Secrets of Desktop Metal’s Upcoming Production Machine

By on May 21st, 2018 in printer

Tags: , ,

 A mock up of Desktop Metal's upcoming production 3D metal printer
A mock up of Desktop Metal’s upcoming production 3D metal printer

While the eyes of the 3D printing world may be on Desktop Metal’s Studio system, it’s their production system that is most interesting. 

Their Studio system was (well, is) revolutionary as it provides a method of performing 3D metal printing at vastly less cost and trouble than traditional powder-bed 3D metal printing. They achieved this by leveraging the well-understood filament extrusion processes and adapting them to metal in a two-step process that involves a post-print sintering stage. 

And to make it most appealing to buyers is that they’ve been able to bundle all required equipment and materials together to provide a fully functional environment for near immediate use in almost any office or workshop setting. 

But behind the buzz about the Studio system, they’ve also been slowly releasing more details of their upcoming production system. We were finally able to discuss this system with one of their representatives to find out more. 

The first thing to know is that the production system is entirely different from the Studio system; it uses a completely different 3D printing process, yet will still focus on speed and part cost. In the cost sense, the production system will hold a similar objective to the Studio system, but that’s about all that’s the same. 

Instead of an extrusion process, the production machine will use a powder process involving a liquid binder selectively jetted on the powder bed surface. The binder will hold the loose powder together into the desired shapes, whereupon they are removed and post processed into completely solid metal parts. 

The post processing includes three stages. First the part is de-powdered, as you will certainly find a quantity of loose powder stuck in crevices. This produces the “green” part, which still contains binder mixed with the metals. Next there is a thermal treatment that removes the solidified binder to create the “brown” part that is solely metal material. Finally, this brown part is placed in a sintering furnace to bind the particles together. 

That’s the basic sequence of events. But there are many twists along that path. Let’s review some findings.

The powder being used is not a specialized (and expensive) metal powder designed for 3D printing use. No, instead it is simply standard industrial metal injection molding (MIM) powder. This stuff is available at very low commodity pricing almost everywhere. Pricing in the range of USD$10-60 per kg is typical, and this is an order of magnitude less expensive than specialty 3D metal printing powders. Even better, it’s available in a very wide variety of metals and alloys, perhaps numbering in the thousands. 

The production system will use a “cart system” for moving powder and build chambers around the workshop. These carts will be sealed with inert gas to avoid corruption of the materials and for safety reasons, too. Once depowdered, the prints, which are built on trays, are removed for further processing. 

Depowdering is done automatically and manually. Desktop Metal says around 90% of the loose powder can be removed automatically with their system, but the remaining 10% can be picked off manually with their depowdering station. 

Excess powder is recovered, as is commonly done in 3D metal printing systems. But in the Desktop Metal concept there is a difference: the required ratio of reused to fresh powder is 80%! This is very different from most other 3D metal printing systems that demand more than 50% fresh powder on each job run. 

The Desktop Metal Studio System includes an innovative furnace that is suitable for placement in an office. Will the same furnace be used for the production system? It turns out, no, it will not. 

The production system will use its own furnace system, likely because of the need to accommodate the production system’s print trays. One reason for this is that their probable customers most likely already have a suitable furnace on site. If not, there are plenty of good options available. 

The speed of this system is interesting. They say they will be able to complete print a complete build chamber (330mm on each side) in about four hours. That’s spectacularly fast compared to even fast plastic 3D printers, and should prove very useful for those producing quantities of objects. One of the reasons for the speed is that the binder jet arm prints in both directions, as it swings back and forth. 

But that’s just the 3D printing operation itself, and the complete lifecycle includes the other steps mentioned above. Desktop Metal explains that the entire sequence from start to finish will take about 16 hours, which includes cool down. That’s far less than one day to produce many custom made metal parts!

Desktop Metal explains they’re hoping for achieving low volume runs of between 100-1000 units and a costing of about USD$0.20 per cubic centimeter of material. This price point should open up 3D metal printing is quite a number of new industries that up to now have not considered 3D printing due to the costs. It seems that the automotive, aerospace and medical industries will be joined by others in the future. Desktop Metal says the system could potentially be certified for use in those industries, but it’s just that they are pursuing different markets first. 

The production system will be quite intelligent, beyond untying I had anticipated. It is able to adapt and compensate for detected issues. The system will image each printed layer and compare it to the expectation to immediately see if anything is amiss. If so, they are able to compensate to some degree by dynamically cleaning the nozzles and using alternates. 

Installation of the machine will require a suitable workshop environment, equipped with 480V three phase power. You won’t be seeing a Desktop Metal production system in your office anytime soon. 

Currently Desktop Metal is only showing off an alpha mock up of the device, but the hope is to issue beta units in the first quarter of 2019. The cost of this system? I’m not entirely certain, but I suspect the all-in cost would be in the USD$1.0-1.5M range. 

Via Desktop Metal

By Kerry Stevenson

Kerry Stevenson, aka "General Fabb" has written over 8,000 stories on 3D printing at Fabbaloo since he launched the venture in 2007, with an intention to promote and grow the incredible technology of 3D printing across the world. So far, it seems to be working!