Blog 6: Une hélice en beignet
Going subsonic: Why drones have their distinctive hum and how propeller design can help fix it.
Blog 6: Une hélice en beignet
As promised, I finally began my drone testing, starting with my Push/Pull and toroidal propellers. I’ll save the results for the end of the blog, but this week’s central topic will be about what I want to test and how best to do that with design and print considerations. As became evident when printing a normal drone propeller, the inability to create overhangs without support material meant that either the bottom or top of every sharply-angled propeller would be littered with imperfections, a big problem for thrust (more drag and vortexing) and noise (higher pitched sounds of bumps cutting through air).
This is another reason Push-Pull propellers are helpful. Since one side of each is flat, I was able to place the filled out side on the print bed and produce a flat, support-free propeller. Having a flat side worked for propellers designed to isolate performance characteristics, but for a real propeller, which needs to take advantage of pushing and pulling to perform well, This solution is counterproductive.
The solution I’ve found has been to use toroidal propellers for my comparative tests. Toroidal propellers have a few factors that make them much, much easier to print than a typical bladed propeller. First off, they have flat surfaces on the top and bottom, meaning they can hold onto the print bed at every point on the bottom (more on this later). Next, because of the loopy nature of the toroidal propeller, the two halves of a blade can connect to each other, making the entire structure stronger with the same amount of material. The gentle slope on my design also affords a support-less angle of 45 degrees or less overhang, the suggested angle for 3D printers. Finally, something that is not immediately obvious, the slimmer shape of the propellers makes them much faster to print and easier to remove once finished. I never discussed this, but a large surface on the bed plus a rigid structure means that peeling the print off the printer is extremely difficult and leads to pain for the remover. A toroidal propeller just needs a squeeze on one propeller and the entire surface unsticks immediately.
With my adoration of loopy propellers out of the way, I must discuss some downsides that I have experienced or expect to have in the future. First is material strength. My first design for the propellers had 1.5mm thick walls, which made them heavier, though more rigid. To cut weight, I shed off some thickness to make 1mm thick blades, but with the 2-blade version, the spin speed was enough to tear an entire blade off, making it dangerous and unusable. I fixed this by reverting to 1.5mm walls, which has been damage free so far. Next, paradoxically, is the problem of bed adhesion (now is later). The blades are so thin that they can peel up from the surface, ruining their print shape and possibly coming off the printer and making a mess. I fixed this by adding a brim around the print. This is just an area one layer thick that adds bed adhesion without causing problems with getting stuck like my traditional propellers had.
Now for our results. This may come as a surprise, but the push and pull propellers were both unideal, possibly due to making a heavier propeller with worse air interactions. My toroidal propellers performed exceptionally, creating strong fan-like winds and only exploding across the room once during my trials. I was able to experience the low droning sound typical of toroidal propellers, which indicates success in making a more pleasing propeller, if not mathematically quieter. I have not taken exact measurements yet, as most of my time has been focused on printing, designing, and coming up with more test variable ideas, which will be reflected eventually in my spreadsheet.
Propeller spreadsheet (Changed, but still no data): https://docs.google.com/spreadsheets/d/1QnVORPgaP6eOXAQWGW8OxpOkJtlu77fsbD5rLiyG7hs/edit?usp=sharing.