-
Notifications
You must be signed in to change notification settings - Fork 234
Breathability Testing
Since we began testing the filtration performance of different materials, one question that has come up is breathability. How much effort does it require to breathe through the material? If the material is very good at blocking particles, but it is too difficult to breathe through, it makes the material makes not useful. Breathability is the amount of negative pressure required to achieve a target air flow through the material. Less pressure required means the material is easier to breathe through.
The basis for our test setup was this paper. However, because we don't have access to the same equipment, we made a several simplifications. We also didn't have large sample of all materials that were flat and could be consistently wrapped over a large intake port.
For the auto-updated testing result, please go to this page.
Here is a picture of the test fixture taking a sample. Here's a video of the sampling process.
Below is a diagram of our breathability test setup:
The simplifications are:
- We couldn't find a regulatable particle generator. The fog machine and humidifier both trigger over-saturation of our particle counter. Our attempts to attenuate the amount of particles from these sources were very inconsistent, even in mixing chambers. So, we just used ambient household air on the intake.
- Our particle counter uses a small pump to pull air into the sampling chamber, but when we had the particular counter on the negative pressure side of the filter, it was giving us erratic readings. Likely, the pump in the particle counter was not strong enough to overcome the negative pressure and possible air was being pulled backwards through the sampling chamber. So, we decided to separate breathability from our existing particular filter results at 3L/min of flow. Ambient air samples are taken before and after each material sample.
- The reference paper uses sampling area of ~59cm^2. However, we didn't have large flat samples of all of our materials and pre-formed masks are hard to flatten without creating wrinkles that would create leaks. So, our sampling area is 20mm in diameter, which is approximately 3.14cm^2. A smaller area will require higher negative pressures to achieve the target flow rate. However, it is our current opinion that this should still lead to valid relative breathability results when comparing against different materials. 3M N95 mask are included as a reference. This smaller area allowed us to be more consistent with clamping treatment, the amount of surface area engaged, and minimized leakage due to inconsistence sample holding.
- Our target flow rate was 90L/min, which 3.17 ft^3/min also from the reference. Which is comparable to a strong breath.
- The controller is an Arduino Nano
- the blower from an old Respironics CPAP that I had. It's possible to usd CPAPson eBay for around $80 or just a similar blower for around $30 on Alibaba. My blower could be controlled using a standard hobby brushless DC ESC like this one, but I think many would work.
- the flow sensor is a Sensirion SFM3200 and is a fairly easy to use 5V I2C device. You can get this from Arrow or Digikey.
- pressure sensor is an MS5611 24-bit barometric sensor. This is a 3.3v I2C device.
- The display is a 4 row LCD Display also very easy to interface with over I2C.
- I used a variable DC power supply set at 16v volts to give the blower motor plenty of power. Note, the Arduino Nano's voltage regulator has a maximum of 16v, so if you set the output voltage too high. You may break your Nano onboard regulator.
- As a button, I used a long lever switch because they are easy to press.
- The sample ring clamp uses a medium sized spring work clamp
- the remaining parts are 3D printable. This is the main body and the ring clamp part1 and part2 and bolt flange.
- This manometer was used to help check the MS5611 was working correctly. But once the sensors were verified to be aligned, this was no longer used.
I use an online tool called EasyEDA to create my circuits. Here is the link to this project. Since we have couple of 5V I2C devices (LCD display, and flow sensor) and a 3.3V I2C device (baro sensor) we need several I2C SDA/SCL pins, 5V, 3.3V, and ground connections. There's also connection for the button, and the blower ESC.
It's a relatively simple circuit just with lots of connections. I created a PCB since I have a small mill, but this is very optional. Though even with the PCB, it's still looks a bit messy with all the wires. It is helpful to use double-sided foam tape or [VHB tape[(https://www.amazon.com/3M-Scotch-5952-VHB-Tape/dp/B01BU7038A) to hold at the parts steady, and prevent damage/shorts from manipulating the wires and connectors too much. Connectors are a common failure point of any electrical system.
Here's a link to downloadable files.
The main chamber is attached to the intake of the blower and liberal amount of hot glue to ensure and air-tight bond. The pressure sensor is also hot-glued into place in bottom port, covering the hole. The top port with the threads was used to connect the manometer to verify the measurements from the pressure sensor. This was sealed off after verification. The bolt flange is assembled as 4 pieces that which surround the intake port. This gives a clamping area to hold the material samples.
The spring clamp (with the rubber tips removed) is used with the ring clamp parts to help secure the sample around the intake port with repeatable pressure around the rim. The 3D printable parts have slots for the jaws of the clamp. Then hot glue is used to hold everything in place. The bottom part of the clamp has V-grooves that center on the rounded bolt heads on the flange. This helps self-center the ring over the intake port when you can't see it under the material sample. This improves repeatability and ease of clamping a sample with one hand.
The firmware for the Arduino Nano is located here. It initializes the flow sensor, the pressure sensor (zeroing on boot up), and the blower ESC. Then, it switches over to measurement mode. Pressing the lever switch cycles between target flow rates. It is set to 0L/min, 34L/min, and 90L/min per the reference paper. It uses a very simple PD control loop to adjust the blower speed so that the measured flow rate will reach the target flow rate. The LCD display shows the target flow, current measured flow, current pressure in the main chamber, and the whether the system has stabilized and is ready for a reading or if it is still adjusting the blower speed.
Here's a video of the sampling process. Once the motor speed stabilizes, it is ready to read the measured pressure required to reach the target flow rate. A lower negative pressure means more effort is required to pull air through the sample material. So the closer this number is to zero, the more "breathable" it is. The pressure required for a 3M N95 mask is included in the results for reference. The relative pressures are more meaningful than the absolute pressure measurements.
It only takes 3-4 seconds for it to adjust the blower speed and stabilize, making it fairly quick to sample multiple areas of the material, and switch materials. The pressure is very quite a bit based on how tightly held the material is held, if there are wrinkles and seems in the sample area. Even moving to different parts of the sample without any visual differences can result in 10% changes. So how the mask fits on your face, and how much surface area is being used will greatly affect the breathability of a material.
Results - link
We are continuously testing new materials as people send them to us, for the auto-updated report please go to this site.