("Supplementary Data" to you science types.)
Ensuring a good fit:
In occupational settings, the all-important fit of the mask to the face must be professionally tested before use and every year thereafter. Here is another link re fit testing (both last accessed 02-06-2021). I got my mask at a hardware store outside of an occupational context but did a DIY fit test with powdered stevia sweetener puffed out of a rubber bulb just after giving the bulb a good shake. The stevia seems to lose its sweetness slowly after removal from the packet, so I grind it fresh for each test. Will try saccharin next. (02-20-2021: saccharin test passed, but self-administered.) NB: I have two years of experience in wearing a respirator occupationally. If you are new to this, you should practice a lot in a safe setting until you know what a good and bad face seal feel like and the moves you need to get a good one. My big discovery was that moving the mask up or down on my face often did the trick.
I learned how to wear a respirator in a paint shop where there are smelly fumes that you can smell if they leak into your mask. This facilitated the learning process and I am surprised that OSHA relies on the sense of taste for fit testing when using smell would be so much more convenient (no need for a "nebulizer" that often blocks up) and would make a more stringent and foolproof test. You just swap organic vapor cartridges for filter cartridges for the duration of the test, then swap 'em back afterward, and use a harmless odorant such as rose water.
Making a flared exhalation tube:
I first prepared a 6-cm dia. plastic funnel by sanding the places where future hot glue would go. I cut off the hanger tab and then encircled the funnel with a bead of glue to help hold the filter. I then cut off the stem and drilled small holes near the bottom to pass the hold-down wire. I then assembled the flared tube on the mask and used hot glue to seal the seam between the funnel and the straight part of the tube (cut from a 1.5" dia. aspirin bottle along the top and bottom of the label). I then secured the ends of the hold-down wire with more glue. (See below). The straight part of the tube fits snugly into the exhalation valve without leaks, but getting it to go in required a combination of force, accurate alignment, and stuffing in the parts that boinked out. A lubricant might have helped; haven't tried it. Something else to try would be filing the end of the tube on a bevel.