J. Phil Thien's Projects

The image shown here: http://www.cwimachinery.com/dustfx-2hp-cyclone-dust-collector-review-item-cwi-dcp020h/ shows a kind of "trash can with funnel on the bottom" thing. Which is a cyclone.

If you subtract the impeller on top and go by the height of the inlet, I'd say it's about 1D1D, which is actually a pretty bad cyclone, comparatively. But still, it clearly is one. It will serve OK for shavings and large particles (eliminate 100% of shavings and ~99% of large particles), but it will be rather bad at separating the small particles (expect 30-40%, not 99%). That's going to be the HEPA filter's job. In order to properly separate small particles in the cyclone (i.e. with a ~95% rate), you would want a cyclone approx. twice the height.

The same will be true for the Thien separator, if you put one before this cyclone. The Thien separator is not super awesome because it separates so well, let alone small particles. Actually it's pretty bad at that, too. It is super awesome because it does not need to be 2 meters tall, and yet it can catches a very sizeable amount of matter. A larger outlet will impact its efficiency, yes. But that's not really something to worry about. What comes out is not high quality air anyway. Nor is the air that comes out of the cyclone high quality. This is why one needs a HEPA filter.

Less matter in the air overall means that the cyclone's efficiency will be increased. Fewer particles entering the cyclone means even fewer leaving it. Fewer particles in the air after the cyclone means even fewer particles after passing the filter, and a longer lifetime of the filter, too.

Normally, for a separator, the recommendation would be to keep both small because a smaller inlet and a smaller outlet improves the separator's efficiency. For the pipe/hose, it's not so clear what the "general" recommendation would be because there is a delicate play between pressure loss and volume/airflow.

However, luckily things are not that complicated here, figures seem to be pretty clear.

The DustFX already has a cyclone built-in, so the purpose of the trash can separator would not be being super efficient, but merely reducing the heavy particulate matter load. This enhances the cyclone's efficiency and shifts its cut-point downwards, which is a good thing (both for your HEPA filter and your lungs). But even the worst separator in the world will catch upwards of 95% of heavy particulate matter, the difference in that class is just not noticeable at all. It is the small and very small particules where you figuratively separate wheat from chaff, and the baffle separator isn't very good at dealing with these anyway (nor is that intended).

Insofar, it doesn't really matter whether the outlet is a bit wider (and thus the pre-separator is a bit worse). On the other hand, not changing the pipe diameter without need means to not needlessly introduce turbulence and pressure loss, which is a good thing. Plus, it's easier, cheaper, fewer parts, fewer leakages. So for the small bit between separator and DC, I'd just keep it at 6 inches.

Generally, having a thick pipe after a thinner pipe is not such a brilliant idea because that means that the airflow drops, and if you have a somewhat naive build with an insufficient blower, you may have heavier particulate matter drop out, clogging the pipe. That isn't a problem here, however. Unless the manufacturer is lying, the DC has almost twice as much volume per time unit than is needed to achieve 20m/s on a 150mm pipe (which is a "good" value). Also, after passing the pre-separator, the air doesn't contain a lot of heavy particulate matter any more, so there really isn't anything that could drop out in the first place. Therefore: don't worry.

About the 4 inch hose, I wouldn't worry, either.  The DC is supposed to provide "9 inch water lift" as per the technical data on their website, which is about 2230Pa, so a pretty common figure. Most DCs that have pipes anywhere from 4 to 7 inches are in that ballpark. Unless your hose is super, super, insanely long, it'll pull through those 4 inches just fine. Since you have that diameter on the hose already, you can as well make the inlet on the trashcan the same. Smaller is better anyway, so going from 4'' to a 6'' inlet, which is extra work, doesn't make a lot of sense. You would generate extra turbulences, entry velocity, and reduce the separator's efficiency. Not like efficiency really matters for the pre-separator, but why make a sacrifice when it's actually easier to just use the smaller diameter right away.

You only need about 40mJ/cm³ of light at 254nm to get a 4 orders of magnitude reduction. That's not quite sterile or perfect, but it's pretty good, and trivial to achieve.

Masks do not even achieve 2 orders of magnitude, and they're quite effective already. Viral load has been, and is an important factor with every comparable disease. As long as you stay significantly below the threshold, you are good to go. Breathing in a few hundred virus particles doesn't affect you. It's millions of them that you need to fear.
That's the reason why masks don't need to be 100% perfectly good to actually work -- contrary to the abundance of pseudo-scientific fake news that claim that they're useless (which was even spread by governmental officials in Feb. 2020 here). They're not useless in any way because they just do that one thing. They shift your exposure from "way above threshold" to "below threshold". Not always, not in all conditions, but most of the time, in most situations.

With the power necessary to operate a heater and a vaporator and a cyclone, you can do 100 times the dose in UV light needed to get 6 orders of magnitude reduction in your ventilation pipe. So that's basically "sterile as good as it gets" inside the pipe. UV germ reduction is a solved problem, having been in use in the food industry and in laboratories for many decades. All that's "revolutionary new" is that you want a somewhat higher dose to be on the safe side. Not suitable for COVID gold digging, sorry.

A Dyson is ridiculously small to fight fine dust as you've described. This will not work, whether or not you add a cyclone to it.

Also, a Dyson is in my personal opinion a pathetic vacuum which is more designed to fool people with snake oil than to actually work. It has cyclone magic powers with a total of 14 "cyclones". Only just, the design doesn't look like these will actually do any separation at all (air blowing straight down into them at an angle that actually cannot conceptually work). The only really good thing about the Dyson is its HEPA filter.

For your problem, you will need a much bigger (10-20 times) airflow than anything the Dyson will produce. And you will need a professional to figure out the very non-trivial problem of getting the flow exactly right. Because, you know, more is always better, but more is not always better. More airflow will eliminate dust better, but at some point, the airflow itself impacts quality of life. Also, getting the right suck/flow without a noise level that will drive you insane is a very non-trivial problem.

As for a cyclone, this can be used to pre-separate coarse and some of the fine dust, enhancing the HEPA filter's performance and lifetime. But you will need a HEPA filter in any case, too. Also, you will need something that "sucks" the right way. Vacuum cleaners and shop vacs are usually built to provide a significant pressure gradient because you may need to suck in bigger stuff which is lying on the floor, and that just doesn't work otherwise.
Dust collection and even moreso air purification doesn't need to suck up large particles so urgently, but it needs a high air volume (the CFM figure is primarily what one needs to look at, the pressure is not that much important, although of course you must be aware of pressure loss due to ducting and the cyclone). After all, what it does is filter out particles that are light enough to be in the air on their own already.

Cyclones isolate larger particles (such as PM10 and up) much better than smaller ones. Unluckily, it is not the large particles that we are usually interested in as they tend to eventually fall to the ground anyway, they're easy to cope with, too, and not a health issue (your nose, if nothing else, filters them). It's the smaller ones (PM2.5 and PM1) that we worry about.

Depending on the build and on the load, you can expect a no-crap cyclone to remove anywhere from 80% to 95% of the particles that we are mostly interested in. For PM1 and below it's much less (like 50%), but luckily, to our advantage, there is not so much of these in the air, comparatively. So, overall, we get out with 80-95% in the range that is important to separate. That's enough to enhance your filter's lifetime, but it is not enough to be considered "safe, problem solved".
For that, we  want a 5-sigma or 6-sigma reduction, and for this you will need to use a HEPA filter after the cyclone. There is no other way. Even then, most affordable HEPA filters will only achieve 6-sigma for the combined range, for very small stuff they often only achieve 98-99%, too (or worse). There's unluckily some legal leeway in the test procedures, too, so even what's printed on the filter is usually only 95-99% truth and 1-5% lie. But, that's as good as it gets.

The most important technical term for what you are interested in for the cyclone is called "cut point", that's the size of particles of which 50% get separated (anything larger gets separated as well). In the end, you want more than 50% of course, but research uses that figure, so what we are looking into is "pushing down the cut point" in order to actually push the total amount of stuff that we don't want closer to zero.

Things that affect the cut point are, among others:

• the load (so e.g. in the shop where you have kilograms of particulate matter in the hose, it makes a lot of sense to add a pre-separator)
• the speed of airflow (more is better)
• the cyclone's diameter (more is worse)
• the cyclone's length (more is better, but only up to some point because of pressure loss)
• the size of the outlet relative to diameter (less is better)
• the size and geometric proportions of the inlet (smaller and slimmer is better, a 7/5 ratio being considered the optimum)
• the angle, if any, of the inlet (for example Pentz cyclones achieve higher separation due to the explicit downwards lead-in)

What works quite well (70+ years of use in industry and agriculture) for fine particles is what's called 2D2D or 1D3D style separators. These are cyclones where the upper, straight bit is either equal to (1D...) or twice (2D...) the diameter in length, and the tapered bit is either twice (...2D) or three times (...3D) the diameter, the 1D3D separator being only marginally better alone, but significantly better in series. There's an awful lot of mostly-Chinese research on the topic with chaining 2D2D-1D3D appearing to be the best performer.