DIY 3D printing in metal is a lot more complicated than we thought. And this video from [Metal Matters] shows two approaches, many many false starts, and finally, a glorious 78.9% success! (And it’s embedded below for your enjoyment.)
The first half of the video is dedicated to the work on a laser welding system that doesn’t pan out in the end at all. But the missteps are worth watching as well, and they hammer home the difficulties of melting metal reliably with nothing more than coherent light. Things like reflection, the difficulty of getting good process control cameras, and finally the whole thing slumping as multiple layers stack up on each other make this approach to 3D construction look nearly impossible.
Indeed, around halfway through the video, the focus shifts toward a metal-powder sintering machine, and this one is a success! Metal dust is deposited layer by layer, and fused with a totally different laser. The tricky bits here range from esoteric problems like making the laser fuse the metal dust without blasting it, to simple things like the geometry of the scraper that ensures even layer heights. And once you’ve got all that down, getting a good pattern down for 2D infill in metal is non-trivial.
A sweet half-scale metal Benchy emerges at the end, so why does [Metal Matters] call this a 78.9% success? Because that’s the density of the final print, and he is shooting for 100%. But we wouldn’t be so harsh. We’ve seen how far he’s come since the first machines, and this is a huge advance. We’re looking forward to the next video update in a year or two!
Thanks to [Jonas] for the tip!
Summary of the video: I did that and it did not work. Then I did that, and it did not work. Than I did that, and it did not work, too. I than did that, did not work. ||I did that, did not work.||
Then I printed a benchy. Did not work. Printed it again, did not work again. Printed it again, did not work again. Printed it again, did not work again. Printed it again, worked.
Real hardcore stuff.
the pacing was rough but I’m really excited that he posted an update after a year
I totally appreciated the detail in the failures.
It’s nice to know what at least one pretty smart person didn’t get to work. Doesn’t mean it’ll never work that way, but at least you can see where the hurdles are.
Obviously it should have been done the way that worked from the beginning, right? What are you trying to prove?
LOL 😂
Whereas everything you’ve ever done has worked first time?
It’s useful to see what doesn’t work and why / what the problems are.
What the hell is a benchy??
Not to worry. Took Edison a while to perfect manufacturing the light bulb in quantities
Wonder if putting it into a partial/full vacuum while printing would help
Scaled down a bit (looking at those fingers I think it might already be scaled down almost enough)…
Ultimate Monopoly player piece!
Thanks, now I want one..
I think “you’d be ahead of the game” using a lost wax (or lost PLA) and cast those pieces, over this finicky laser process.
I’ve long wondered if there is a route to hobbyist-level selective sintering by ditching lasers altogether. There are cheaper and simpler ways to melt stuff.
I’m thinking of, for example, those demos of modulating a row of gas jets using ultrasound; or a variation of the thermal print heads used for receipts; or an electric arc controlled by similar techniques.
If a raster SLS printer could be built with that sort of hardware, then there’s no limit to how cheap (and therefore accessible) it could get. With a steerable high-power laser, there will always be a fairly high floor.
With gas and thermal printing, it’s impractical if not impossible to reach the temperatures necessary to melt steel. As for electric arcs, they are finicky in their own way – they have a tendency to wander towards high points, and the melt pool size is proportional to the electrode height which means any decently precise print would require having the electrode very close to the workpiece (and therefore running the risk of getting contaminated by some of the powder).
Just thinking out loud here, but what about coating the electrode (which should be super fine at this point) with a thin layer of some material that does not degrade with arcing but does not allow things to get stuck to it? (thinking PTFE, idk how it degrades nor what the breakdown voltage would be, but something of the likes). Then, if done in a proper vacuum, you could have all the actual heating energy concentrated on the “tip” of the arc (there would not be an arc/spark as there would not be air molecules to heat). If I had a vacuum pump and enough time I would love to try something like that out. The achievable resolution would likely not be good, at all. But it is an interesting experiment.
Ptfe starts to degrade at food cooking temperatures. An arc will vaporize graphite, which otherwise might be suitably non stick. But I think you need an atmosphere to have an arc… and you need the passage of current to make heat…. hmmm
I don’t see why it would be impossible if you preheat the powder near the melting point. You’d just be tipping it over the edge to fuse.
LoL, I just forgot that you actually need a conducting path, else you would just have a big capacitor. I definitely need some sleep.
I do like how this is coming together, but id never have this in my home. With cnc, the chips already end up everywhere. This metal dust is not just intrinsically expensive, its also going to end up literally everywhere. And I mean everywhere… in your living room, bedroom, lungs, and so on… oh and it’s probably carcinogenic… I would honestly like to do metal 3d printing at home, but this is the main reason I’ll take a pass.