Yesterday, two eagerly-anticipated 3D printing platforms were unveiled by Desktop Metal, a company that has garnered a great deal of funding—just shy of 100 million USD—and consequently a great deal of publicity.
The R&D work had hitherto been shrouded in mystery. Only very basic information had been released, namely that it was a ‘revolutionary’ metal process, destined for office use (hence the ‘desktop’ moniker) at a very competitive price point, and that some very clever people were behind its development.
Having read all of the press coverage following the initial announcement, the general consensus seems to be very positive, if not excitable. Of particular note are the impressive build speeds quoted, the extended range of metal materials available for this process (courtesy of working with metal injection molding [MIM] powders) and the easy removal of metal support structures from parts which for metal AM has long been a holy grail endeavor.
There was some surprise that Desktop Metal lifted the lid on two different additive processes via two different hardware systems, both enabled by a single, novel core technology for post-processing. In terms of the actual metal 3D printers, one is in line with expectations in that it is office-friendly, very cost-competitive and due for commercial release towards the end of this summer; it’s called the DM Studio system.
The second platform, called the DM Production system, is expected to be commercialized later in 2018 and, as its name suggests, it is bigger, more expensive and intended for production applications. In terms of the core technology that enables both printers to produce metal parts, this is a post processing system, generically termed a sintering furnace.
Two 3D printers, one common denominator
The two 3D printers are based on different 3D printing processes. The DM Studio system utilizes an extrusion process which Desktop Metal has dubbed bound metal deposition (BMD). Essentially this involves depositing metal ‘rods’ (aka powder) in a specially developed polymer binder to build the part, at a layer resolution of 50 microns (µ) within a maximum build volume of 300 x 200 x 200 millimeters and at speeds up to 16 cubic centimeters (cm3) per hour.
The DM Production system, meanwhile, is a powder bed process, utilizing binder jetting technology. Not too surprising when you consider the patent holder for the original ZCorp technology—one Ely Sachs—has been working on this development. The process adds yet another acronym, SPJ, standing for single pass jetting, and involves a bi-directional binding approach, whereby two full-width print bars (containing over 32,000 jets) work in conjunction with powder spreaders to spread powder and print in a single quick pass across the build area, jetting millions of droplets per second. It is this that has resulted in Desktop Metal quoting speeds that are more than ‘100 times faster than today’s most common metal 3D printing systems, to produce up to 8,200cm3 per hour.’ This speed, coupled with a build area of 330 x 330 x 330 mm and a layer resolution of <50 micrometers (µm) voxels, is what is getting people excited—in combination with the use of MIM powders that are readily available and not as cost-prohibitive as the specialty metal powders required for additive laser based, powder bed fusion systems.
The above two paragraphs describe the actual 3D printers that build the parts up, but the parts produced by both processes fundamentally depend on a subsequent post-processing step that involves them being sintered. To this end, Desktop Metal has developed a patented sintering system, also described for clarity as a ‘specially designed microwave’. The sintering system combines silicon carbide (SiC) heating elements with high-powered microwaves to reach temperatures of up to 1,400 degrees centigrade (°C), and, according to Desktop Metal, temperature profiles can be tuned to every build and material.
So, for the BMD process, this sintering step removes the polymer material and binds the remaining metal rods in a carefully controlled, fully automated process. Importantly, the sinter system operates closed-loop thermal control, which along with its ability to fit through a standard doorway, is what makes it office and studio friendly. The point to understand here, though, is that the sintering process involves heating materials to just below their melting point, to the point that the powdered material fuses together. I’ll come back to this.
In terms of capital costs, the complete Studio system—comprising 3D printer, debinder and sintering system—is being offered for 120,000 USD. However, in a familiar move, there is also a subscription model available for this hardware (as a service), for 3,250 USD per month, over a four-year period. The Production system, which, as previously stated, will not be available until next year, is being offered at 420,000 USD, again including all the ancillary equipment required, but considerably less competitive.
The other really interesting point to note from Desktop Metal with these systems is the development of a unique, and therefore patented, support removal feature. Removing metal supports has long been a difficult and time-consuming post-processing requirement for metal AM. What Desktop Metal has done is a significant improvement in this regard. The company has developed a specific method for printing a ceramic release layer of material between the part and the support structures. Thus, when the part is sintered following the build, the ceramic material converts to sand and the metal supports can subsequently be quickly and easily removed—by hand, no less. Although you might want to wear gloves.
An important distinction: sinter or melt?
But let’s go back to the sintering process. This is where I have a few alarm bells ringing, because over the years in the additive manufacturing (AM) industry, ‘sintering’ has become a very familiar term, thanks largely to ‘laser sintering’ being the term of choice for early laser powder bed processes from DTM (acquired by 3D Systems) and EOS. The terminology transferred even as direct metal laser sintering (DMLS) evolved from the plastic technology base. Thus, as selective laser melting (SLM) emerged, little emphasis, apart from in the highest echelons of academia, has been placed on the fundamental differences between ‘sintering’ and ‘melting’—rather, they have been used interchangeably in many instances.
In terms of the actual processes, particularly for the production of critical applications in metal, this is not something that should be overlooked.
As highlighted above, the definition of the sintering process is that it does not fully melt the process material, rather it heats it to a level just under the melting point, allowing it to fuse at a molecular level, meaning that there is a level of porosity involved, which can be controlled. This is in contrast to melting, which for additive laser (or electron beam) processes involves full melting of the powder to form a homogenous part that is demonstrably proven to be closer to full density with the implied strength implications that this brings.
However, any additively manufactured metal part can suffer from internal stresses due to the very nature of being built one layer at a time. I do not think the Desktop Metal processes are going to eliminate this problem, but like their higher energy, more expensive counterparts in this space, they will likely benefit from the wide-ranging R&D for in-process quality assurance (QA) and traceability solutions required for critical applications.
To sum up
I do believe that Desktop Metal’s initial reveal is worth getting excited about. It offers new technology platforms that present innumerable opportunities for a much wider market base for economically prototyping with metal materials. Although, I would temper that with the fact that it is just a starting point and for production applications with the larger, more capable machine—that is itself a year away from production—some caution is required. Like all AM processes that have gone before, this one will likely find its place within the ecosystem, it will probably accommodate some novel, high-profile applications and will undoubtedly contribute to further the evolution of 3D printing and AM. At this point, though, I am prepared to go on the record to say that I don’t think it is going to transform the entire industry in the way that some coverage implied yesterday.
That said, I am very much looking forward to seeing this technology in action at the RAPID + TCT show in Pittsburgh in a couple of weeks and putting more questions to the chief executive officer (CEO) Ric Fulop.