Masks, again

I assume that until this bug is over, it’s not over, and until then I will wear a mask around other people, and I would like it to be a good one. I keep reading papers. I’ve learned that an earlier paper about the goodness of high-thread-per-inch cotton had a flaw, where they failed to neutralize the charge on their test particles, which made it (very) easy for a mask to trap the particles and this overstated the effectiveness of tight cotton weaves. More recently, I read about the performance of mask materials over time, and that was depressing; if you use an medical N-95 mask very much at all, it probably loses a lot of its ability to filter small particles because the electrostatic charges on the fibers go away. Medical wrap (Halyard H600) has the same problem. This is especially true with alcohol, which is a polar solvent that can conduct electricity.

What this chart says is that for the best protection, you want a new N-95 mask; nothing else is better, and the different between 99% percent efficiency and 98% percent efficiency is twice as many particles, so a real N-95 really is much better. If, however, the mask is worn more than once, especially if it is exposed to alcohol in small amounts (for example, from hand cleaner) then it might not work nearly as well. For a mask that I’m going to wear more than once, “EX101” looks interesting; it turns out to be a high-grade air filtration media used to make air filters for diesel engines. It does, however, resist breathing more than other materials.

Two other things matter, a lot, in a mask. One is how much it resists your attempts to breathe through it, and the other is how well it is sealed to your face. Once you get to 99%-of-particles-filtered efficiency, small gaps can reduce the relative performance of the mask by a lot. In masks I’ve made in the past with high-thread-count cotton, if they get damp from sweat or condensate, they fail in both ways; their resistance to breathing goes way up, and when you exhale, the elastic cannot keep them sealed to your face and you blow unfiltered air out the sides (you can feel it, it is very annoying). On the plus side, this is a validation that the mask design I use can seal tight. I’m a little nervous about masks that only hook on ears; I’ve worn some of them, and they don’t feel like they’re sealing that well, and shopping, a few times I have seen people wearing such masks where from the side I could see between the mask and their face, all the way to their lips and nose.

Another problem with mask made of materials that resist breathing is that they can collapse onto your face, which reduces their surface area and makes inhaling even harder. It also feels disgusting if you’ve been wearing the mask for a while and there’s some amount of condensate in/on it. However, I figured out how to modify an existing design to let me prop it off my face with coffee sticks, and I’ve since improved it; the coffee sticks are inside now, which looks better and is a little simpler to sew, and more of the seams are on the inside, not the outside

So, this is how I got to where I am now, making masks from two less-filtering fabrics surrounding an inner layer made from EX101, and washing them with alcohol as necessary. I wear those masks whenever I go indoors, and for short/slow bicycle rides. For long or fast bicycle riding, the breathing resistance is too high, so I either don’t do that, or wear a mask that is less good (but I am outdoors, moving quickly, and staying away from other people by default).

Here’s the pattern and the instructions, 2 pages, print single-sided, the first page on card stock if you can, and a series of photos from me making a mask using these instructions. Cummins filtration sells 50 packs of EX101 filtration media (it’s about $45). There’s other mask information at the University of Minnesota, where they have other designs that use this filtration media.

And to the obvious question, why am I doing this, why am I not buying a mask instead? Two reasons. First, any N-95 mask that I know is trustworthy and good, I am probably taking out of the hands of someone who needs it. You might say, “but I see a lot of those for sale, I don’t think there’s a shortage”, but that may just mean that we have different levels of trust. Second problem is, how much do you trust the people selling you the mask? Anyone can claim anything, I look for signs that they are or are not thinking about what a mask is supposed to do. When I look at the masks that are for sale, if I see one with ear loops — I don’t really trust it, because an N-95 that isn’t sealing, isn’t really N-95. Or I see a mask (STILL!) with an exhale valve on it, which means that they’ve completely lost the plot with respect to stopping this epidemic. Or I see a mask that claims to have an insertable filter, but it just looks like a cloth pocket, and really, is that going to seal to my face? There’s no regulations about what is sold to the general public, there’s standards, but is anyone checking claims? We’re very much in the land of caveat-emptor capitalism, and right now the way I caveat, is I read papers, buy supplies from trustworthy sources, make my own masks. (Do I trust Cummins Filtration? Pretty much, especially given the independent research into their filter media. It makes plenty of sense that an engine air filter would work for a long long time.) This is a market failure, government could help, but for a little while longer, our government is crap.

Anyone that knows me, if they sew, if they want some of the filter media, I currently have more than I need, ping me. Either research one of the U Minn mask designs (there’s a weird open-source license they want you to do, it’s odd, but harmless, just takes a few steps) or I’d recommend mine. No guarantees, of course, only I did my best in the time available to do it, my evaluation is it fits my face, I can tell when it leaks, it doesn’t collapse when I inhale, I can get enough air through for moderate physical activity. If I could figure out how to give it more non-collapsible surface area without adding too much to the volume, I would, but this works well enough.

Here’s a newly-made mask using the design above, showing the two props, the interior seams, and the loops for attaching laces/elastic:

I’ve done a few experiments along the way. Here’s a picture (below) of one such experiment; it’s a mask that I use for longer/harder biking, that has no filter and so outdoors only, stay away from other people. I had just finished biking a few miles outdoors in mid-30F weather, so there’s a lot of condensation (much below freezing, long term, expect ice — that’s what ends up in my facial hair in a normal winter. Facial hair, bad for a mask seal). Experimentally I tried to use more props, on the theory that (1) aligning them with the elastic/cord attachments would be “better” and (2) two props per side would do a better job of keeping the mask from collapsing. I decided against this; I didn’t particularly notice “better”, and the prop higher on my face can actually be a little irritating because it rides directly on top off my cheekbones and is still a little hard through the fabric. This photo also shows how much condensate ends up in a mask doing physical activity in a cold outdoors. I’m not too surprised by the wet/dry condensate patterns, except that there’s more than I expected down near the bottom where it catches my chin.