be any better for water-tigthness anyways.

 

[dik]

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Feel any better?

-Dik