On a recent morning, Rick Lake stopped in the middle of giving a tour of NIST’s Machine Shop to ponder a part sitting on a shelf. It was about 25 centimeters long and made of metal coated in a sheen of purple and blue. Along the top, gaskets, nuts and bolts stuck out like spines on a steampunk caterpillar. At the top, one big opening poked out like a head.
“It’s a manifold made for integrating gasses,” Lake explained, and each opening was supposed to provide an attachment for a nozzle. The dramatic, almost psychedelic color was the result of coating the entire piece, made of stainless steel, in vaporized glass. The coating was experimental, and was meant to reduce reactions between the metal of the piece and the mixture of the gasses in the scientist’s experiments.
Although interesting in appearance, it had failed. “It was impossible to hook up anything to the piece, because each thread got clogged,” Lake said. There was no way to coat only the underside. The entire piece would need to be redesigned.
Invention in the movies takes a moment and comes like a flash to a single, remarkable protagonist. But in real life, innovation is often slow and takes a large team of people working together. Technology helps move the process along. But in the end, you are still likely to encounter unexpected challenges that demand creativity, persistence and patience. And sometimes, as was the case in the groovy-looking manifold, you must go back to the drawing board, literally.
Perhaps no one appreciates this more than the people who work in the Machine Shop, which sits in building 304 on the Gaithersburg Campus. These are the people who often end up helping the innovators and inventors of NIST take an idea from dream to reality.
“A lot of times, customers come in and kind of bounce ideas off us,” said Rick Lake, who is the shop’s cost estimator and Job Coordinator. “We give them some feedback. We then have our designers generate a computer drawing so that it’s a little more formal. And if it’s a bigger project, there’s an assembly of everything that’s translucent so that they can see what the inside of the part is actually going to look like.”
“Or it could be they come to us and say, ‘We’re going to do this experiment, and we’ve got this idea of how we’re going to do it, but we need some help,’” added Mark Luce, the division chief overseeing the shop.
The building itself is approximately the size and shape of an airplane hangar. An acrid smell of machine grease lingered in the air the morning Lake and Luce were showing visitors around. T-shirt-wearing machinists worked at more than a dozen workstations and attended to various interactive project screens, watching as pieces of shiny metal pivoted around on small platforms inside blue cabinets with see-through doors. In some cabinets, a mix of water and oil cascaded over cutting surfaces to prevent overheating. In others, filings of shiny metal were being scraped away like bits of carrot in a kitchen grater. In the background, heavy-metal music mingled with the hum of large motors, the sound of steel pipes being moved, and the occasional flare of a welding torch being fired up.
On one workbench, a set of clamps and plastic encasements held up two large pieces of thick, heavy aluminum about the size and shape of a turkey roasting pan. Along the exterior, various square compartments glinted in the light. When fitted together, they will house electronic instruments needed to photograph the phases of the Moon.
At another workstation, machinist Luther Vulgamott stood in front a of Computer Numerical Control machine housed inside a box roughly the size of an ice cream truck. With the confidence of a concert pianist, Vulgamott punched instructions into a complex key pad, causing the platform inside to move in carefully choreographed movements. Atop the platform, a piece of metal about 20 centimeters (8 inches) long turned from a rectangle into a complex shape complete with divots and screw holes.
“If you can draw it, we can make it,” Vulgamott said with pride, displaying two other made-for-tech pieces he’d fashioned that same morning for work in the NIST Center for Neutron Research. One was no bigger than a pencil eraser, the other, about the size of a postage stamp.
Interestingly, NIST has always had a machine shop, though those who worked here decades ago might not recognize many of the tools currently on the shop floor. In the early 1900s when the organization was still known as the National Bureau of Standards, the machinists were likely helped by a few blacksmiths, and everything was done with fundamental machines from the first and second Industrial Revolutions—lathes, mills and drill presses—that allowed craftsmen to shape the pieces manually.
When the Gaithersburg campus opened in the 1960s, the NIST machine shop employed approximately 50 people, who were assigned to tasks like model and pattern making, woodworking, instrument making, glassblowing, optics and electroplating. An old black-and-white photo hanging on the wall shows the factory-like look of the shop before it opened for business in 1966.
Although the shop people still do some manual milling and lathe work and even use a few of the heavy lathes in that early photo, they have, for the most part, traded in their rulers for high-tech computer modeling software. In fact, most modern machinists must get a college degree in fundamental manufacturing before they can begin the formal four-year apprenticeship needed to operate the tools of the trade. And at NIST, the certified machinists also need to be flexible and creative.
“We need to have very qualified people in the shop because they have to be able to program their own part and machine their own part, and—in essence—inspect their own part when it’s all done,” Lake noted.
This is different from those who work in modern production facilities in private industry, where speed is prized above all else.
“In production they’ll have one person who does all the set ups, and then they’ll turn it over to another guy who loads, hits a button and then unloads it when it’s done,” Lake, the estimator, said. Producers like that make thousands of copies of that one thing every day for weeks, months or years.
In contrast, those who work at the NIST shop make prototypes. They consult with the research staff and see an idea through from start to finish. It might take 10 years to get the device just right for the science experiment to work out. And they might only make one copy of the new device, ever.
For example, when physicist Jim Fedchak needed to create a dynamic expansion chamber for his work on a cold atom vacuum standard, he knew the machine shop would be able to help him. Fedchak’s research is used in advanced manufacturing and applied to things like the semiconductors used on the tiny chips inside of smartphones. The shop has helped him numerous times since he began at NIST in 2003.
“You can go to outside shops for lots of things, but quite often you only get what you submit and there’s not a lot of feedback,” Fedchak said. If, for instance, the scientist makes a mistake on the drawing or wants to adapt the design based on new information, it isn’t always easy to correct problems or make changes mid-process when you outsource, which can slow down research momentum and increase costs.
“It is very useful to have them here,” Fedchak said of the shop staff. “It is great to be able to go down and talk to the people in the shop face-to-face.”
Mechanical engineer and chief of the Quantum Measurement Division, Jon Pratt, also has lots of praise for the NIST machine shop and their work. The past few years of Pratt’s career have centered around building a better way to realize the kilogram in the form of a new and improved Kibble Balance, formerly known as a watt balance. The complicated work has involved some incredibly complex math and will eventually impact how mass measurements will be made in the United States and around the world. But what many don’t know is that Pratt began his time at NIST as a post-doc 20 years ago, working on manufacturing research in the machine shop.
He said the experience gave him an increased appreciation for the value of being able to fix and make things by hand and for those with an intuitive knowledge of how pieces of precision instruments could come together. The process of getting the best Kibble Balance has taken more than a decade of continual adjustments, tinkering and reconfiguration.
“In my land of balances, I still need people who can make amazing parts,” he added. During the years, he spent hour upon hour in the shop, going over every detail of each new balance with the team and watching as it was all made and put together.
Pratt said he’s fought against outsourcing shop work for years.
“If you are making the same part over and over again, [outsourcing] works just fine. But what we’re doing here, we’re making custom stuff—the best in the world, and you’re really going to be asking people to do things that are a little out of the ordinary because no one’s ever seen it before, nobody’s ever thought about doing it before,” Pratt said. “It just works a lot better if you can just have people sitting around a table actually talking about it.”
Colin Wrenn, an engineer and researcher from the NIST Center for Neutron Research (NCNR), said the machine shop is essential to performing the wide array of work the center’s users perform. One week, they might be working on a part for a major automotive company who needs to test a new coil spring assembly. Another, they may be helping to investigate the Big Bang theory, or the way medicines react to human fat cells. Although most of the time the engineers on his team bring the machinists finished drawings and CAD models for the parts they need, the shop often provides key insights as to whether a piece can be manufactured as designed or not.
An engineer might be able to “do a lot of nice things on CAD, but it doesn’t mean they can make the part work. Those might just be pretty pictures,” Wrenn said. At the shop they can usually tell quickly if something isn’t going to come together. “The machinists in the shop are quick problem-solvers.”
“We’re a user-based facility,” Wrenn said, noting that the NCNR is one of only two facilities in the U.S. able to do certain types of neutron research, and that many come here from outside of NIST to conduct unique tests. “It is vital that the shop stays here. Without them our ability to stay competitive would be severely impacted.”
Although some wonder if 3-D printing and additive manufacturing will make the work the shop does obsolete one day, those running the shop envision that, while they will increasingly use those technologies in concert with the Additive Manufacturing Lab (ADML), they will never entirely be replaced by them.
“People often come in now with a part they’ve made on a plastic printer, wanting us to make it out of metal,” Lake added.
“They are making toys, they’re making replicas,” said Luce. In the end, customers need functional metal parts that enable testing and experimentation.
Luce said he negotiated to get the ADML lab housed in the shop because “we know 10 years from now, in the future as technology matures, that’s going to be an option and it is going to be an option that needs to be provided by this organization.”
In the meantime, when you ask Luce and Lake what they like the most about their jobs, the answer comes quick and is said almost in unison: Variety.
“Every day something new walks in the door,” Luce said. “And in some cases, we only get to see a part of it, but six months from now we get to see how what we did impacted the development of this device, and maybe it ended up creating a new standard that’s going to affect some technology. That’s part of the satisfaction of doing this.”
“We are NISTers, like everyone else who works here,” he said, rubbing his hands together and grinning.