Spreading uniform thin layer of fresh metal powder

I looked and did not see anything posted on this subject yet, and I feel like I have just had a brain storm and wanted to see if it is already old news.

With the desire to have high precision parts I saw 5u spot size being mentioned. To get the full accuracy in the Z axis you will want each new layer to be not much thicker than 5 microns and very uniform. I imagine that drawing a squigie across to roll out the next layer of metal particals is likely to make a mess! I was thinking about how to control the dispersal of powder as the leveler moves across and back. Since the bottom of the powder box/build area is being lowered every time, you know exactly how much new powder is needed across the whole build surface. And it should not change much from the center of the build out to the edges, so you don't have to modify the dispersal based on the currently printing part.

My first thought was to have an auger or roller system that would spread out the powder into a thinner and thinner, but hopefully uniform layer. But the more I thought about it the more that it seemed almost as unlikely as the squeegee. So then I thought about other ways to cover the whole surface with a very thin but very uniform layer of power. And that is when I thought that spraying a 'mist' of powder from a number of nozzles well above the work surface to let them settle. But even that would tend to be thicker in the middle, thinner at the edges. And then I remembered how very uniform thicknesses of paint are applied, using an electrostatic powder coat paint approach. Since the bed is either un-sintered metal powder or sintered metal object, it will be easy to apply a uniform voltage across the whole surface. Typically in powder coat painting the object being 'painted' is grounded and the power electrified. The approach works well with insulating powder as long as the target is conductive. But even with a conductive powder, if it is all heavily negatively charged when sprayed out it will spread apart from electrostatic repulsion, then be drawn to the comparatively positively charged bed so that it should spread out and be very uniform. You will still want to draw a leveling bar across to prevent random buildup of high or low spots, but now it is just intended to move a tiny amount of powder if any at all. Normally spraying all that powder around would create a health hazard, but we are already talking about sealed build area with inert gas fill so spraying out the powder and negatively charging it should not cause any additional problems. You may want to wait a little longer for all the powder to settle before sintering the next layer, but a draw bar used to shovel out the new material then back again to level it will have to go slow or make multiple passes to get such a thin layer to come out as even so that should not really slow the build process down much and it should improve the accuracy of the Z axis build.

Mike

Just to clarify, the chamber will contain a vacuum rather than an inert gas, but I guess a mist could still be created with a strong enough charge.

Thanks for the clarification Kempy. I had read through some of the threads, but not all. Last I had read there was still a desire for an inert gas environement, using argon, nitrogen, or hydrogen, with a preference for the later two since they can be generated on site.

Do any of the current professional machines work in a vacuum? I wonder about the added strength of the chamber. And especially that any static charge buildup on the powder will cause them to wander farther afield, get stuck in all kinds of strange places, and coat the interior by the time you finish a piece, even without the charged powder layering approach. And considering how strongly the particles are drawn to grounded potential, the powder coat paint approach might require putting a positive charge on the bed so that it will draw the particles away from the rest of the interior. Or the rest of the interior might be negatively charged as well to repel the charged particles. It would also be a good idea to find out which polarity the tribo charge build up from rubbing together or rubbing against parts of the printer. That way the same polarity can be used for the powder coat electrostatics, and if the rest of the interior is similarly charged, it will minimize powder particles getting into everything.

Mike

From my understanding, the existing Arcam EMB prints inside of a vacuum. Also, the bed will be at a potential 63KV above that of the cathode.