Invention, optimism & realism: The back story of SLS 3D printing with Dr Joe Beaman

Dr Joe Beaman, a mechanical engineering professor at the University of Texas, worked with the late Carl Deckard to develop Selective Laser Sintering in the 1980s.

by

"Let this guy in, I think he's got some potential."

Dr Joe Beaman is reminiscing. It’s what you do when you lose a friend.

He’s thinking back to when a master's student came to him with an idea, to when they developed a 3D printing process together, then a machine, then a business, to when they thought they were just a year or two away from printing a motor vehicle, to when the press interest reached such an overwhelming point that he had to stop answering the phone.

The person Beaman experienced all that and more with is gone, but the memories remain, the story waits to be told, and the phone is back on the hook.

“Carl Deckard came into my office with this concept,” begins Beaman, a University of Texas (UT) professor of mechanical engineering, interrupting himself as he jogs his memory. “We were actually trying to solve a problem: how do you make the first one of something and not have to wait six months [for the service provider to deliver it]?”

Deckard had first become aware of this rigmarole having worked in a Houston-based machine shop called TRW Mission in the summer of 1981. Here, he recognised the potential for companies to save time by creating casting patterns out of CAD models rather than handcrafting them. A few years later, while studying a Mechanical Engineering Master’s Degree, Deckard was walking into Beaman’s office.

It is at this moment that Selective Laser Sintering (SLS) was born. When Deckard passed through the door of Beaman’s office, he explained his concept of fusing powder to build parts layer by layer, kickstarting a chain of events that would see the technology commercialised by the end of the decade. Yet, it’s a conversation that wouldn't have happened if not for some killer foresight from Beaman.

“When I first met him, he was just coming into graduate school and I had to actually vouch for him to get in,” he recalls. “His [undergraduate] grades were okay, but not terrific and so I remember talking to the graduate advisor and saying, ‘let this guy in, I think he’s got some potential.’”

This potential coincided with significant developments occurring through the 1980s in the technological landscape. Among them were the increased accessibility of computer solid modelling and personal computers, which would facilitate the slicing of 3D models, but perhaps even more important for the growth of Deckard’s invention was the democratisation of CO2 lasers. Beaman remembers a company offering a CO2 laser at around $5,000: “before, they were more like $100,000. That triggered what Carl and I were thinking of doing and then he came up with this concept of using powders.”

Deckard then set about sourcing components to build a machine that would allow him to enact his vision. Having successfully printed a part direct from a 3D CAD model and presented it to Beaman, Deckard went on to receive his master’s in 1986. The machine he used at this time featured a 100-watt YAG laser which was being regulated with a Commodore 64 computer after some Deckard tinkering.

At the turn of the academic year, with Deckard now beginning his PhD studies, a grant of $30,000 was received from the National Science Foundation (NSF), allowing him to add a counter-rotating roller for more even powder deposition. Meanwhile, a lamp that was installed primarily to illuminate the interior of the machine turned out to be powerful enough to heat the bed surface. “We started making pretty good parts after that,” Beaman insists. “Before, we made some cr*ppy looking parts which were distorted and everything else.”

Bit by bit, Deckard and Beaman, with the help of another student, Paul Forderhase, were edging closer to a machine design they would be happy to show to investors. The wheels had been in motion for commercialisation since October 1986 when Deckard met with Paul McClure, the Assistant Dean of Engineering at UT and future CEO of DTM Corp, and Harold Blair, a business owner from Austin. A business name, Nova Automation, was already being attributed to the work Deckard and co carried out into the late 1980s. But there was a hurdle in the way. As a public university, UT didn’t allow faculty members or students to own a business created at the university, considering it to be using public property for private gain.

“And there’s some rationale for that, right?” Beaman concedes. “I’m using university facilities, why should the public pay for that? But we were going all the way to the highest authority in the university system and they agreed to make it legal. I was the first faculty member, [Carl] was the first student, that were actually allowed to own equity in a private company built on campus. That’s another thing that happened which was serendipity in some sense – it happened to be the chancellor was an engineering professor and he came by and got excited about it.”

In fact, he – that’s Hans Mark, an aerospace engineering professor – kept coming back and was later recognised by McClure for his support in ensuring UT protected the intellectual property. McClure, himself, also played a big part in the development of SLS and was a name that kept cropping up as Beaman rolled back the years.

He, according to Beaman, was responsible for getting the grant from NSF and also connected UT alumnus Dave Bonner - the VP of Research at manufacturing outfit Goodrich Corp - with the rest of the Nova Automation team, which led to an investment sum of $25 million and the founding of DTM. To help convince the likes of NSF and Goodrich to back Deckard’s endeavours, McClure also put a lot of work into getting the technology into the press.

Beaman maintains some of the newspaper clippings to this day. One story published in the Austin American-Statesman in 1987 pictured Beaman and Deckard next to an early prototype of their machine alongside the headline: ‘Revolutionary’. Two years later, the Wall Street Journal reported on the ‘growth of Desktop Manufacturing’s following’, as DTM readied to challenge 3D Systems with the launch of a ‘faster machine that can produce parts out of a broader range of materials’ – the birth of the additive manufacturing (AM) industry’s incessant marketing one-upmanship, perhaps.

There was another story, Beaman remembers, published in the New York Times. “When that article hit, I started getting phone calls from around the world; they went by the time zones. In fact, I got so many calls I had to quit answering the phone.”

What Beaman could take from his phone ringing off the hook at all hours, though, was that SLS technology was gaining traction. While they were having to catch up to 3D Systems, who had been in operation since 1986, DTM was still one of the few companies promising a means to turnaround prototypes within a few hours. When SLS was first introduced publicly at the Autofact trade fair in 1989, companies like Ford Motor Company and Pratt & Whitney/ United Technologies Corp began to take notice too. Not long after, Beaman sold the first machine to Frank Zanner from Sandia National Labs, which the company’s project leader, Clinton Atwood, would use to produce investment castings.

This spate of interest led to the creation of the Solid Freeform Fabrication Symposium, which still takes place in Austin every year. The first time it did in 1990, Beaman remembers, there were no more than 20 people sat around a table – including representatives from Ford, United Technologies and the US Naval Research Lab – delivering presentations with shared manuscripts.

The conversation around that singular table focused, quite level-headedly, on functional prototypes, with the occasional query about metal 3D printing technology. “How do you get a prototype faster that was functional at the same time? They didn’t want parts for touch; if I had a snap-fit, it needed to be a snap-fit,” Beaman explains. And yet, away from the symposium and their potential clients, Beaman and Deckard were thinking bigger than prototypes, and bigger again than Sandia's castings. 

“We were overly optimistic,” Beaman laughs. “We thought we were going to be printing automobiles in two years. That didn’t happen, by the way.”

It wasn’t the only time optimism was overwhelmed by reality. In the time since Deckard started his PhD, he and his team had sought to take advantage of the Texas State’s newly established Advanced Technology Programme, designed to invest in high-tech after the oil bust, to enhance their SLS machinery. A design named ‘Godzilla’ was abandoned, but the 'Bambi' machine was built between 1987-88 – it would remain as a research platform at UT – and then the Mod A (showcased at Autofact) and Mod B followed. Between 1989 and 1992, DTM unveiled three more generations: 125, Beta and SinterStation.

Amidst all that, the financial arrangement with Goodrich was set up, Nova Automation became DTM, and McClure was replaced by John Murchison as CEO. Beaman took a leave of absence from UT between 1990 and 1992 to take on an Advanced Development role and began to realise, while so much graft had been put in to get SLS to market, there was no let-up as a venture-backed commercial entity.

“We discovered machine business is a tough business,” Beaman says, “especially if it’s capital goods because we were at the level where it would take a fairly high-level person in the company to sign it off. And how many machines do you need to make prototypes? We weren’t making parts for final production at that point and the market went up, then went down, and a lot of companies tried to get into forming service bureaus.

“DTM had some financial troubles, but if we had just sold parts and not machines, we would have been fine because there’s never a reduction in the demand for parts. The trouble is, you can’t get your capital back very fast by selling parts and Goodrich was used to selling freight carloads of material. The machine was going to be an avenue for them to do that.” 

Selling machines wasn’t the only cause for concern in the 1990s. A year before Deckard’s departure in 1993, and not long before Beaman’s in '92, DTM discovered a similar patent to that of SLS filed in 1979 and granted in 1981. It belonged to Hico Western Products and was filed by the co-owner of that business, Ross Householder. This idea focused on pouring sand and cement into a square matrix a layer at a time and solidifying it with water, but with a job opportunity arising in Saudi Arabia the year the patent was granted, Householder left the idea behind. When he came across a report of SLS in a magazine, he contacted DTM and negotiated a deal to sell the rights.

Beaman says he and Deckard were unaware of Householder’s patent when they got to work developing SLS. And while this patent has been cited hundreds of times since and may have posed a risk to the future of DTM, Beaman believes the kudos rightfully goes to Deckard for not just formulating an idea but getting SLS to a point where it could be applied on factory floors.

“Carl’s real emphasis and significance were he was the one that actually made it work,” he says. “He was down the lab doing the processing and building the equipment. I really give attribution to someone who actually manages to make the thing and make some parts.”

Though Deckard and his peers had managed to get SLS into the commercial space, acquired some early customers and purchased similar patents to protect DTM’s flagship product offering, Deckard and Beaman didn’t stick around. Beaman returned to UT; Deckard decided to channel his inventive side once more after three years as a professor at Clemson University.

Deckard’s next idea was a hybrid rotary reciprocating gasoline engine that had only one major moving part and sought to replace engines in small hand-held applications. While he never managed to make a success of this project, SLS continued to grow gradually, even if the company supplying it to market stagnated.

Goodrich bowed out of the DTM project in 1999. Unsatisfied with machine sales, and in turn materials sales, the majority shares were sold to a private investment firm, who would then sell the firm to competitor 3D Systems. Rival SLS firm EOS continued to prosper, companies like Farsoon and Prodways also entered the market with similar processes, and there are now thousands of SLS machines installed around the world. Some 20 years after his departure, Deckard returned to the AM industry to expand the materials choice for these users with Structured Polymers, a company developing polymer powders from chopped pellets. Evonik acquired the company in 2019.

Although never again active in the commercial side, Beaman continued to contribute too. After his return to UT in 1992, and perhaps harnessing some of the thoughts shared at the symposium and maintaining the early enthusiasm, metal 3D printing was in focus. Beaman says he and another professor named Dave Bourell first managed to print a metal structure on the Bambi machine – “of all places” – before Suman Das came under his supervision. For Das’ masters and PhD studies, he designed and built two metal AM machines, powered by a process we now know as Selective Laser Melting, in collaboration with Lockheed Martin and Rolls Royce. In Germany, the Fraunhofer Institute was also working to develop metal laser melting technology at this time. 

For Beaman, it was vindication that at least some of his and Deckard's early optimism came to fruition. At the first Solid Freeform Fabrication Symposium, they were delegates from the aerospace and automotive industry who asked about metal. Beaman and co were 'adamant' they could, Sandia backed them up too, and while they weren't printing entire cars, Das did manage to process titanium and nickel superalloys. They were further buoyed by an ‘eye-opening’ research paper published by an English university in the mid-1990s which suggested, in theory, it would make economic sense to 3D print a volume of polymer parts up to around 14,000 than manufacture them through injection moulding: “That's when people said, 'man, we should really try manufacturing'.”

To that end, Beaman has kept in tune with the development of 3D printing technology. He has worked on the development of superalloys with Sandia National Labs and has also researched optical coherence tomography to inspect parts layer by layer. Another of his subjects has centred around process control, such as controlling laser powder by looking at the temperature profiles in the bed, which Beaman says they underestimated back in the DTM days.

One thing that Beaman didn't underestimate, however, was potential. One thing that never changed is his penchant for doing new things. It’s why Beaman has continued to do his bit to push 3D printing technology forward and why he was so receptive to the ideas of a young Carl Deckard, who he sums up as a ‘creative genius’ and, on a personal level, ‘like a son’.

“I always tell faculty; I like someone coming through the door with a new idea. What’s nice about SLS is it’s kind of simple in a way; doing things in layers and drawing patterns on powder. That’s a lot less complicated than an engine.

“That company [Structured Polymers], I think it’s going to do okay. They have a neat process where they can do a better job of the distribution of the particle sizes. It doesn’t surprise me that Carl was involved in it because it’s different. I don’t know who else would have thought of that.”

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.

Featured Additive Manufacturing & 3D Printing Insights