Welcome! Log In Create A New Profile


SLS powder printer workthread

Posted by peter_ 
SLS powder printer workthread
April 03, 2010 12:43AM
Hi folks,

I thought I'd make a workthread for interesting information related to the SLS powder printer that I've been describing in a few posts to the builder's blog recently -- the forum post is probably a bit better way to communciate, rather than the comment system on the blog. smiling smiley

Here's the latest link:

selective laser sintering printer: part 4 (putting it all together)


first tests building a selective laser sintering printer: part 3

first tests building a powder-based stereolithography printer: part 2

first tests building a powder-based stereolithography printer: part 1


Edited 1 time(s). Last edit at 04/03/2010 12:45AM by peter_.
Re: SLS powder printer workthread
May 19, 2010 03:40AM
Hey, I'm not the only one trying to pull this off after all!

I would like to compare notes with your progress, perhaps we can work together to figure this one out. I have done allot of reading on how SLS works from the types of laser needed, the temperatures needed for melting points on my materials, and even possibly building my own high powered CO2 laser.

It doesn't look like there is really any cost savings in building your own anymore, as their are plenty of cheap Laser tubes on eBay that are being built in China for dirt cheap (for lasers anyway).

I researched the patents that were published on many different SLS printers currently being used, and it looks like 50 - 100 watts is the norm. I even read a few other early on printers that used 20 - 30 watt lasers, and those supported printing in most mediums like ABS plastic.

You can find these tubes on eBay for around $200. Then you have to build your power supply, I read that you can get these from neon lights on the cheap. You also need to add the proper mirrors, and lenses, as well as a cooling method which is normally water. I am estimating that I can build (safely) a commercial grade CO2 laser for right around $500.

That may not be a cheap project for some seeking 3D print, but for me, it is well worth the price for the benefit of owning an SLS printer in my home. Not to mention how much these run commercial anyhow.

Now, I am guessing that 90% of this project is going to be pretty straight forward. We already have a working model that does what we need with CandyFab. Were basically just using the design, and swapping out the heating element for a Laser for Sintering. My biggest concern is configuring the software to talk to the controller for our laser, and knowing when it needs to be turned on and off.

Re: SLS powder printer workthread
May 19, 2010 03:59AM
... i made some tests with sintering powder with a 50Watts fiberlaser (1070nm wavelength) and a 5Watt Diodelaser (975nm) - it works with dark/absorbing materials but you need a really thin layer or you'll receive single molten droplets sitting on top of some not molten particles not connecting the surface eye rolling smiley

The diode-lasers have a spot of maybe 0.2 to 0.1mm diameter, the fiberlaser of 0.02 ro 0.005mm.

With 2Watts power with the diode and <1Watt with the fiberlaser i melted powders from dark plastic, powder-shugar mixed with carbon black and 0.1mm big black spheres from molten rock.

Yesterday i found out that my local plastic-granules-company has a lot of thermoplastic colour pigments as dust, what's easy to apply in thin layers but tend to 'contaminate' all touched parts with the corresponding colour spinning smiley sticking its tongue out

I tested with a black pigment, it melted and settled to the base surface ... when solidified, it's hard and brittle, but i have to make some more tests with other pigments too to define material specifications ...

Re: SLS powder printer workthread
May 19, 2010 05:01AM
I am most attracted to the ABS Plastics, and found this [www.3dreplicators.com] easy walk through on chemically grinding ABS plastics down to a fine powder. Looks like this may be the cheapest route, and can likely be formed from scrap ABS parts if you have access to any.

While reading through patents today for anything related to SLS, I found an SLS concept printer that was file in 2005 to create the same process at much cheaper costs and less print time. This one was a really good read, so I am attaching it to this post.

For those that don't know, SLS requires you to heat your printed pieces as the finalization process to strengthen the bond between the sintered layers. This patent has a second pass on the sintered areas to apply a quick dry binder glue before covering it with the next layer of powder. This allows the piece to have strength upon print completion, and removes many hours of curing, and the cost of a high temp curing oven.

They also listed the very simple idea of adding color to each layer with another pass before the sinter that sprays the specified area with ink jet cartridges. I am not sure if this is the current method of adding color to SLS prints, but that seems easy and cheap enough for me!

Being a patent, this isn't exactly a blue-print to recreate any designs, but these patents do give some additional knowledge of how these systems work that can be pretty helpful if your in the mass knowledge search mode. I wish they listed the type of glue binder, and ink being considered, but it looks like those will require some trial and error if used.
open | download - 11_093_078_Method_and_apparatus_for_rapi.pdf (179.3 KB)
Re: SLS powder printer workthread
June 21, 2010 07:47PM
Hey guys, thought I might chime in. I'm an SLS Technician and can answer any questions you have (can even sort out some powder if you need it).

Laser power itself isn't nearly as important as the heating of the powder bed. I'll give you an example. Our nylon melts at around 188C. The machine itself heats the build area to 183C. The laser itself is only responsible for the last 5C temperature difference, just enough to push the temperature past the glass transition point. While its possible to do some LS work with just laser heating, it will be extremely prone to warping/curling/fouling etc. By doing most of the heating externally, you're also able to speed the laser up significantly.

I've done successful SLS Nylon-11 prints with Laser Powers down in the 3-4 watt range, at the same speed that I usually use 12 watts at - around 10m/s laser travel.

stlouistechy - You're on the right track, but your scraps of knowledge are from all over the place tongue sticking out smiley . Most SLS systems that are only doing plastics only require 20 watt lasers at a maximum (10 watts in usage is normal), and they don't require a post curing step (you may be thinking of SLM (the metal SLS)).

Any questions at all, please don't hesitate to ask. I love talking about the tech.
As Proto states, pre heating is critical of the build chamber, I personally would stay away from lasers and have a
print head printing black ink onto the powder and IR lamp to fuse the powder, no need for nitrogen gas and can
use lower grade powder, and its faster.
google high speed sintering. or look here [www.lboro.ac.uk]
Re: SLS powder printer workthread
June 22, 2010 06:02AM
... it's the desired resolution against complexity of beamshaping methods whats interesting here.

With lamps, blinds and lenses i'll receive a spot of maybe 1mm diameter or a parallelized beam of 50mm diameter and a blind or beamer-optics for the cured zones, where the resolution can be something between 0.3 and 0.05mm - the finer, the expensiver ...

With a laser i have smaller spots - 0.2 to 0.1mm with the 5Watt-diodes i'm working on, or 0.02 to 0.005mm with my fiberlaser ... so with the laser i can sinter much finer structures ...

Yes lasers can have high definition due to the small spot size, but the drawback is that with such a small spot,
rastering can take longer, unless some sort of optics are employed similar to sla where they draw the contours with a small spot and raster the internals with a larger spot.
The resolution of high speed sintering is mainly down to the print head resolution, and the benefits of hss are cheapness as well as speed, but lasers give sharper detail and open up the possibility of sintering metals etc, maybe a combination of both??????.

P.S I have seen some work done on shaped or tuned lenses that greatly improve the melt pool in polymers and metals.

PPs. I am just finishing building a baird mechanical television, I love this old technology, can't help thinking there must be
a way to use the nipkow disk for 3d printing.

Re: SLS powder printer workthread
June 24, 2010 04:31AM
Hi John,

... as i use a CNC-mill for my paste-dispenser and the laserwork, i can combine all of the technologies - imagine the paste-dispenser or FFF-extruder outputting a 'coarse' sheet, the millhead or lasermodule finishing the contours and heighth to lets say 0.01mm accuracy and an aditional powder-dispenser and the laser as sintering/melting device embed some metal-structures inside the plastic- (FFF-Extruder) or ceramic- (Paste-dispenser) parts grinning smiley

This sounds like pure sci-fi or an excerpt from the early developing-stages of the 'enterprize-fabber', but i have all single modules there and working, what's more to do is the software for combining all the single methods in a homogenous fabbing-chain ...

Re: SLS powder printer workthread
June 26, 2010 11:17PM
I'm really liking these ideas, and it's great to have so many knowledgable folks interested!

The IR-sintering with black ink printed onto a white surface sounds really interesting... any thoughts on where one might obtain a suitable lamp? the powder printer itself would likely have to be built out of something that could withstand some heat, like aluminum or such, which would limit it's ability to be easily fabricated or to reproduce... but it's still a very interesting idea!

For a first SLS printer, I would be happy with 0.5-0.25mm resolution, and 0.1mm would make me extremely happy. (I can't imagine a resolution measured in microns for some time, Viktor... it's best to start tractibly and make one's mistakes cheeply. smiling smiley ).

I'm also very much in favor of hybrid designs -- things for which there is the potential to print using more than one material, where (hopefully) at least one material is usefully conductive. It seems likely that for a given part, the bulk of the part might be made out of something structural (like ABS thermoplastic), while there might occasionally be other material (like a confuctive paste) deposited.

The inkjet front also has me excited, since Adrian announced it recently. While it seems like printing solid parts might take a little longer than on a conventional reprap (at least at the start), the idea of potentially printing (flexible?), multi-layer circuits fairly quickly (they are, after all, typically rathr thin) once folks had usefully conductive materials worked out really, really excites me. Instead of using a board mill, one could simply print out their board...

But back on topic... any idea where to get a few lbs of inexpensive powder material? glass filled nylon, abs powder, anything? The method that ends up winning out (SLS, IR SLS / inkjet hybrid) might end up depending on what material is easily available...

Re: SLS powder printer workthread
June 26, 2010 11:49PM
But back on topic... any idea where to get a few lbs of inexpensive powder material? glass filled nylon, abs powder, anything? The method that ends up winning out (SLS, IR SLS / inkjet hybrid) might end up depending on what material is easily available...

Plaster-water is a good test-substance. Cheap, easy to source, and non-toxic.

-Sebastien, RepRap.org library gnome.

Remember, you're all RepRap developers (once you've joined the super-secret developer mailing list), and the wiki, RepRap.org, [reprap.org] is for everyone and everything! grinning smiley
One possible source of heat lamps could be halogen, oven hob, patio heater or the sort that fit into small enclosures that look like floodlights but are heaters, Viktor would be best to advise if this is a suitable wavelength, I do know that they pulse them when used for HSS, but may not be necessary for initial trials.

I use a Zcorp at work and the parts are good for such as architects to look at, but not much use as parts for engineering use,without a great deal of post processing or secondary applications, polymers have much broader applications and as it is important to have the powder at an elevated temperature whichever way it is fused, so will need a machine to withstand elevated temperature I think.

Getting the powder is a problem at present, I do know that it was, and may still be a by product of some cleaning process,
much like flyash is a by product of cleaning coal power stations. Virgin material is mixed with used material until it has to be disposed of, this used to be a problem and may still be for commercial fab shops, but getting small quantities will be difficult,
One possible source may be rotary moulding, they use powders in moulds which, (you guessed) rotate to make large tanks etc, powder coating may be worth looking at but have never handled the powders they use, so may be barking up the wrong tree.

I think you make an important point about materials, its not much use building a system if the materials are hard to come by or are very expensive.

Hope I have helped and not muddied the waters
Re: SLS powder printer workthread
June 28, 2010 04:21AM
... for IR-sintering you can use an highpower aerial IR-lamp from distance or NiCr-wires floating close over the surface.

You have to apply enough heat for melting the topmost surface of the powder, so it melts to the solid object too.

I found two glass-tubes with embedded NiCr-wire in a toaster, which should be powered in series by the powerline (230VAC).

The tubes are maybe 10mm in diameter and 200mm in length, when powered, they glows orange-red and emits enough heat for melting plastic in distance up to 40mm - but you can reproduce this with naked NiCr-wire too, only the lifetime of the open NiCr-wire will be limited in respect to the embedded wire in the glass-tubes ...

So moving one or two tubes (or red glowing NiCr-wires) over the surface with some mm per sec. should be enough to melt/sinter the surface ...

Your machine sounds absolutely brilliant, bringing together all those aspects into one 3d printer will
mean you can build just about anything, I think you are way ahead of the rest of us, please don't leave us behind.
Re: SLS powder printer workthread
June 28, 2010 06:53AM
Hi John,

... its true, i have all this brilliant bits/tools here, waiting to be used ... but beside my dayjob and family-business simply not enough time bringing this to a short end eye rolling smiley

Im working on 3D-RP and associated themes (with some delays and sidetracks) since 1986 or so, but the last 3 years its something speeding up ... so there is some hope grinning smiley

Re: SLS powder printer workthread
June 28, 2010 10:21PM
I could probably set you guys up with a couple of different nylon's and nylon mixes if required. But I'm not sure they'd be ideal. They work well in SLS machines because SLS machines are designed to use them, and a DIY application may just amplify the inherent difficulties of the process (warping and curling).

The most important first step in my opinion is engineering a suitable chamber for this to all take place in; I can't stress enough how important the external heating is in order to control curling and warping. This could be as simple as an electric oven coil over the top of your teeny little round build chamber. But its critical that this is extremely well monitored and ideally set up on a duty cycle to hold that temperature. When this heating arrangement is engineered to the point where you can glaze your material on the part bed surface completely evenly, and then drop back a couple of degrees to a point just shy of glazing, then it'll be ready to implement some of these IR-heatlamp or NiCr wire sintering systems.
Re: SLS powder printer workthread
June 29, 2010 03:41AM
... i have two temperature controllers from Watlow which can drive any current until 25 or. 50 Amps and PID-controlling the end-temp with +/-1 deg. or better, so an ambient temperature created by a heated bed and/or additional heaters at the surrounding walls/ceiling around the building area aren't a problem.

The only complexity comes from the heat-tolerance of the electric/electronic parts in the enclosure - here i have to add some coolant to prevent overheating ...

I know this is sort of of topic but......
Some years ago I visited rollsroyce and they had a large electron beam welding plant, the welds were in the region
of 100 mm deep in some exotic materials, the vacuum chamber was the size of a small room, ever since that visit
I have liked the idea of using electron beam.
So.. what if we had a cylindrical chamber, small to start with, filled with nylon powder, and two electron guns set at an angle
so that where the beams intersect we get fusion of powder, not on the surface but deep within the powder,
It would need a vacuum pump but thats not difficult, and as the chamber is packed full of powder curling may not happen?,
the main question in my mind is, will the electron beams intersect, or because they are both negative and therefore repel,
will they diverge.
Another thing to throw in here..... When mist makers first came out, the ultrasonic ones, I played around with one, I discovered that at the focal point it generated enough energy to melt plastic, and cause cavities in silicone rubber about 8mm below the surface, by slowly moving the object by hand I was able to create a small long cavity within the silicone rubber block with no contact to the outer surface, this must be similar to those glass blocks that are laser ablated,

I know people will be saying don't talk about it, do it, but like Viktor I just have to many things to do, but someone may
have the time or inclination ???.

P.S a note to Sebastien, would it be worth having a section titled "What If" where people can share general ideas.
Re: SLS powder printer workthread
June 29, 2010 04:18AM
Hi John,

... the electrons will heat the surface of the powder and along the complete path too before intersecting, so a 'deep' engraving/heating as with a laserfocus in glass won't be possible eye rolling smiley

A vacuum-chamber and an electron-gun salvaged from an old tv would be posible, but the complexity is something more than with a diodelaser and the benefits aren't so clear confused smiley

I was thinking along the lines that the beam energy is such that it does not fuse the powder, but where the beams
intersect the energy is enough to fuse powder, if it could work it would allow the possibility for true 3d printing instead
of layer by layer?.
Re: SLS powder printer workthread
June 29, 2010 04:04PM
Hi John,

which material is transparent for electrons?

You can't use electron beams in air because of the absorption of the electrons on the fly ... and solid materials are even more dense, so the electron won't pass the surface eye rolling smiley

Hello Viktor
Not sure how to answer the question because I have not experimented with electron beams, and will not have time to do so in the near future, but maybe a low energy beam will effect the polymer powder or resin but not fully fuse it along the path of the beam, but may do at the intersection point?, leaving the powder or resin usable for the next build?.
some form of Tomography may be possible, like they use in cat scanners, or if it only works for layer by layer building
that would be a start.

Re: SLS powder printer workthread
June 30, 2010 03:29PM
... you need a really good vacuum for the process, whats highly complex and expensive in the desired dimensions, so sintering with diodelasers is much simpler and cheaper to realize ...

proto Wrote:
> I could probably set you guys up with a couple of
> different nylon's and nylon mixes if required.

i'm trying to make my own powder, for sls. I'm searching for recipes and sources.
Can you help?
Re: SLS powder printer workthread
January 16, 2013 06:46AM
Has anyone looked at using high power white LEDs? 10W are easy and very cheap to get, 50-100W are more expensive. The only real problem is that they do not produce a nice clean point, the light is spread over anything up to 140 degrees. So a collimator and lens would be needed. They also have the advantage to being visible and being multi-colour they might be easier to use on coloured powders.
Re: SLS powder printer workthread
January 16, 2013 08:04AM
... high-power LED's are multi-chip-emitters - i have some with 9Watts @405nm, that were composed of an array of 5x5 single emitters with a combined square size of maybe 5x5mm.

For focussing them on a spot of below 1mm diameter you'll need a lens-array, where 5x5 focussing lenses are positioned on the right positions to the single emitters, what's a really complex/fiddly adjusting!

I was involved in the development of the adjustment/assembly stage of 20Watt-IR-diodes, forming a ring-focus laser with 80 of this diodes combined to a single focus spot and combined 1600 Watts of power - this was a really costly process ...

Fiber-coupled single-emitter IR-diodes are much easier to use and already in a price-range achievable for DIY: - 9Watts on a spot of 0.1mm diameter for below 300 USD!

Sorry, only registered users may post in this forum.

Click here to login