Additional technical information is also available at the MIT Center for Bits and Atoms coronavirus tracking page here.

Team contributors

Product development co-leads: Matt Carney (MIT Media Lab Biomechatronics) and Aaron Cantrell (Principal Designer at Cofab Design), Philip Brown (Wake Forest Baptist Health). 

Design/development/manufacturing team:  Jesse Jarrell (Owner at Kaos Softwear), Dave Carlberg, Che-Wei Wang (CW&T), Jake Horsey (Cofab Design), Mike Stone (Cofab Design), Steph Whalen, Patrick Kennedy (Tooling Kaos Softwear), Nelson Madaleno (Simoldes), Paul Voss (Associate Professor Smith College), Nick Moser (Maskproject.tech), Joel Stitzel.

Testing: ATORLabs, Frank Hernandez

Funding support: Media Lab Member Consortia, Wake Forest Baptist Health funded prototype molds, Kaos Softwear donated LSR mold and mask production, Cofab Design has also funded substantial prototyping.

Strategic support: MGB COVID Center for Innovation,  Hollingsworth & Vose (Filter Supply),  AFFOA.org, Glidewell Dental,  Merck KGaA, Kilpatrick Townsend (Legal),  Dassault Systems Solidworks.

Additional support: Suchit Jain (Dassault Systems - Solidworks), Marie Planchard (SW), Abhishek Bali (SW), Center for Bits and Atoms, UMass Amherst, UMass Lowell.

Goal

We aim to satisfy the four primary criteria of front-line healthcare workers: particulate and bacterial filter efficacy, breathability, and liquid protection.

• Bacterial Filtration Efficiency—ASTM F2101
• Sub-Micron Particulate Filtration—Determine particulate filtration efficiency as directed in Test Method ASTM F2299. Applied at 35cm/s air velocity.
• Resistance to Penetration by Synthetic Blood—Determine synthetic blood penetration resistance as specified in ASTM F1862.
• Inhale/Exhale resistance, residual CO2 NIOSH

Two hidden dangers

There are two hidden dangers in face mask design that cannot be detected without analytical equipment:

1. Re-breathing of residual CO2
   
2. Attenuation of particulate filtration efficiency at increased air velocities
   

If there is too much dead-volume between the face and filter, CO2 from the end of breath can be re-inhaled at the next breath, slowly decreasing your O2 intake over time. This is invisible to the user but can lead to cognitive impairment.

Most filter-media are tested at 5.3cm/s air velocity; however, actual air velocities may be much higher (over 50cm/s). For the same lung breathing capacity, reducing the cross-sectional area of the filter increases air velocity due to conservation of mass. Higher velocity particulates have higher kinetic energy and thus can penetrate deeper or even through filter media. Filter media that relies on electrostatic Van der Wall's forces, which are the most common N95 filter, are more susceptible to this attenuation than barrier media. 

Because a thing looks like a thing or, in this case, breathes like a thing, does not mean that it does the thing.  Please be aware of this in your own respirator designs!

Our team is addressing these hidden dangers with both testing and simulation in order to rapidly converge on a rapidly deployable and safe face mask design.

Testing

NIOSH breathability and CO2 testing has been performed at a facility in FL. We have sent three prototypes for testing to evaluate performance of both filter material as well as geometry of the mask itself.

We are also coordinating Filter Efficiency testing of NaCl challenge particles through AFFOA.org. Another MIT Lab is performing the experiments at expected air velocities to mimic expected conditions in the device. We are also running particulate challenge experiments with Associate Professor Paul Voss (Smith College) in order to get quick results to show trends we can use for rapid evaluation of design changes.

Simulation

Computational fluid dynamic (CFD) analysis of face mask flow can accelerate design and provide insight into unseen phenomena.

These simulations are qualitative, at the moment, but we are working to map the simulation to match experimental data. Once agreement is found,  the mechanical design can be rapidly iterated without the longer process of building prototypes, shipping, and then testing. There is always a balance between time spent analyzing simulation data and just building and prototyping. What is clear here is that the CFD tool provides unique insight into the fluid dynamic behavior and interactions between airflow, filter, structure, and breath.

(These simulations were created by Suchit Jain VP of Strategy & Business Dev. at Solidworks, and using Solidworks 2020 with Flow Simulation.)

Current Status

2020-09-28

• pre-NIOSH data shows >99% particle filtration efficiency and all other breathability tests passing.
• Establishing a quality management system.
• Preparing mass-production of units.
• Low-volume manufacturing and pilot tests at hospitals.
• Multiple filter vendors identified and providing 10k's of samples and test data.

2020-05-07

• Open Standard Respirator, Model 1 (OSR-M1) first prototype tooling in production.
  
• Low-quantity injection molded face pieces now molding LSR (silicone) in small, medium, and large sizes.
• We expect first articles of the rigid components in about two weeks.

2020-04-21

• Clinician evaluation of first articles at Wakeforest Baptist Health
• Initial results from NaCl filter challenge at 45cm/s from MIT Lab thanks to MGB CCI and AFFOA.org coordination. Showing promising results.
• Design Freeze of final version 1.5 (later renamed to be OSR-M1).
  

2020-04-15:

• First injection molded rigid components shipped.
  
• Diving into model rebuild as a master-model in preparation for v1.5 design freeze.

2020-04-12:

• First LSR molded components were made by Jesse Jarrell and Patrick Kennedy at Kaos Softwear.
• Injection molds are being cut for the rigid components. First articles are expected early in the week.
• Filter testing is undergoing at four different labs: WBH, Smith College, MIT
• Awaiting NIOSH test results from previous week.

Additional References:

Characteristics of Respirators and Medical Masks