Centrifugal laser sintering

Centrifugal laser sintering
founder: timthelion
depends on:
software license: GPL3
hardware license: GPL3

I believe that centrifugal laser sintering might be the future of rapid prototyping and short run productions. Centrifugal laser sintering is a laser sintering process involving a rotating bed with a slight vibration or tremor into which powder is poured, a laser mounted on a single axis to sinter that powder, and a rotating pick and place machine for small parts.

This type of laser sintering has a large number of advantages over flat bed laser sintering:

  1. You can do an automated, rough cleaning of the parts right in the machine simply by opening some holes and spinning the powder off.
  2. Since powder is leveled by centrifugal force rather than by a leveling blade, you can do pick and place of small parts directly onto the powder. This allows you, for example, when sintering a tin powder, to place electronic components onto the powder and immediately sinter them into place. They won't be knocked out of position by the leveling blade.

Basic approach

Each machine can handle a practically unlimited set of powder types. However, for normal operation it is necessary that the powders be separable in order to allow for their re-use:

  1. Build material which is heavier than water(metal).
  2. Support material which is soluble in water(salt).
  3. Build material which is lighter than water(plastic).
  4. Ferrous build material which can be separated magnetically.
  5. Any number of centrifugally and chemically separable materials.

The materials are poured into the centrifuge while it is spinning. The laser has the ability to preciscely measure the height of the material. When the material is the right height, we stop pouring the material, and start sintering one layer.

We repeat, until we get to a layer in which we would like to place a small pick and place component. Our pick and place arm takes the component to be placed, accelerates it until it is spinning at the same speed as the centrifuge, and then places it onto the powder(like going into orbit ^ _ ^). We then sinter the component in. Additional layers of powder then cover the component. Since the powder is leveled by the vibration and rotation of the centrifuge, there is no need to mechanically level the powder in a way which would displace the component.

Once we are done sintering, we open up a series of vents in the centrifuge(or we move an inner, mesh type container, horizontally away from a solid one) then we pour salt into the centrifuge to replace the build material. When we are confident that as much build material as possible has flowed out of the centrifuge, we close off all of the holes(including the center one) and slow the centrifuge to a stop.

Once the centrifuge has be stopped, we take this torus, salt filled container, and carefully submerge it in water. We now open up the torus container and take out the parts.

Waste powder is separated, by dissolution in water/siphoning etc. and re-used.


It would also be possible to have the pick and place components rotate along with the rest of the centrifuge. I think that this would simplify design, but it would be worse in the long run. Sometimes the tapes get stuck, and it would suck to have to stop the whole build just to fix a problem with the pick and place. Another option would be to put the pick and place setup on a rotating platform that rotated parallel to the sintering centrifuge. This may be the best of both worlds.

Leveling methods

Leveling methods are going to be a very important topic for centrifugal SLS. While the original idea was to use vibrational leveling, there are several problems with it. It is hard to produce, causes unwanted sinking of pick and placed components etc. The key point here, however, is that we cannot use a leveling blade if we want to pick and place components onto the powder or do multi-material sintering. Here are some alternative leveling methods:

  • Wind based leveling: Blow air at the dust in the centrifuge to make the particles move to their lowest positions.
  • Electrostatic leveling: Charge particles before dispensing them into the centrifuge, this will make it so that they push away from each other into a very even spray. (This could also be used to lift powder from the dispenser onto the centrifuge, which would simplify the process of opening/closing the powder dispenser. When you want more material, apply the electrostatic charge, when you have enough, turn it off.)
  • Capacitive leveling: Create a charge differential between a static plate which is suspended inside the centrifuge, and the centrifuges chamber. This will lift particles up slightly and maybe they would come to rest leveler than before.
  • Careful powder dispensing: Don't level at all, just be careful to dispense powder very evenly.

Building the thing

Centrifugal laser sinterers have many risks associated with them:

  1. Dust can/will get into the air, causing health concerns.
  2. Need for high voltage.
  3. Lasers

However, if we start very simple than it may be possible.

Start by attemping capacitive powder dispensing without doing any sintering

I would start out not trying to build a centrifugal laser sinterer at all. Rather, I would create a vacuum tight box, with our centrifuge in it, and a tray with powder in it. I would pump out all the air with a cheep vacuum pump(seems they go for 20-$50 on ebay). I would charge the powder tray to +20 000 volts or so, and the centrifuge to -20 000 volts and see if the powder doesn't migrate up to the centrifuge and stick (migrate, as the air molecules do as described in this video. Perhaps we'll need a much higher voltage than that, as we're trying to move solid matter and not just air.). (hopefully I can steel a voltage source from the tesla people when they aren't looking).

If our capacitive powder dispensing trick works at all, we would move on to trying to turn the dispenser on and off. Can we dispense some powder into the centrifuge and then take away the static charge and have the flow stop in some reasonably controlled fashion?

If not, we try some simpler dispensing method, like gravity, for instance.

If the capacitive powder dispensing trick does work, then GREAT! We've invented a way of dispensing powder with esentially no moving parts :D.

Once we get powder dispensing down, THEN we can think about lasers and sintering

It seems to me that the best bet is to get an 900nm infrared diode laser with a fiber optic cable pre-attached to it. CO2 is more powerful, cheeper, and works better, but has a semi-limited life-span and requires high voltages(which, unlike the case of our capacitive dispenser, actually have enough current to do damage). For directing the beam, I don't know yet. I think it would be nice to be able to use a spinning square of mirrors like in a laser printer. Yes, this requires having the laser off most of the time, so it'll slow down print time, and there is the optical issues as well, but if we do this right, then we won't have to have any of the nasty moving parts a rep rap has, with steper motor BS ;) .

Once we get a laser installed in our centrifuge, we will do a bunch of tests just burning black lines in paper to test the laser. We will need to create a good firmware with some kind of optical rotary encoder for the centrifuge. We'll also need the ability to measure how much powder has been dispensed into our centrifuge, probably we'll do this by doing some laser based distance measurements (another laser mounted fixed inside the centrifuge).

Known problems with this approach

Dust dust dust dust dusty dusty dust dust. Powder, better known as dust, is bad for small electronic components. It's a bit insane to try do laser sinter electronic components onto tin powder as that same tin powder may cause a short.

Centrifugal forces on powders, especially during the end stage where we remove the excess powder from the container, cause powder to fly through the air, creating heath concerns.

Centrifuges are often used for separating materials with varying densities. When we use multiple powder types, it may be hard to ensure that the laser sinters a layer of high density powder before that powder seeps through a previous layer of low density powder.

Similarly, when we pick and place components, it will be necessary that they are always placed atop of already sintered surfaces. If they are placed atop powder, they will sink into the powder. A traditional flat bed laser sinterer with just vibrational leveling might do better at this.

Multiple powder types pose a problem when the powders melt at radically different temperatures. This problem is intractable, but it doesn't prevent all use of multiple powder types.

Vibrations will cause ball bearings to wear out quickly and might create precision issues. (Perhaps one could use air flow rather than vibration for leveling).

Crazy Sci-Fi wet dreams

The year is 2025 and the MaaS(Manufacturing as a Service) revolution is in full screen. You bring up a website and upload your cad file. A price is instantly calculated. You enter the quantity of items you want, and you press print. Five minutes after that a law enforcement drone arrives at your location to arrest you for copyright infringement.

In a long white walled factory outside of Kladno, a 4 meter tall centrifuge begins to spin up. This is a state of the art machine featuring:

  • Multiple lasers to ensure fast build times.
  • Dozens of powder types.
  • A multitape pick and place system with automatic tape transfer, allowing you to choose from any of the thousands of electronic components in stock.
  • Fabric, thread and wire laying mechanisms allowing you to produce strong, light, composite materials.
  • Reprap like FDM extruders with multiple material types, making it so that you can produce items with hollow, air tight, cavities.

The outer compartment of the centrifuge is on a rail. Once the build is finished, it slides out of the build area. Dust sprays out of the tiny holes in its sides, and is replaced by salt dust. Then the compartment slides further along its rail, into an outer torus envelop. This envelop is then rotated, so that the centrifuge is horizontal rather than vertical, and is lowed into a water basin. The salt is dissolved, and the torus is opened. A robotic arm lifts your assembly out onto a drying line.

24 hours after the print command is issued a UAV is at your door with the finished assembly. This is the first time your assembly comes in contact with a human being. Due to high competition in the area, and low overall costs, the price of the item is less than if you had bought it in a store.

Crazy Pol-Sci wet dreams

The samething as the Sci-Fi but the machine is at home, can self replicate, everyone has several and it uses infinitely recyclable materials which can be recycled at home so that there is no need to even buy powder.

Except where otherwise noted, content on this wiki is licensed under the following license: CC Attribution-Noncommercial-Share Alike 4.0 International
Recent changes RSS feed Donate Powered by PHP Valid XHTML 1.0 Valid CSS Driven by DokuWiki