3D Printed Masks for COVID-19 Front Line Workers

From March-September 2020, I worked with a coworker and a local nurse to 3D print masks, ear savers, and other PPE for community members and organizations in need.

GoFundMe Link: https://www.gofundme.com/f/3d-printed-masks-for-whatcom-healthcare-workers
No longer accepting donations-Link contains information about the project throughout the process

    At the beginning of the COVID-19 lockdown, many people were trying to find a way to help out. The 3D printing community was no exception. As a 3D printing enthusiast, I began to see designs for 3D printed masks boasting N95 performance circulating 3D printing forums. One particular open-source design caught my interest, the Hack the Pandemic mask by Copper3D (the design has since changed). I was interested in the design because the masks were printed as a flat layout that could then be thermoformed to a user's face for an effective seal. They were also designed to be used with Copper3D's PLActive copper infused filament. Since copper is naturally antimicrobial, this mask eases sanitization concerns. The website claimed that using simple makeup removal pads as filters could replicate an N95 mask.

First Iteration

    Because of all these factors, I sent this design out to my coworkers. One of my coworkers forwarded the design to his wife, a local nurse. She expressed a huge concern about the availability of PPE at her institution. Only those with direct COVID contact were given N95 masks. Those working in the facility but not in direct contact with COVID were given surgical masks that often had to be reused. With little known about the virus, an N95 mask was desirable for all of the COVID-adjacent workers. Based on this expressed need, I began 3D printing a couple prototypes of the initial Copper3D mask design.
    The three of us began assembling the masks and made some initial improvements. We added epoxy to the seams on the mask that is bleach resistant so the masks can be easily cleaned. We started looking into a bleach resistant strap system as well. We also quickly developed a manufacturing system with a mannequin head to easily pre-form all of the masks.

The Filter Problem

    One of the biggest unknowns for these masks was the filter material. Even if the masks form a great seal, they will only work as well as their filters. N95 filter material would be ideal, but that was all being used in production already, so we needed to find a good substitute. Based on research, we gathered multiple materials to try. The first was oil absorbent pads which are made of non-woven polypropylene in a very similar process to N95 masks. We also gathered knock-off N95 materials, HEPA air filters, and ​hospital-grade OR sterilization wrap.

Second Iteration

    The first iteration of masks had a lot of interest from hospital workers. At this point, we started a GoFundMe to help financially with our efforts. We ended up raising over $3,500 with which we purchased 4 more 3D printers, filament, filter material, strap material, assembly material, and anything else needed in the process.
    The initial generation of mask seemed promising, but had some issues to solve. It was tough to breathe out of. Someone had modified the open source design to include two filter ports, so we adopted that design. We also found that it took practice to get a good seal with just thermoforming the mask. We added closed-cell neoprene foam tape around the mask and found that it fit most faces with no adjustment. Small adjustments could still be made if necessary. Neoprene rubber straps were added as well. All materials were selected to be bleach resistant and were left in a bleach solution for 24 hours to prove their durability.

Filter Testing

    To test the filter materials, we initially set up a homemade N95 fit test where we put a mask on a volunteer, put a bag over their head, and sprayed perfume into the bag until the volunteer could smell/taste it. This test failed miserably as the volunteer was overwhelmed with perfume within seconds. We then resorted to testing the efficacy of the seal by blocking the filter holes and having a volunteer put their head under water. The seal was proved effective as water did not leak into the volunteer's mouth and nose area.
    We then reached out to a local company who certifies N95 masks, and they let us use their equipment for quantitative data. The machine uses a tube fitting and sensor to measure the amount of a particle that enters the mask when the particle is released into the atmosphere around the mask. The volunteer is asked to complete a series of actions including deep breathing, head movement, and reading out loud. A result of pass or fail is returned, with a numerical value. Under 100 is a fail, so a higher number is more desirable.
    In our first round of tests, all of the potential filter materials failed. We were able to get our hands on some 3M style P95 dust filters (pictured on the right) for testing. These filters when paired with the proper mask meet N95 requirements. We cut these filters to fit inside the filter openings on our mask, and once again, the test failed. Then as a baseline, we modified the filter holes on our mask to accept the 3M filters and the mask passed. We realized that the geometry of our filters was producing failed results. The main difference we noticed was the surface area of the filter. The screw-on filters in our masks direct air through a very small part of the filter material. The 3M filters have a large amount of surface area exposed to incoming air. We took this knowledge and proceeded to a new iteration.

Final Iteration

    After the first round of testing, we redesigned the filter caps to accept a filter with a larger exposed surface area. The filters pressed into place with an outer ring that allowed for a complete seal with maximum surface area exposed. Once again, these masks were tested. The results were better than before, but they still all failed overall.
    At the same time, a different style of 3D printed mask was being created and tested by another group. This was called the Montana Mask, developed in Billings, Montana. This group was able to certify that their mask was as effective as an N95 mask when N95 filter material was used. They were using the same testing methods that we were. We made one of these masks and tested it with the same filter materials. We were able to get duplicate results from the Montana mask and our mask. so we concluded that our mask was effective given proper filtration. Of our filter materials, we found that the OR sterilization wrap gave the best results.

    Throughout this testing and development process, the target market for these masks changed. Hospitals were understandably reluctant to allow non-certified masks in their facilities. Therefore, we began designing with COVID-adjacent people in mind. These places included fire departments, pharmacies, police offices, and community members who may come in direct contact with COVID. We made sure to be very clear that our masks are not certified and are only as good as their filters, but we believe that they can provide more protection for the user than a standard surgical or cloth mask.

Ear Savers

    The masks generated a lot of interest, but our most popular offering by far was ear savers. These simple 3D printed devices hold the mask straps on surgical style masks to prevent wear on the ears. This was an open-source design and there are many iterations. We began handing these out to medical offices and essential business all over the place. At one point, I was carrying a stock with me to hand out when I had to visit an essential business. We were also able to print other pieces of PPE on request, such as visors for face shields.

Part of a delivery of 500 ear savers to a police department

Results and Donations

Members of the Big Lake Fire Department wearing the masks that were donated to them

    In the end, we donated around 100 masks, over 1,000 ear savers, and around 100 face shield visors. We offered our dual filter mask, the Montana mask, and the 3M style mask as options to the organizations where we donated, and our dual filter mask was by far the most popular. In August, 3D printers were no longer needed to fill the gap in manufacturing created by the COVID pandemic. We were able to effectively help community members fill this gap as mass-manufacturing caught up. In August, 2020, I was able to donate the printers, materials, and remainder of the funds to the engineering department at Sedro-Woolley High School, my alma mater. All of the funds were used for development, manufacturing, and distribution of the products, and the equipment purchased was all donated at the end. The three of us involved in the project did not take any compensation.
    This project was an extremely practical example of product development. We started with an existing product and innovated on the technology through research and testing. We then adjusted the products we produced based on the needs of our customers in the community. By thoroughly understanding these customer needs and where our masks could be most effective, we were able to supply our community with the protection the need free of charge, and at a low cost to those who donated.

The printers in their new home at Sedro-Woolley High School