Behind the Scenes at Desktop Metal

- Is this metal 3D printing technology set to shake up the industry?

by

At the 1927 Solvay Conference, 17 eventual Nobel Prize winners including Marie Curie, Erwin Schrodinger, Niels Bohr and one Albert Einstein, met to discuss the outstanding preeminent problems in regards to the newly formulated quantum theory. It is this meeting of minds that conferences, like AMUG, are attempting to replicate to address the concerns affecting the additive manufacturing industry. 

The challenges facing additive manufacturing are simply too broad and threading the needle through from materials development to post-processing, with all of the stakeholders that involves, is nigh on impossible. However, what if one company focused on solving the whole process chain and brought in a team of experts to do so? On the back of AMUG, I visited one such company... 

Desktop Metal CEO, Ric Fulop, has assembled a team that wouldn't feel out of place debating quantum amongst Einstein's scientific realists and Bohr's instrumentalists. The difference being this team are pulling together to solve some of the most significant rate-limiting steps for metal 3D printing. Desktop Metal is going to make metal 3D printing a magnitude more affordable, faster and office friendly. 

"Our thesis is that metal 3D printers, today, are similar to 1970s punchcard computers," explains the Ric during a tour of Desktop Metal's HQ, 30 minutes outside of Boston, in Burlington, MA. "It is a technology that requires dedicated facilities like large argon gas tanks outside your premises, three-phase technologies that require you to invest $800k in the machine with $200,000 of post processing equipment.”

These problems had frustrated Ric in his previous company, A123 Systems - a lithium-ion battery company that was Boston’s biggest IPO in the past decade. In 2013, having been the first investor in a number of 3D related companies including ProtoLabs, Onshape, and Markforged, Ric stopped investing and started doing. He got together with some MIT professors, Who are among the world’s leading experts in materials science, metallurgy, advanced manufacturing and 3D printing and began brainstorming how they could make metal parts more efficiently. 

The fruits of that labour will be presented to the public for the first time at RAPID + TCT in Pittsburgh (May 8-11) but Ric and his team opened the doors to Team TCT for a sneak peak of the most hotly anticipated 3D technology I can recall. 

Hotly anticipated is, in some ways, an understatement. We have seen many a technology launch, but anyone with a keen eye on the market will have noted the $100 million in equity funding that Desktop Metal has received from prestigious investors like Google Ventures, BMW, Kleiner Perkins and GE. Shrouding the tech in secrecy has only added fuel to the flame. Until our visit, all we knew of Desktop Metal's technology is that it would be laserless. 

Get your FREE print subscription to TCT Magazine.

Exhibit at the UK's definitive and most influential 3D printing and additive manufacturing event, TCT 3Sixty.

Desktop Metal has, in fact, created a host of technologies (135+ patents in process) and has threaded the needle right through the eye of many of metal 3D printing's problems. There's a new material science that leverages the world's most voluminous metal powders, hand-removable supports, software that takes your part from orientation right through to specified metal microstructures, using machinery that can be utilised in an office environment including the heat sintering furnace. 

“A Huge Technology Change”

There are two 3D printing systems at launch, a desktop version, and a production system; both are enabled by a proprietary sintering technology that is the same for both systems.

"The microwave-enhanced sintering technology is a huge technology change," explains Ric. "This is what allows us to make this system office friendly. It has a reducing atmosphere that makes it possible to use low-cost powders like metal injection moulding (MIM) powders. Right now people have to use low oxygen powders with a narrow particle size distribution that are very expensive. By having a reducing system in the furnace, we can use conventional powder metallurgy materials like MIM, which is what many of the parts in your iPhone, computers, and cars are made with."

During the tour Ric introduced us to the majority of the team, reeling off a biography of every team member no matter how junior or senior. One such member was Mark Sowerbutts, a sintering expert running the furnace project, who started his career at Oak Ridge working on uranium oxide sintering for fuel pellets and armour piercing projectiles. 

"I have worked on demanding and exciting end products like the space shuttle main engines, but as a furnace itself, this is the most exciting," says Mark. "It has many demands that are challenging. Usually, they give you all the power you want, all the water you want and forklift it in, this has got to go in through a single door, plug into a wall and sinter parts that are tiny or large, parts that are complex or simple."

Considering a laser alone for a powder-bed fusion machine can cost hundreds of thousands of dollars, to reduce the cost of the technology we have seen something of a move back to the indirect printing of metals that do away with a laser. Users are expected to use processes developed for traditional production methods, and it is often unclear how much additional cost or real estate space this will take. 

Desktop Metal is clear from the start you need both the printer and the sintering furnace, and the pricing is upfront. For the desktop printer and the furnace, you will pay $120,000 for what the company calls the Desktop Metal Studio System. You can add more Studio printers to each furnace. The sintering system is fully automated, can reach temperatures of up to 1400°C, and, in a point worth repeating, can be utilised in an office environment without the need for external ventilation. 

"We are not just doing half the process," says Ric. "(With traditional sintering ovens) you have no control, and you need a dedicated metallurgist to make good parts. If you do not know how to sinter correctly your gas flows are wrong, your oxygen control is wrong, and your part is not fit for purpose.

"Our system has a metallurgist built into a box..." - Ric Fulop

"Our system has a metallurgist built into a box. We are taking some of the world's best metallurgists like Chris Schuh and the 14 PhD metallurgists and engineers that we have here and wrapping all that know-how into the hardware and the software, so you do not have to do any programming."

Two-pronged attack

At RAPID + TCT, Desktop Metal will show both the Studio and Production Systems in action, the Studio system will be shipping in September 2017 with the Production system following a year later. They are completely different technologies but use the same sintering technology that can be used to sinter parts from both machines. 

The Studio System is akin to FDM printing, albeit very top end FDM. It uses a patented Bound Metal Deposition (BMD) extrusion process to make accurate and repeatable parts, extruding Desktop Metal's media of MIM metal powders bound by a polymer mix - a chemistry developed in-house that allows parts to be printed in an office-friendly environment. It has a 50-micron layer resolution, automated bed levelling, and a heated build area. 

The Studio System is designed 'for engineers to engineer in the place they engineer.' It is both affordable and user-friendly and could be truly transformative in the manufacture of low-volume part runs. The Production System, however, is not after those low-volume runs, Desktop Metal believes this machine will totally transform the mass manufacture of metal parts. 

The Production System uses what the company is calling Single Pass Jetting (SPJ). Created in unison by co-founder and inventor of the binder jetting process, Ely Sachs whose work spawned Z-Corp and ExOne, together with Paul Hoisington, a scientist with over 100 patents in the field of inkjet printing.

Like the Studio system, it is enabled by a proprietary sintering technology. It uses metal powders together with Desktop Metal's binder and two full-width print bars containing over 32,000 jets, printing at a speed Desktop Metal say is over 100x faster than today's most common metal 3D printing systems. Ric says SPJ is similar to what HP is doing in plastics. 

"Our system can print at roughly 500 cubic inches per hour," says Ric. "In a laser-based machine with a similar build volume, you are looking at making twelve of these (Ric shows me a hockey puck sized part) a day. Because we can nest parts like SLS  and don't weld to a build plate, we can print 112 parts in one four and half hour run which is 560 parts a day."

With this level of output from the machine, the key to success is being able to process quickly. The printing mechanism is decoupled from the build box, which is encased in a cart that can be swapped and transferred for sintering. 

Both systems have another bonus for those wanting to print complex parts in metal, supports that are removable by hand. Desktop Metal has created and patented a unique ceramic release layer that sits in between supports. The ceramic is bound with polymers, and when sintered turns to sand, when supports are removed parts bear no witness marks.  

MIM-icking success

Core to Desktop Metal's philosophy is affordability, and with so many patented technologies one may think costs could begin to add up, but arguably the most important factor in Desktop Metal's whole process is a decision that was taken very early on, to use Metal Injection Molding (MIM) powders. 

"Why did we base our technology on MIM?" Ric asks. "In my last business I was a supplier to the car industry, we made batteries for hybrid vehicles and electric vehicles. We were growing up at around the same time as Tesla, in a Tesla electric car the battery packs have thousands of little cells whereas in a conventional electric vehicle has these big batteries. We were makers of the big batteries. I asked their CTO, JB Straubel, Elon (Musk, CEO) and some other guys in the company, 'Why are you building you electric vehicles with 18650 laptop cells?" They wanted to be at scale from a cost structure from day one, not wait until there was the volume in electric cars to bring the cost of batteries down, which would take two decades. 

"Their mission was, 'let's make this real,' so they started with a raw material that was already at scale. That philosophy stuck with us; we looked at current metal technologies, and the entire business is built around a custom metal powder that is very expensive." 

That philosophy stuck with us; we looked at current metal technologies, and the entire business is built around a custom metal powder that is very expensive - Ric Fulop

By basing the whole business around MIM materials and then designing a process that is compatible Desktop Metal has an economy of scale. The raw materials for MIM are roughly 80 percent cheaper than powder-bed fusion simply because the production of MIM powders is three orders of magnitude higher than powder-bed fusion powders and involves less processing to produce. Thanks to this prevalence of materials, Desktop Metal will be launching with 30 alloys, and there are over 200 metal materials compatible with its systems. 

"Added to this is the fact that there are 40 years of research on sintering of MIM materials, says Ric. "As a result, there are massive data sets on mechanical properties."

Closing the loop

The software ties all these features together and closes the loop from a CAD file to a finished part complete with the microstructures mentioned above. To do this Ric, brought in the tenth employee at Solidworks and founder of Xpress3D, which went on to be RedEye and then Stratasys Direct Manufacturing, one Rick Chin.

With a thorough demo, Rick Chin talked us through the software system that has taken inspiration from websites like Kayak with slider interfaces that allow the user to see individual payoffs and drawbacks of each orientation.

"We want to make it so that our users are efficient with the technology from day one," says the software guru Rick. "Whenever you are adopting a new technology there's this period of trial and error. In 3D printing that involves wasting material, until you build up an intuition about how your printer works. We are capturing that intuition in the software."

Whereas other 3D printing software merely takes into account the printer, Desktop Metal's software, which is all ran in the cloud for optimum performance, gives you the ability to understand your parts final mechanical properties once it comes out of the furnace. The software works in conjunction with all the hardware and is continuously feeding back from the printer and machines to improve the estimations given beforehand. 

Rick is one of the many stars that Ric Fulop has assembled like Ernest Solvay picking Einstein and co. for his 1927 conference. We haven't even begun to mention people like; Yet-Ming Chiang, an MIT professor and co-founder; A. John Hart, another co-founder and MIT professor; CTO Jonah Myerberg, who spent the last two years working on the the Porsche 919 Le Mans team that has won the race for the last two years; Matt Verminski, who helped robotics company, Kiva, become a leader in self-driving warehouse robots and was sold to Amazon for nearly $800 million; Marc Minor, who helped Carbon off the ground with one of the most watched 3D printing videos on YouTube to date; Animesh Bose, a pioneer of powder injection molding who has co-authored a book in the area of Powder Injection Molding; Tuan Tranpham, a man many of you know from his long history in 3D printing and has come from Arcam on the back of the acquisition by GE; Peter Schmidt heading up design, MIT Ph.D. founder of Original Machines Studio. 

The list of names is exhaustive, Ric describes the team as, 'not having training wheels,' and in a brief encounter with Ely Sachs, he described Desktop Metal as the 'most exciting project' to date. He can say that again.