Friction Stir Welding

November 19, 2019 6:11 AM

Friction stir welding is a technique for joining two pieces of metal (often aluminum alloy). The process is potentially counterintuitive to watch.
posted by Sokka shot first (22 comments total)

Yes, that's weird. I would like a clear stationary image of the probe (?). Is it smooth, or what?
posted by Kirth Gerson at 6:35 AM on November 19, 2019

the wikipedia article shows some, Kirth Gerson: https://en.wikipedia.org/wiki/Friction_stir_welding#/media/File:Advanced_friction_stir_welding_and_processing_tools_by_MegaStir.JPG
posted by the antecedent of that pronoun at 6:45 AM on November 19, 2019 [2 favorites]

It looks like they took something that was going wrong and figured out how to make use of it. "Goddammit, this fucking aluminum keeps getting mushed all the fuck up when I try to work it!" "Heeeeey, that gives me an idea..."
posted by clawsoon at 6:56 AM on November 19, 2019 [4 favorites]

If you don't watch closely it looks like something rubbing on the surface is all that's happening. But at around 2:00 you can see a cutaway animation that shows there's a pin stuck in fairly deep into the material being welded. That must be the "stirring" part. See also at 2:49 where you see video of the pin being inserted into the material.

There is a lot of force being applied by the welding tool. What I think is most interesting is it's described as "below the actual melting point". I'd think with this much mechanical energy being put in you couldn't help but melt it! Not to mention welding is usually understood to involve melting. Not here though, they're just "stirring". Weird.
posted by Nelson at 7:19 AM on November 19, 2019 [2 favorites]

You don't need to melt a metal for it to become pliable. I think this is more akin to stirring thick clay, though I could be wrong.
posted by Orange Pamplemousse at 7:29 AM on November 19, 2019 [2 favorites]

I was wondering what it looks like when the tool leaves the workpiece at the end of the weld and they finally show a brief glimpse at the very end: it looks like a horrible sucking wound. That's, uh, not fantastic, but I guess the benefits of being able to weld thick aluminum without any hazardous light or fumes outweigh it?
posted by Mr.Encyclopedia at 7:36 AM on November 19, 2019 [1 favorite]

AFAIK, this technology/method came about as an offshoot/radial-direction mod to the same process used to join cylindrical things together - say, a thin valve stem and a thick valve head. I recall seeing the latter about 1998 or 1999 in the context of the redesign of railroad tank car vacuum relief valves. Basically, chuck up one of the two pieces in a lathe, put the other on the feed base, spin the lathe up, and smash things together!
posted by notsnot at 7:53 AM on November 19, 2019 [2 favorites]

> I guess the benefits of being able to weld thick aluminum without any hazardous light or fumes outweigh it?

There's smoke or some kind of fume coming off the plate near the end of the video, so even if the metal isn't melting, it's getting pretty hot and emitting something.
posted by ardgedee at 7:59 AM on November 19, 2019

Part of the benefit is that it doesn't warp the workpiece the way traditional welding does, but with all the crazy workholding you have to do I wonder if traditional welding would budge that material anyway.
posted by phooky at 8:01 AM on November 19, 2019 [3 favorites]

Why didn't they show us the other side, huh? Checkmate!
posted by seanmpuckett at 8:06 AM on November 19, 2019

it looks like a horrible sucking wound

It looks like a very clean weld to me, apart from the tailings being kicked up at the edges of the shoulder. And it can be polished down to be almost invisible.
posted by adamrice at 8:16 AM on November 19, 2019

Nice. I'm probably the last person on the planet to have discovered the plethora of fascinating manufacturing videos on YouTube. (How They Make the Modern Aluminum Can.) Luckily, I have an appointment, so my morning will not be wasted.
posted by kozad at 8:28 AM on November 19, 2019

It seems more like something you could do with plastic, it's amazing they're doing it with metal. A lot of mooshing going on here.
posted by Bee'sWing at 8:28 AM on November 19, 2019

> it looks like a horrible sucking wound

It looks like a very clean weld to me, apart from the tailings being kicked up at the edges of the shoulder.

I think Mr.Encyclopedia is referring to the hole left behind by the stirring pin, which is visible in the upper middle of that square of weld for less than half a second at 3:31, just before they superimpose the logo.
posted by hanov3r at 8:31 AM on November 19, 2019 [1 favorite]

I was also wanting to see the tool extraction point. The one "sucks" part of an otherwise elegant process.
posted by Meatbomb at 8:32 AM on November 19, 2019

The exit wound is substantial, but in production I'm guessing you can just use a runoff tab, as you would for normal welding.
posted by aramaic at 8:37 AM on November 19, 2019 [4 favorites]

Here's a slightly more low-tech demonstration by everybody's favorite foul mouthed Canadian, AvE.
posted by TheCoug at 8:53 AM on November 19, 2019 [3 favorites]

SONUVA DIDDLY
posted by hanov3r at 9:08 AM on November 19, 2019 [3 favorites]

There's smoke or some kind of fume coming off the plate near the end of the video, so even if the metal isn't melting, it's getting pretty hot and emitting something.

Probably residual oil on the surface.

I'm fascinated by friction machine operations. I don't know how long friction welding has been a thing, but it seems obvious if you've seen a CNC mill crash into the table while it's running. Tool/spindle and vise/workpiece become one of you stop it and allow it to cool.

In more controlled situations, you get welding or burnishing with very good results.
posted by 2N2222 at 9:25 AM on November 19, 2019

One of the main reasons why they use friction stir welding is because it does not require heating the work to oxidation state temperatures and does not require the use of filler metal as found in wire or stick welding.

This directly also helps with workplace safety because you're not getting fuming hot metals as well as flux and shielding gases from coated stick electrodes or other hot/arc welding processes. There's also no danger from arc flashes.

You don't even need flux or shielding because the temps are low enough.

This eliminates a major portion of most of the dangers and risks inherent in almost all hot/arc welding techniques, as well as helps aid with quality control as well.

There's also no filler metal which means lighter, stronger welds.

There's a reason why one of the first primary uses for stir welding was on spacecraft and spacecraft parts. It's a much lighter and stronger weld.
posted by loquacious at 11:12 AM on November 19, 2019 [3 favorites]

I sorta thought this was like Cold welding. Similar metals will fuse together. The thing that keeps this from happening is that there's an oxidation layer on the outside (because metals oxidize). If you get enough of the un-oxidized metal mixed up, there's just no difference from a solid piece of metal. They ran into this on an early space mission where the hinge of a hatch was so polished and smooth that it got a bit stuck. There's a video somewhere of two pieces of gold under a microscope basically oozing together like amoebas to form a single piece of metal. Because, you see... metal atoms placed oh so close to other metal atoms just clump up into that same solid metallic structure. At least once you get past the surface oxidation and smoothness issues to actually manage to get the atoms from the two pieces suitably close enough.

At least that's how I think about it. Just using pressure and rotation to stir up and distribute the surface oxidization layers and force enough of the metal atoms into close enough direct contact so that the two pieces fuse into one.
posted by zengargoyle at 2:52 PM on November 19, 2019

Listening to this video made me feel like I was starting to understand how the turbo encabulator works.
posted by avapoet at 4:34 AM on November 20, 2019