A pile of clear, pear-shaped respirator shells grew on a worktable in the Frith Lab in early April. Genevieve Gural and Rod La Foy called it the “graveyard,” where they left flawed shells.

Other things were growing in the lab, too. Empty coffee cups from three different Blacksburg shops stood on the next table over. In the back, dust billowed in brown clouds and settled as a machine carved the pear shape into engineered wood, for the next version of a shell mold.

The things that were piling up around Gural and La Foy — materials, tools, wood dust, hours worked, failed attempts, lost sleep — all came about in a months-long project: the novel design and production of an N95-alternative respirator.

Led by Chris Williams, head of the Design, Research, and Education for Additive Manufacturing Systems (DREAMS) Laboratory, a team of Virginia Tech mechanical engineering graduate students has created the reusable respirator using rapid tooling and 3D printing resources from on-campus makerspaces and research labs.

In efficacy tests run by local company NanoSafe, co-founded by Matthew Hull of the NanoEarth team, the design has performed as well as or better than an N95. As the DREAMS team adjusts the final design, they’re making sure that their tools and fabrication methods can be scaled up for rapid production.

Long term, the design will serve as backup. If the region's health systems need more respirators, the team can produce a batch for them. And if other universities or institutions with similar resources aim to help health care systems in the same way, the DREAMS lab’s design provides a low-cost option for producing an N95 alternative, made with tools that are easy to use.

“We have a respirator design that offers N95-level protection and is comfortable to wear, but can be sterilized and reused, which reduces the total volume needed and also reduces waste,” said Williams, the L.S. Randolph Professor in Mechanical Engineering. "We’ve also designed it in such a way that it can be rapidly produced. About one shell per minute. It goes to show that with the fabrication capabilities at our university — and other universities — it’s possible to pull these components together and fabricate a useful object.”

The project is among 10 COVID-19 response projects started by Virginia Tech faculty, staff, and students in March to design, produce, test, and deliver medical supplies to local healthcare systems. In the months since, the students working on the novel respirator, all part of Williams’ 22-member lab (save for La Foy), have broken up into small teams or gone solo to work on each of the respirator’s parts: the shell that goes around the mouth, the silicone seal that fits the shell to the face, the lock-in housing that holds the filter material, and the ring that attaches straps.

Novel respirator design
Photo courtesy of Rod La Foy.

To figure out how they’d make these myriad parts, the team weighed speed, cost, and efficacy. No one knew how an N95 worked when the group started the project, but they were all highly trained in manufacturing. The DREAMS lab researches new materials, machines, and processes for 3D printing. Williams described their expertise as deep, but broad: they look closely at the relationships between printers and the materials they process, for many different printing techniques. Their lab space is set up for nearly every 3D printing process that exists. 

“When a design opportunity presents itself, we’re able to make the right decision of which process to use,” Williams said. “We knew at the very beginning that we had the capacity to help. The technologies to help. But we didn’t know how we could best help immediately.”

To 3D print or not to 3D print

Though Williams is a 3D printing expert, his team looked at a larger set of manufacturing methods for the respirator’s design and production. “It’s funny, our expertise in 3D printing doesn’t mean we want to 3D print all the time,” Williams said. “What it means is we think hard about when it’s appropriate and not appropriate. We know the strengths and weaknesses of 3D printing, both from a materials standpoint and a process efficiency standpoint.”

Respirators created entirely with 3D printing have emerged all over the country. But it quickly became clear to Williams that in the context of designing a respirator with N95-level protection at scale, a fully 3D-printed respirator would have too many weaknesses.

Because 3D printing involves building up layers of material to form a part, rather than reshaping or cutting into stock material to give it a particular shape, Williams said it’s inherently subject to porosity — tiny holes between the layers where bacteria, and in present day, SARS-CoV-2, can get in. That would potentially create hiccups in producing parts like the shell, the respirator’s largest component. On top of that risk, the large-scale printing of every part would be expensive and time-intensive.

The team decided to use 3D printing not to make all of the respirator’s parts, but mainly to give them their shape, through molds. For the shell, Gural and La Foy used the Frith Lab’s CNC routing and 3D printing systems to prototype molds, and chose vacuum forming to make quick copies of the pear-shaped part itself. Vacuum forming is a rapid tooling process that involves heating a sheet of plastic and pulling it down with suction over a mold, giving the part a fully dense, smooth shape.

Vacuum forming respirator shells
Vacuum forming involves heating a sheet of plastic and pulling it down with suction over a mold. Material shifts, buckles, expands, and contracts inside the machine. Photo courtesy of Genevieve Gural.

The students then combined five molds in a circular set they called the “pizza” to allow for production of multiple shells at a time. Once they had the molds’ dimensions finalized, they used an industrial-scale 3D printer from the DREAMS lab to print molds ready for large-scale production.

“We’re using 3D printing where it makes sense, which is to make complex shapes at low production volumes, and vacuum forming where it makes sense, which is to make quality parts at high throughput and low cost,” Williams said. He and the team went through the same weighing of options for each mask part.

‘This experience is impossible to simulate’

As Gural and La Foy worked on respirator parts, they spent most of their spring in the Frith Lab and the DREAMS lab in Goodwin Hall. La Foy, a mechanical engineering Ph.D. student who works as a senior graduate teaching assistant in the Frith Lab, volunteered to help Gural when she asked the engineering education department and the lab's director, associate professor of practice Mike Brown, for access to the space and its tools.

Gural and La Foy often worked through the night in the otherwise empty labs. As the COVID-19 pandemic shut Blacksburg down around them, they saw few people other than one another and a couple of coffee shop baristas. La Foy would hear pipes cracking in the Frith Lab, which sounded to him like footsteps, and he’d turn to see if Gural was there.

To Gural, the rattle was an alien sound. She’d heard it when she worked in the lab as a sophomore. She’d spend her last days as a mechanical engineering master’s student working to that sound, before moving on to a job at SpaceX in June. La Foy has continued working out details for shell production.

“We did a lot of initial testing, and I want to say trial and error, but it was mostly error,” Gural said of the first few weeks of the project. The very first shells were particularly bleak, Gural remembered. "It looks like a tent," she said of one. You can kind of see an impression of a circle at the top of it. Very sad.”

Genevieve Gural in the Frith Lab
Early in the pandemic, Genevieve Gural works in an empty Frith Lab, before the "graveyard" shells have piled up. Photo by Peter Means for Virginia Tech.

Together, she and La Foy problem-solved their way over hurdles as varied as missing plugs, broken emergency stop buttons, and a vacuum-forming phenomenon called webbing, wherein small elements in the geometry of the shell molds can cause the hot plastic that’s sucked down over them to form ridges, making it difficult for the silicone seal to adhere to the shell. 

Gural would film webbing on her phone and slow it down to study the videos with La Foy. Eventually, after they tinkered with the molds’ positioning and tried vacuum forming the shells again, the webbing was gone. Gural tossed her phone onto the lab counter, threw her hands in the air, and walked off, triumphant.

“It’s astounding, really, how much energy they’ve put into this,” said Williams of the students. “We teach product design and manufacturing processes via design projects every semester, but despite my best efforts, I never could have simulated this in a classroom setting.”

“It’s not hard to get motivated on a project like this, where the impact is immediate and the need is great,” said mechanical engineering Ph.D. student Sam Pratt, who designed a silicone seal for the respirator. The part is made using casting: liquid is poured into a 3D-printed mold and cures into a sterilizable, silicone seal. Pratt worked with La Foy and Gural as the three went through versions of the shells and seals, to make sure the seal fit the new shell and vice versa. 

“None of us were particularly hung up on how cool or important ‘our part’ of the system was, which meant that iteration happened really fast, and ideas — good or bad — were tested almost as fast as they could come up,” Pratt said. “It was a really exciting demonstration of the power of a prototype-first mentality. If you have an idea, build a prototype and prove it works. Usually before the end of that day, everyone on the team had seen it, played with it, tried to break it — that's how we were able to run through so many concepts so fast.”

For La Foy, part of the motivation for the time he’s spent on iteration after iteration is to feel less powerless. “It’s depressing and frustrating to see all of the suffering going on right now,” La Foy said. “But the more that people are willing to work together to get through this, the better we will all be when we make it through to the other side of this crisis. Working on the respirators has been one way that I can use the skills I have to help others out.”

Members of the team of current and recently graduated students who have worked on the novel respirator include: Sam Pratt, Genevieve Gural, Rod La Foy, Viswanath Meenakshisundaram, Derick Whited, Liam Chapin, Bemnet Molla, Kendall Knight, and Amanda Wei.

This project and other COVID-19 response efforts received support from the Fralin Biomedical Research Institute at VTC. Further development of the respirator design and testing is supported by the Office of the Vice President for Research and Innovation COVID-19 Rapid Response Seed Fund.