Tuesday, July 16, 2019

An Electrostatic Nanoparticle (?) Precipitator for UMMD

This is a project I started and then abandoned.  I recommend you don't do anything similar.  The idea was to capture particulate emissions from my printer using an electrostatic precipitator (ESP).  As the project progressed, I kept reading more and more scientific papers about the process and about the type of device I was using.  In the end, I came to the conclusion that an ESP that emits ozone is a very bad way to capture nanoparticles because the ozone will react with everything in the environment and may produce nastier stuff than it captures, including more nanoparticles!  I have added links to many of the papers I was reading at the end of the post. 

Please note- absence of odor is not a reliable indicator of the efficacy of a filter.  There are plenty of harmful things that you can inhale that have no odor.

What follows is stuff I was writing as I was working on the project:



In the last few years, there have been several research studies of the particulate and gas emissions from 3D printers (see the list below for some papers of interest), many suggesting unhealthy levels of both, especially if you print ABS, though at this time the long term health effects are unknown.

As a result, a lot of people are trying to make air filters that will capture the scary nanoparticles and volatile organic compounds (VOCs) produced by 3D printers.  Most go the route of using HEPA filters made for vacuum cleaners to capture particles and activated carbon filters to capture VOCs.

The one thing they all have in common is a lack of any objective measurements of the results.  Instruments that can count nanoparticles in the air are uncommon, expensive, and few people know how to use them well enough to get valid results.  So amateur attempts to mitigate 3D printer produced environmental air pollution is a guessing game at best.  And no, your nose is not an adequate instrument for testing, unless your only measure of success is elimination of odor.  Maybe your filter works, maybe it doesn't.  Maybe it captures the nanoparticles, or maybe it only captures the bigger particles that are currently assumed to be less harmful.

A different approach


I found a few research papers on air scrubbing systems that are used to remove nature's nanoparticles, commonly referred to a viruses, from the air in clean rooms and research facilities.  They use a combination of electrostatic precipitation (ESP) and "soft" (low energy) X-rays to electrically charge the particles and remove them from the air.  ESP's are commonly used to remove dust from the air in homes and commercial buildings, and to scrub particles from smoke stacks in industry.

In one paper, the author made comparative tests of the efficiency of electrostatic precipitation alone vs electrostatic precipitation plus soft x-rays.  He tested it at different voltages in the precipitator and found that above about 8kV, the ESP alone approached 100% efficiency at capturing the nanoparticles.  At lower voltages, the ESP alone wasn't so efficient and the soft x-rays, presumably because the tinier particles don't always get charged in the ESP, pushed the efficiency back up to 100%.

ESPs can be made very inexpensively.  Why would anyone want to go to the trouble of adding the soft x-rays, greatly increasing the expense of the system?  At the very high voltage where the ESP is 100% efficient at particle capture, there will be some corona discharge (sparks).  That corona does a couple things.  First, it appears that it manages to apply a charge to even the tiniest nanoparticles so they can be removed from the air, hence 100% efficiency at particle capture.  The other thing it does is produce ozone.

Ozone is triatomic oxygen and is reactive with many things in the environment including VOCs.  It also makes up a pretty large part of the brown haze in the air over cities on polluted days and isn't very healthy to breathe.  Ozone is commonly used to remove odors from homes that have had fires, gruesome criminal activity, and unfortunate accidents that result in bad smells caused by VOCs.

Oxygen prefers to be O2, not O3, so ozone happily gives up the extra oxygen atom to almost anything nearby that's willing to accept it.  That means ozone is unstable and and has a half-life of just a few minutes.  As temperature increases, the half-life decreases, so inside a heated 3D printer the ozone produced won't be around for long.  Hopefully, the extra oxygen will attach itself to VOCs, breaking them up, instead of attacking the rubber drive belts.

ESP construction


The image below shows the construction of the ESP used in one of the papers I've linked above and below.

Diagram of ESP with Soft Xray emitter used in this study.
It's just a metal tube with a wire running down the center, and has I/O for air flow.  Pretty simple.

How I built It


I chose to make a similar thing, but without the X-ray emitter.  I arranged a 40mm fan at the end of a piece of metal pipe (the collecting electrode) about 32 mm in diameter, and a thin wire down the center for the negative electrode.  I used a 12V to 20 kVDC converter, purchased for $10 via ebay, to provide the necessary electrical charge, and stole 12V from one of the DC-DC converters in the printer that I set up to do stuff like this.

I wanted the whole thing to be easy to clean, so I built it so that the pipe could easily be removed without having to do any major disassembly.

After a few failed and suboptimal attempts, I settled on a design printed in six parts.  There's a mounting bracket to hold the assembly on the printer's Z axis frame, an end cap, spring bar, a HV mount, a HV contact, and a fan mount.

The bracket has a ridge that fits into the frame t-slot and there's a single screw/t-nut to hold it in place.  It has slots for zip-ties that will hold the rest of the assembly in place.

The bracket screwed to the back of the Z axis frame and waiting for the rest of the assembly to be mounted.

The end cap fits on the top end of the pipe and holds the spring bar that puts tension on the central wire electrode.  The end cap and wire connection have to be removed to take the pipe out for cleaning.

This is the end cap and spring bar that is used to tension the central wire electrode.  The spring pulls on the wire and prevents is from touching the pipe.


The HV mount is a close fitting tube into which the pipe electrode slides, and also mounts the HV converter module.



The HV contact part fits over the pipe holder and the pipe and has a spring that makes contact with the pipe when it is inserted into the tube.

The HV contact has a spring inside that touches the pipe when it is inserted into the assembly.


Finally, the fan mount has the electrical connection for the central wire electrode, an air baffle that forces the air coming into the pipe to spin, and holds a 40 mm fan to blow air through the whole assembly.

This is the fan mount.  The blades force the air to spin as it flows through the pipe.  The negative electrode wire feeds through the hole in the center.
How do you mount a square fan on a round tube?  Fusion360 makes it easy using the loft function.  I drew the square-with-rounded-corners fan shape on one sketch and about 40mm above it, I drew a circle that would become the outer surface of the printed fan and tube mount.  Then I used the loft function in the "create" menu to connect the two as a solid, and finally, I used the shell function under the "modify" menu to hollow it out.  The resulting print varies smoothly from the square fan to the round pipe.  I used the same function to make the blade inside the fan mount that twists the air flow.

Here's the assembly set up for initial testing.  Left to right- 40mm fan, fan mount, HV contact, HV mount, end cap, spring spring bar.

There were a couple problems to deal with in this design.  I needed the pipe and tube to be easily removable for cleaning so I couldn't solder the HVDC wires to them.  Making the electrical connections foolproof and reliable was a bit of a challenge.  Also, I wanted it to be very easy to reinsert the pipe even if I couldn't see down inside the assembly because of its position in the printer.  That meant I had to design it to guide the pipe into the correct position to make electrical contact without effort.  I ended up with a spring as the electrical contact for the pipe.  It sits in a groove at the bottom of the pipe holder and when you push the pipe into it, the spring contacts the pipe that was sanded to bare metal.

The central wire electrode is just soldered to the HV lead coming from the converter module.  The end of the wire has a loop that hooks onto a spring at the far end of the pipe.  The spring is held in place by a printed plastic end cap and the removable spring bar.


Here's a look down the pipe with the HV converter running.  You can see the purple glow of the corona discharge along the central wire electrode.  It produces a little bit of fresh-smelling ozone that will hopefully break down VOCs from melting plastic in the printer.

Does it work?


I'll be printing ABS with it over the next few months and see if there's any ABS-stink while and after it runs.  I'll run a clean rag through the pipe to see what sort of particulate stuff it manages to pick up.  I don't have anything to count nanoparticles in the air, so this will be like everyone else's build-it-and-hope-it-works approach.


Relevant articles (some may be pay-walled):


Acute health effects of desktop 3D printing (fused deposition modeling) using acrylonitrile butadiene styrene and polylactic acid materials: An experimental exposure study in human volunteers


Characterization and Control of Nanoparticle Emission during 3D Printing

Ultrafine Particle Emissions From Desktop 3D Printers

Characterizing 3D Printing Emissions and Controls in an Office Environment

Destroy VOCs (Chemical Pollutants) at their Source | SanusAer Ozone Generators