Check end of summary for regularly released updates.

VERSION 2 FILE REPACK IS LAUNCHED 4/12/20:
-Now 4 sizes instead of 6, and renamed: Large Wide, Large Narrow (most people so far), Small Wide, Small Narrow (next most common, smaller faces)
-Masks all fit the same filter box
-Increased depth for better nose fit
-Grill is thinner for faster print with less waste
-Filter box redesigned
-All masks have 2 elastic strap attachment points and they are lower to fit different different rubber edge liners, tubing, etc
-Press-and-glue square to hold first 2 layers of filtrete in place to increase speed of installation and minimize glue usage
-fusion360 .fd3 files finally provided as masks were remade for me by Everett W. in Nevada, HUGE thanks to this incredible designer
-attachment loop if you want to add another elastic attachment site to the nose of my original mask design using a 3D pen

Youtube original tutorial at: https://youtu.be/ySOCNycv-aA

Youtube Post-production: Molding, Sanding, Sealing and Upgrades https://youtu.be/9XwoQl0alVE

Version 2 READY-TO-PRINT .gcode files for Ender 3 (220 x 220 print bed): temporarily down for repair, please slice stl files on your own for now

If for some reason you want access to the files from version 1, they will remain accessable here: https://drive.google.com/drive/folders/1He3TSlDc5UCXDhZF0_TLeBUN2oqIc1jD?usp=sharing

I'm a doctor who loves 3D printing. We have a nationwide PPE shortage as a result of COVID19. I'm ready to distribute this optimized design for free for in-house printing in hospitals and elsewhere. My hospital is buying printers and will be expanding fast. You may use these designs and please share with everyone. You may not sell for profit but may accept donations to cover the costs of the supplies.

I've calculated average cost per mask+filter to be $1.94
$0.97 PLA + $0.12 3M filter + $0.31 MERV 13 filter + $0.25 glue + $0.29 tourniquet or elastic.
Additional Costs with processing and upgrades: Suggested upgrades include spraying 1 - 2 layers on the outside of the mask with clear acrylic ($0.08 per mask), adding window/door foam to the edge ($0.40), and adding a thin layer of silicone around the box-mask junction - just enough to seal it, but still be pulled apart if filter change is desired ($0.03). Additional cost = $0.51; TOTAL = $2.45
In these approximate calculations, I am not including cost of supplies such as sand paper, garbage bags, gloves, organic fume respirator, and wear and tear on machines.

If you wear this OVER a surgical mask for comfort and to filter the air an additional time in case any virus gets through cracks or micropores despite sealing the plastic, this would add another $0.58. Here in the hospital we have a few doctors experimenting with a layer of Halyard H600 between the mask and their face. One even put velcro to hold this layer along the inside of the mask. That stuff is featured in my other video, theoretically pushing this mask up to filtration numbers >99% though physics lab tests are pending. See other project:
https://www.youtube.com/watch?v=xVYypLYEo_0
https://www.thingiverse.com/thing:4262131

Disclaimer:
There are no guarantees about the efficacy of any home made masks. Use at your own risk. Get creative, stay safe. These are okay to use per CDC ONLY if you run out of other approved masks. Though the goal is to reach and exceed n95 level of filtration, this has not yet been confirmed. Testing is in progress will be updated with any findings. These are a last resort mask, considered equivalent to a scarf. You have to be the one to choose what you wear and I am not responsible for any negative outcomes if you choose to use this. The video and info here are for educational use only.

Collaborations and Thank You's
-Everett from Nevada - this guy helped me do at least 8 revisions of the mask. He is an incredibly talented 3D designer and offered to help me extensively without compensation, just to help us all make a difference in this fight against coronavirus. I had asked for help on youtube and he responded. I couldn't have done it without him. He is to thank for most of the improvements listed at the top. THANK YOU,

-This is a remix of Kvatthro's thing #4222563, in that I started with some of his early files and modified them many times to get to where the designs are today. Kvatthro and I are finally connected and are friends. We video chat about design ideas frequently and share designs and concepts from across the world with each other. He is an incredible guy with a dedicated team in Slovenia. Please see his newest concept to use a ventilator circuit filter that provides over 99.99% filtration, holds pressure at 5 bar without leaks, and provides good saturation. The filter can be changed every 24 hours of use, and are affordable and accessible to hospitals. https://www.thingiverse.com/thing:4242653

-Thanks "Igor & Pascal" - from France comes a second strap loop model, now used as both attachment points

-Please see the remixes - there is a gasket that can be printed from TPU though I have not tried, thanks mrdanielphelan

-The idea for the square to hold the filter in place with minimal glue was sparked by Dr. Premkumar as well as the press-fit piece remix made by TheLastDunedain, thanks.

-ABA-PGT with input from DaVinci and Medtronic for designing and injection molding the new filter box for us. - Bob Hazelton, the speed at which we are making these wouldn't be possible without you. Thank you!!

Donations through GoFundMe: gf.me/u/xsqhra

Links to some of the news coverage:
https://www.wfsb.com/news/local-doctor-using-d-printer-to-print-masks-for-medical/article_de3a805a-72c6-11ea-887e-db303591745e.html

https://www.nbcconnecticut.com/news/local/doctor-3d-printing-face-masks-for-health-care-workers/2246501/

https://www.theday.com/article/20200327/NWS01/200329496

Future plans: flange edges out around face, experiment with dual nozzle printers for TPU material against face while keeping the rest rigid, make the filter box have tabs so the filter retaining square pops in tight without need for any glue at all, and seek NIH approval for community use as is and medical use with added one layer of halyard h600 between mask and face for >n99. Will also upload mask-fit paper templates asap when complete, made for us by Dr. Sean London.

UPDATE 4/13/20: Some may find it tight enough to not need any additional sealant, but a think line of silicone around the connection between the mask and box is still advised. It is easily removable if needed and provides peace of mind about the seal.

UPDATE 4/13/20: ready-to-print .gcode files for ender 3 now linked towards the top.

UPDATE 4/17/20: Filter box is now thicker and much less likely to break. The dimensions have been fine tuned, as users were saying the fit was a little off. Please comment and let me know if the fit between the box and the masks is now working for you.

UPDATE 4/27/20: some people might still be having issues with the filter box fitment. I'm pretty sure I fixed it with the thicker box, but I also uploaded the original filter box file now. Use that if needed.

UPDATE 4/28/20: I uploaded a pdf info document that I'm including in packages I send of these

UPDATE 4/29/20: Per request, filter box CAD files and dimension info can be found here. If the scaling is off and the model is very small, increase by 2540%. Thanks to our design partners ABA-PGT for allowing me to share this so that people can modify to different depths for various filters: https://drive.google.com/drive/folders/1mlc26QV251YRb6mroBG-RZrHo9rdQni7?usp=sharing

UPDATE 5/5/20: We are now using 2 layers of 3M Filtrete MPR 2200 and 1 layer of MERV 13 corrugated filter with wire in it. Our most recent testing is showing this arrives at an n93 under NIOSH testing levels (85 liters per minute), and when worn over a surgical mask, n98.5! These claims need to be verified through repeated testing at multiple labs, but I wanted to get some long-awaited numbers out there of initial findings. This is still for community use, not medical.

UPDATE 5/13/20: I added an updated suggested/example materials list, to coincide with some initial testing results showing that 3M filtrete 2200 in 3 layers MAY exceed 95% filtration (aka higher than n95) and that the Merv 13 layer is now optional, but might provide some other benefits. This is still only approved for community use, though initial testing is showing levels of filtration with the newest filter combination as high as 98% of 0.3 micron particles. Stay tuned.