Supersonic Propeller

Along with the bull-whip project, this project will aim to break the sound barrier at the push of a button.

Here i’ll simply get a “propeller” up to a supersonic tip speed.

The first thing is to figure out how fast a small propeller needs to go. A simple excel spreadsheet tells me how fast the tips need to go for an size of 2 bladed propeller. Furthermore it will give a rough estimate of gearing needed for a given motor kv and voltage.

At 12” diameter, the propeller speed need only be 22,000rpm. That’s actually surprisingly achievable. Handheld routers get to 35,000rpm.


I could use the same gearing i just used very successfully for the solar boat, but in reverse, gearing up from a medium speed to a very high one. Alternatively i could use a much faster spinning motor, and direct drive the propeller.

Neither presents much of challenge at this point.

Interestingly, the propeller shape is very different from what i’m used to building. This one shouldn’t create any lift, and needs to be very easy to pass through the air. The profile chosen is a double ogival, since that’s apparently quite good for supersonic flight. Interestingly, the reason we use wings with fatter rounded leading edges is to do with variable angle of attack. If the angle of attack is known, a sharper leading edge is a better shape. If we advance through the medium with a sharp wing edge at the wrong angle of attack, we create a very low pressure area in the shadow of the sharp edge, breaking up the laminar flow.

Upon simulating the 12” propeller for strength against the centrifugal force it’ll see, i was a little shocked to see a safety factor around 0.5 in the worst places (didn’t manage to record the result). It might have yielded, and spread the load out a little more evenly, but dang. It really would have sucked to get hit with a sharp machined aluminum edge at 700mph.

This lead me to a feature i’ve never used in Inventor. The parametric FEA functionality. It’s really just a batch processor for FEA simulations, but it’s damn useful for situations where you want to avoid a manual search. This is likely on the simplest end of what this tool should be used for, however.

I instructed the application to iterate through a range of distances from the wing tip for a control point on a spline that controls the guide rail for the main wing loft. This meant it was adjusting the concavity of the wing’s top and bottom faces. Above is the graph of the maximum Von Mises Stress in each simulation. The trend is clear, and the accompanying safety factor increase was significant.

The resulting control point location for the loft guide rail is illustrated below.

master_case3.jpg
IMG_20191011_214329.jpg

First tests with an off the shelf propeller (10”), were somewhat successful. It didn’t disassemble, but only got to around 14,000 before it drew 60 amps from the motor controller. That was a result of having gearing that didn’t really make all that much sense. With an 18v power supply, the duty was around 50%, so the effective voltage was only 9v, while drawing 60amps! Decreasing the gear ratio would allow a faster spinning motor, which allows a higher voltage, so with constant power, draws less current. Work smarter, not harder…i think.

This also drove me to finally cut the designed propeller which should cut the air more effectively, and cause no lift induced drag (due to having no angle of attack). Perhaps i’ll be able to cheat my way to 22,500 with that propeller and the same gearing, but perhaps i’ll have to decrease the gearing further, and perhaps move to a belt drive to reduce the gear noise.

Build

Initial fabrication went well, and some unloaded testing worked out alright. I do think i am underestimating just how dangerous this thing will be though. Tip speed will be in excess of 750 mph, and that’s in bullet territory. I’ve got 2” of plywood in the surrounding ring, but a bullet would likely get through that plywood. I don’t know if i have a way to simulate that impact right now, so an abundance of caution (ha!) is the main method for protection. Keeping everyone and anything important out of the plane of the rotation is pretty important until i experience a failure and asses the fallout, and/or test it enough times to be confident it won’t fail. I’m thinking about installing at least one band of rolled steel around the perimeter to (try to) keep any debris/shrapnel inside should the propeller disassemble and break through the plywood.

Testing

First round testing went pretty well. No explosions, and lots of noise. No “booms” to be heard though, rather an incredibly loud, high pitched whine.