Hacker News new | past | comments | ask | show | jobs | submit login
New technique can weld “un-weldable” metals (osu.edu)
164 points by Oatseller on Oct 29, 2015 | hide | past | favorite | 39 comments



I learned recently of another surprising welding technique. It turns out that if you rub together the right kinds of wood at the right speed, you can actually weld wood to wood!

Friction Welding of Wood (text): http://ibois.epfl.ch/page-20697-en.html

Linear friction welding of wood (video): https://www.youtube.com/watch?v=X0k04hjdYuQ


That's what wood glue is for, it's strong as wood if your joint is tight. Personally, I love the look of box joints. Buy a chisel and try one, it's mesmerizing.


If you haven't already, check out this youtube channel: https://www.youtube.com/user/Matthiaswandel



Ah yes, reminds me of the night when I was binge watching Japanese woodworking videos. That dovetail joint, man.


Stronger than wood, in fact. A properly assembled glue joint will be the last thing to break on a wooden part. That reminds me that I should really get back to working on that desk I was building... months ago.


> Stronger than wood, in fact. A properly assembled glue joint will be the last thing to break on a wooden part.

With the caveat that depending on grain orientation w.r.t. the joint, wood movement[1] can very significantly weaken some glued joints to a tiny fraction of their original strength, in a surprisingly short period of time. IIRC, Matthias Wandel (aka the Woodgears.ca guy) referenced above has a nice example of a poor joint vs. glue design in one of his videos, but I forget which one. :-/

[1] Primarily due to expansion and contraction due to changes in ambient humidity.


Welding is also very strong, but it has limitations and weaknesses.


>new welding technique that consumes 80 percent less energy than a common welding technique, yet creates bonds that are 50 percent stronger.

I don't care at all about the energy required. It's peanuts anyhow compared to making steel. You can use less steel with better welds, because you don't need to compensate so much. "50% stronger than previous welding" doesn't really mean anything. "80% strength of the base metal" would be really cool.

>Within microseconds (millionths of a second), the foil vaporizes, and a burst of hot gas pushes two pieces of metal together at speeds approaching thousands of miles per hour.

Explosion welding isn't exactly new. They seem to have developed more accurate(?) or less energy incentive(?) way to explosion weld. The article doesn't say. And how loud is this?

>The technique is powerful enough to shape metal parts at the same time it welds them together

Probably this is disadvantage. Likely they need to make counter recess for the welds, and afterwards smooth things out with something.

This technology seems very cool. But the article is really bad. Anyhow, it's so interesting that I'd actually welcome OP to do this again.

Could I get the same results with very powerful laser peening equipment?


So it's electronically (rather than chemically) powered explosive welding? Neat. Unfortunately, the last picture on the article seems to suggest it causes a fair amount of material distortion, and doesn't weld at the centre of the joint, only the edges. Still, very cool - I'm not sure automotive is the right application area, but I'm sure it will be applications. Corrosion-resistant cladding for petrochemical applications, or maybe even in-orbit assembly comes to mind.


Ohio is smack in the middle of the rustbelt - Automotive is on their mind.

Mixed material cars are here [1]- weight kills fuel economy and perceived power (power to weight ratio).

My immediate thoughts turn to safety and industrial hygiene related to exploding metal. I wonder if this releases more or less metal vapor than other methods. How loud is this?

1. http://articles.sae.org/13986/ 2. Bonus: video on the process: https://www.youtube.com/watch?v=CB2QnbSfmw0

Disclosure - I work for GM, but not in this area; any opinions are solely my own.


Cool. They actually vaporize the foil at 5m18s for those in a hurry. It looks like you could use it in similar places to where spot welding it used in cars but to bond say steel-aluminium rather than steel-steel.


They sublime the aluminium directly, more details: https://iml.osu.edu/vaporizing-foil-actuator-tool-creating-l...


Judging from the video at https://iml.osu.edu/vfa-welding, the side diagram is very much NOT to scale; when they call it a standoff sheet, they're being very descriptive.


The fundamentals of this technique seem to be similar to blast welding, but more controlled. Blast welding is used in all sorts of high strength applications, like vault doors. But its expensive, requires big pieces, and ends up warping the material.

Material processing, specifically metals, often involves heating the metal to high heats, and then controlled cooling. The max temperate and rate of cooling and completely change a metal's properties. Pearlite in steel is the classic example. Welds basically remelt the material and negate any processing gains.

If this can be cheaply replicated, and can be adapted to different configurations, it will be revolutionary. But reading the article, it requires aluminum for vaporization, and looks to only bind two flat surfaces, where one's backside is easily accessible.


it's interesting how very close the margin in the micrographs they show looks to explosive welding.

The ones on page 2053 (warning, PDF, but it isn't 2000+ pages) http://www.jim.or.jp/journal/e/pdf3/47/08/2049.pdf are the ones I'm on about.


Exploding foil detonators have been used for decades in nuclear weapons. So in retrospect, it seems obvious. But hindsight is always 20/20.


How would you minimize galvanic corrosion with this technique, if it were used for cars?

edit: to be specific, this technique would seem to require that you prevent galvanic corrosion through surface coatings, but that doesn't seem to me to be a lasting solution for consumer automotive applications. It would seem like adhesives would be a superior choice since they can insulate the different metals from one another, reducing the need for the surface coating to maintain integrity.


Adhesives (e.g. ITW Plexus) are used in automotive, but your typical car has all kinds of galvanically wrong things going on inside. There are lots of techniques that have been developed to deal with the couples. Automotive industry have also developed more robust corrosion testing standards (e.g. SAE J2334).

Appeal to Authority: I don't work in the automotive industry, but I do use an automotive coating system for day job products that need high levels of corrosion protection.


Nice. That's very similar to the exploding foil initiator used to detonate nuclear weapons. The video doesn't give any of the electrical specifications, though.


So I guess we shouldn't try to use this technique to weld refined uranium to steel?


No, no, it's just used to initiate the conventional explosives to produce the implosion squeeze. All the detonators have to be triggered within about 10ns. Classic resistance-wire-in-primer detonators aren't that tight on timing. So the primer stage is omitted, and the main charge is detonated directly by dumping a capacitor bank into a foil. This is all explained in Wikipedia.


The precisely timed implosion lens is typically used for plutonium bombs, not uranium ones. A fairly bog-standard explosive can be used for uranium bombs (but plutonium is much more easily manufactured than sufficiently enriched uranium)


Plutonium is more easily manufactured than highly enriched Uranium if you're just starting out. If you have the industry set up, then it's easy peasy. In the developed world there's Uranium enrichment facilities everywhere and in nuclear powers they have HEU up to their ears.

Indeed, a lot of modern bombs (which are entirely implosion based) use HEU extensively in their design. Generally, they tend to still use Plutonium in the primary mostly because it saves on size and weight, which is hugely advantageous when it comes to long range ballistic missile delivery. But most compact sub-megaton thermonuclear weapons derive the majority of their yield from HEU in the secondary.


I found this webpage with a video of the weld in action:

https://iml.osu.edu/vfa-welding

Very cool stuff - I'd be interested to see some non-theoretical strength test data and more application-ready packaging.


This seems like just a more refined EXW (Explosion welding) technique, the 'ridge' weld shape is a dead giveaway. EXW has a major drawback though - it's useless for field word or anything not somewhat flat.


Yep, one of the images at the bottom shows that they're welding flat sheets of metal together by using a capacitor bank to vaporize a foil sheet to provide a plasma blast that smashes the target metals together. So, yes, more controlled explosive welding.


Yeah - in their presentation they point out the similarity, and explicitly talk up their achievement as explosive welding at the laboratory scale, without high explosives.


Could it be used in a tac weld situation. Basically they have some long armed machine that lets them tac an aluminium sheet down to a steal frame.


"Also, high-strength steel and aluminum join together with weld regions that are stronger than the base metals."

Does that mean they can create a new "alloy" made entirely of this join?


I took that to mean that any alloying that occurs is only right on the margin of the two metals, the way brazing traditionally works, too. But if they're really planning on making titanium-copper mokume gane [1], I'm all in favor...

1: https://en.wikipedia.org/wiki/Mokume-gane


I hope you're right. That's what excited me when reading this too.

If the joints are stronger than both materials, we can make layered metals that are crazy strong, and selectively apply more layers only when needed (like carbon fiber).


How would that work? Say you have a 'rod', something like this:

    AxBxAxBxAxB
              |
              v
Where A and B are the two different metal, and the x's are the joints that are stronger than either A or B.

Now we apply force in the direction of my crude vector. Won't it just break at an A or a B section?

(I know nothing about all of this, just wondering what I'm missing)


The assumption is that the joint is not a single line, but rather extends into the metal of both A and B to either side, so the whole rod is X.

Option 2:

    AAAAAAAAAAAAAAAAA
    xxxxxxxxxxxxxxxxx
    BBBBBBBBBBBBBBBBB
                    |
                    v


Keep in mind that this is still very much a press release, so of course it manages to sensationalize.

These metal combinations are only "un-weldable" via fusion welding, other solid state welding techniques (like explosive welding, as others have noted) can accomplish this as well. Furthermore, looking at the papers on the technique [1][2] the increased weld strength has only been observed in a few combinations of alloys and isn't significantly greater than other collision based methods. While the energy and scale reductions this accomplishes are worthy of praise, geometry restrictions, intermetallic generation and fatigue behavior within these welds are still issues.

[1]http://www.sciencedirect.com/science/article/pii/S0924013613... [2]http://link.springer.com/article/10.1007%2Fs11661-014-2404-0


I assume this means we'll see more beautiful automobiles in a few years.


Reminds me of making clad via explosion welding: https://www.youtube.com/watch?v=2u51tJdRDK0


That same weld pattern from the article can be seen here in the explosion welding video (~7:50): https://youtu.be/2u51tJdRDK0?t=7m50s


Wonderful taxpayer funded research that will see limited diffusion into industry. So in effect, we all pay twice while innovation at the edges is denied access. Research shouldn't be "in business"




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: