http://www.msm.cam.ac.uk/phase...1.MPG

Says they're welding magnesium plate. I wonder about the penetration.

 

That is pretty sweet, that machine probably cost more than my house.

 

Agreed. It sure doesn't look like it's welding though, and all the way through?

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by kb58 &raquo;</TD></TR><TR><TD CLASS="quote">Agreed. It sure doesn't look like it's welding though, and all the way through?</TD></TR></TABLE>
That's one of the advantages to friction welding. It bonds at the surface. If you have 2 disks and spin one on the other so fast that it melts them together, there's no reason to get anything more than the surface penetrated. In other words, why do you need penetration in something like this?

If you had 2 disks and wanted a full weld across the surface of them both, and had nothing but a tig welder to do it with, imagine how you'd do that. Friction welding is good for when you have large surfaces that you want bonded together.

That said, I don't think that the application in the video is one which friction stir welding is best suited...

But, notice when the weld stops, it leaves a hole? This hole is going to be very close to the depth of penetration. This machine will have small pieces that protrude down into the metal so that it is actually stirring the two metals, not just rubbing on the surface.

 

Agreed. I was assuming since they were welding the two large plates together, full penetration is required. And yeah that's one heck of a "crater" at the end there.

 

Actually, friction stir welding is greatly superior to other types of welding in that it does not actually melt the metal, unlike electric or gas welding. Rather, it cold forms the joint, mixing the metal around the weld like dough. This serves to strengthen the metal at the joint, much in the same way rolling or forging does, meaning the weld is actually the strongest part of the piece, so that in a friction stir welded sample, failure will occur in the base metal and not the joint. Now, if they could only make it affordable...

 

If you look into some different welding symbols, some call for full penetration and some don't. For example, a fillet weld doesn't ever call for full penetration, at least in my years of experience, I've never seen it.

Here's a couple links for helpful welding symbols. You may not need this info, but you may find it interesting.

http://www.gowelding.com/weld/symbol/symbol.htm

This link is from AWS, and includes all the most common symbols and more.
http://www.aws.org/technical/errata/A2.4errata.pdf

There's many cases where the stresses on a part are not such that full penetration is not needed. Quality is not always an issue of doing the best that can be done...it can be more of a matter of satisfying the requirements of the drawing. I have had cases in which I knew a part was drawn wrong and not as the engineering staff intended it...but after arguing with them over so many other cases, I decided to make it as they said and let the part prove them wrong. They don't like it when their mistakes are only brought out after you make the item...but if you call them out prior to making them, they often want to argue, and it serves as little motivation to change their ways in the future since they can cover up the mistake easily before the part is made.

Also, look around on your car, and other items you encounter every day. Very few items will be welded with full penetration. Anything done by ASME Section IV code that is a butt weld or seam weld will be full penetration, as this is the Boiler and Pressure Vessel Code. These things need all the strength of the base material.

 

Interesting reading, but this technique is not intended for everyday uses. The intent is really for aerospace and other high-tech, high-dollar, high-performance uses where every available ounce is pared out of the design and every member is pushed to the limits of allowable stresses. I doubt you will ever see FSW machines making I-beams or putting cars together.

 

The link appears to be down...

 

one big advantage to this type of welding is that it can be used to successfully weld otherwise "non bondable" materials. You can weld virtually any two metals together with this process.

If any of you are big into welding and learning new technology, check into an AWS membership. It's $75 a year, and includes subscription to the Welding Journal. You will also be able to go to monthly meetings (usually in your area) where you can meet people, go on plant tours, and learn about things done locally.

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by Engloid &raquo;</TD></TR><TR><TD CLASS="quote">one big advantage to this type of welding is that it can be used to successfully weld otherwise "non bondable" materials. You can weld virtually any two metals together with this process.</TD></TR></TABLE>

aluminum + steel? Are the different melting temperatures the only reason that combination can't work?

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by kb58 &raquo;</TD></TR><TR><TD CLASS="quote">

aluminum + steel? Are the different melting temperatures the only reason that combination can't work?</TD></TR></TABLE>

I just watched a presentation from ESAB on friction stir welding. You can definetly weld mixed materials. Melting point doesn't matter, since you're not actually melting the materials. Does this mean ALL materials are ALL weldable to each other? I dunno, but you can mix materials... I think they had copper and aluminum welded to each other. I think you can also get decent strength out of hard to weld alloys such as 7075 aluminum, which isn't supposed to be welded for structural applications.

As far as the hole at the end, if you can run the machine off the end of the piece into scrap material, then there is no hole, or atleast the hole doesn't matter.