Ten years ago the concept of having on our desks an affordable 3D printer knocking out high quality reproducible prints, with sub-mm accuracy, in a wide range of colours and material properties would be the would be just a dream. But now, it is reality. The machines that are now so ubiquitous for us hackers, are largely operating with the FDM principle of shooting molten plastic out of a moving nozzle, but they’re not the only game in town. A technique that has also being around for donkeys’ years is SLS or Selective Laser Sintering, but machines of this type are big, heavy and expensive. However, getting one of those in your own ‘shop now is looking a little less like a dream and more of a reality, with the SLS4All project by [Tomas Starek] over on hackaday.io.
[Tomas] has been busy over the past year, working on the design of his machine and is now almost done with the building and testing of the hardware side. SLS printing works by using a roller to transfer a layer of powdered material over the print surface, and then steering a medium-power laser beam over the surface in order to heat and bond the powder grains into a solid mass. Then, the bed is lowered a little, and the process repeats. Heating of the bed, powder and surrounding air is critical, as is moisture control, plus keeping that laser beam shape consistent over the full bed area is a bit tricky as well. These are all hurdles [Tomas] has to overcome, but the test machine is completed and is in a good place to start this process control optimisation fun.
Hardware-wise, the frame is the usual aluminium extrusion and 3D printed affair, with solid aluminium plates all over the place where needed. Electronics are based around a Raspberry Pi (running Klipper) with a BigTreeTech 1.4 turbo mainboard handling the interfacing. The 5W blue laser is steered over the powder surface using a pair of galvanometers, which sounds easier to get right than it will be — we fully expect there to be some ‘fun’ to control the spot size and shape as well as ensure that it stays consistent over the full area of the build surface. Definitely fun times, and fingers crossed that [Tomas] irons out the details and gets some good prints out of it soon!
Those who’ve been around here a while may remember we covered the OpenSLS project a while ago, and whilst we’re on the subject of 3D printing in alternative ways to FDM, here’s a little something about printing with metal, so long as you’re plenty patient!
If only a 5W laser is needed, I would just mount the laser head on an XY stage and call it a day. Galvanometers and mirrors are just a never ending headache, constantly drifting out of adjustment, degrading, etc. I’ll happily accept slower XY motion if I never have to adjust laser mirrors ever again!
These systems tend to heat the whole lot to very very close to melting/sintering temp so you do only need a little bit of added energy to make it happen – a small laser is quite doable, but perhaps not a small enough one to avoid ever having to play with its focusing optics or the mirror to shoot it on target…
I don’t think I’ve ever heard of one using something quite so weedy for SLS before, but its plausible (and lasers have the wonderful world of burst vs sustained power – is it a 5W laser on average but each pulse is 100000W or something equally daft but with a low duty cycle tanking its average output).
(Very worth pointing out no personal experience with these machines, I just read alot and wish I had the space to try making one, or I suppose both the space and money to buy one and try other projects that would be otherwise damn difficult, its a really cool technology)
The Sintratec Kit is using a 2.3W diode laser, and it is a CW laser, not pulsed, so it does work.
The big downsides to ~450nm diode lasers is that you need the dark colored carbon mixed powders to get the absorption, where most industrial systems are using long wave infrared CO2 lasers instead.
Indeed, thanks for the info. Don’t remember ever seeing such a weedy laser for the role before, but it is nice to see such things exist, even if its with extra functional limitations (which isn’t a surprise).
application for fiber laser?
I’d be interested in the as well. Build volume would be smaller, but print speeds would be much faster than a single diode moving in x or y.
Fiber lasers does work, as long as you only use the carbon pigmented dark powders to boost absorption. This is what Formlabs is doing in their Fuse 1.
You do want a fiber laser capable of CW emission though, the pulsed behavior of the engraving lasers are not wanted (it just vaporizes the plastic powder due to the low heat conductivity and short pulse times), and you want a relatively large spot size.
The Fuse 1 runs a 10W laser with a spot size of 200 microns. Most industrial systems with >30W lasers run spot sizes closer to 4-500 micron, to be able to deliver a relatively large amount of power but not vaporizing the polymer with too high local intensity.
The Dr D-Flo videos on SLS were a bit of an eye opener for me about the powder mess and health considerations. Not so keen to have one now.
Wow! Thomas is a super smart guy to do this! I’m so Happy that we have him fighting for us little guys. Heh
Saw a laser engraving setup that did X,Y on a gantry but used mirrors to allow the laser to be stationary. So the “print head” was just a moving mirror at 45 degrees and maybe a lens under it.
The Amada’s (and other industrial cutting lasers) have done this for decades; it’s not unique. Also, yes, there would be a collimating lens near the substrate/surface. A lens is required to direct the beams to a focal point for precision work. You could put the collimator further up, but I’d bet that without a completely evacuated guide box, rayleigh scattering would cause a drop in effective output power.