How would you design a cyclocopter?

CampRobber

Active member
Looking good! A few thoughts from things I've read or learned:

(3 and 4) I discovered that racing drone motors work the best. Not sure what you're using, but racing drones tend to have very powerful and weight-optimized motors.

They're diatone mamba toka 1606s and weigh ~16g, or about half that of a standard miniquad motor. I'm just not sure if the rotor I built is better matched by a bigger motor. One problem I kept running into is the t-mount pulley on the 1606 melting. I made those "fan pulleys" but they don't work very well... maybe having the motors hidden from rotor wash is not a great idea.

How long have you continuously run yours at 100W power?

And then I'd recommend taking out that center piece that connects all the pitching linkages. It's back a few pages, but I tried that a year or so ago and it completely messed up the pitching and reduced thrust.

The piece in the middle is TPU and supposed to be flexible in the directions required. You think that's causing problems? It seemed like it was working pretty well. How much torque is there on the control input when yours is running? On mine it was pretty significant, but, nothing servos couldn't handle.
 

2jujube7

Well-known member
They're diatone mamba toka 1606s and weigh ~16g, or about half that of a standard miniquad motor. I'm just not sure if the rotor I built is better matched by a bigger motor. One problem I kept running into is the t-mount pulley on the 1606 melting. I made those "fan pulleys" but they don't work very well... maybe having the motors hidden from rotor wash is not a great idea.
How long have you continuously run yours at 100W power?
I'm using a 2004 motor that weighs 16g, so that sounds about right. Not super long, maybe 20 seconds. It's the peak power consumed, so any type of prop attached at that power consumption will heat up. I made the motor mount and drive pulley out of ABS for higher heat resistance.

The piece in the middle is TPU and supposed to be flexible in the directions required. You think that's causing problems? It seemed like it was working pretty well. How much torque is there on the control input when yours is running? On mine it was pretty significant, but, nothing servos couldn't handle.
The fact that it is now a 5 bar linkage instead of a 4 bar linkage messes up the AOA of the blades.
 

2jujube7

Well-known member
It's been a little bit, but I've rebuilt 2 3D printers that were causing me trouble and holding me back from working on this project. I'm now at a 5:1 thrust/weight ratio per rotor, and I just started back on building up 4 rotors for the second try at a cyclocopter.

I have all the electronics, hardware, and printed parts that I need, so it should really just be making new airfoils and then putting everything together. (y)
 

CampRobber

Active member
It's been a little bit, but I've rebuilt 2 3D printers that were causing me trouble and holding me back from working on this project. I'm now at a 5:1 thrust/weight ratio per rotor, and I just started back on building up 4 rotors for the second try at a cyclocopter.

I have all the electronics, hardware, and printed parts that I need, so it should really just be making new airfoils and then putting everything together. (y)

Any progress? I got a new printer earlier this year and am thinking about another attempt at cyclocopter this winter.
 

2jujube7

Well-known member
Any progress? I got a new printer earlier this year and am thinking about another attempt at cyclocopter this winter.
I have all the rotors built, but I had to stop things for school. I'll be on break starting mid December, so I plan to pick things back up then. I also found a professor at my university that might have interest in research, so after the break I might look at that.

So yeah, not really any progress.
 

CampRobber

Active member
I have all the rotors built, but I had to stop things for school. I'll be on break starting mid December, so I plan to pick things back up then. I also found a professor at my university that might have interest in research, so after the break I might look at that.

So yeah, not really any progress.

Has your design changed much since the last photos you posted? You mentioned a 2004 motor but it looked like you were using something bigger with a 5mm shaft earlier.
 

2jujube7

Well-known member
Has your design changed much since the last photos you posted? You mentioned a 2004 motor but it looked like you were using something bigger with a 5mm shaft earlier.
Yep, I've been through like 3 different motors. I'm currently using a racing drone 2004 motor with a 2mm shaft. It's pretty well designed to be light.
 

2jujube7

Well-known member
It has been quite a while since I've had time to work on this project, but I achieved stability and responsiveness with the PID tuning today. I had the cyclocopter set up on a gimbal for the PIDs, which allowed me to iterate without crashing it. I'm looking to attempt tethered flight tomorrow, and I should hopefully be posting a short video of the results. It's very responsive so achieving stable flight is looking very doable.

@NickRehm - I got some experience this summer PID tuning in dRehm Flight with an 8ft long 3d printed VTOL F35 -- after struggling with the initial learning curve (quite a few crashes) I am quite happy with the software. I worked in some custom code to actuate the F35 nozzle with stepper motors (so that I have control over multiple rotations i.e. more range of travel than a servo) and it worked great!
 

NickRehm

Member
Great to hear such promising progress! Would love to hear more details on the F-35 if you decide to start a thread on it
 

2jujube7

Well-known member
Not quite sure how much I'm allowed to share as I've been doing the F-35 over the last 2 summers as research with the AFRL (Air Force Research Laboratory). It started off from these plans: link, but we designed a custom nozzle that swivels just like the real F-35 nozzle. Other F-35s that I've seen just rotate the rear EDF, so designing and implementing that was pretty fun.

It's pretty funny though -- my boss went to flite fest this year and got pretty disappointed that she didn't meet you when I saw afterwards that you went.

EDIT - video:
 
Last edited:

2jujube7

Well-known member
Attempted a flight, was able to get it off the ground. Probably just didn't have enough altitude to be able to have the room to level it out. Not super bummed that it crashed, as I was at a low throttle on takeoff and was able to demonstrate good control with the flight controller in bench testing.

I've already been working on fabricating the next iteration that will utilize carbon spars in the blades, bigger gear reduction, more voltage, and a 2 DoF offset that will allow me to change the magnitude of thrust while remaining at a fixed speed. (ie near instant control response) I am running my motor on 3s and it is 3s capable, so the focus will be to develop a very stiff cyclorotor so that I will be able to go up in power. I had previously tried a 5s batt, but there was too much slop so it only lead to ~10% increase in thrust.

 
Last edited:

LitterBug

Techno Nut
Moderator
Think I would wear heavy long pants, shoes, and eye protection around that.... But Awesome build! Keep us posted on the progress!
 

2jujube7

Well-known member
Think I would wear heavy long pants, shoes, and eye protection around that.... But Awesome build! Keep us posted on the progress!
Ha, would probably be smart. The foam airfoils wouldn't do much damage but a flying piece of plastic has the potential to cause some damage.

Currently mocking everything up on the CAD for the new design, have some finishing touches to do with the new parts but the main ideas are all there. I have a new transmission - I'm going to couple 2 4:1 pulley systems together to achieve a resultant 16:1. This will hopefully allow me to double voltage (which results in higher motor rpms, so I am converting it to more torque through this reduction). This should also keep the motor heat down.

I also have a new vectoring mechanism, which will be on the inside of the rotor instead of the outside. The new linkage gives me control over 360 degrees of rotation, as well as pitching from 45-20 degrees. Theoretically I can use the same linkage design (with a few changes) to do 45-0 degrees, but it would probably be much less stiff so I'm going to look at that after I verify that this works. The only real downside to this design will be that I will have to include a 2nd servo (SG90 so an extra ~10g). This is not great as each rotor will weigh 90g without the steppers, so that adds on an extra 20% to the weight. Should probably try to find a lighter way to do that, let me know if you guys have any recommendations for mini servos or steppers.
1721671223859.png
1721671169241.png
There will be no center shaft, as the rotor will be supported externally by the frame. I am doing this in an attempt to stiffen everything up, as that was an issue on the last copter. I'm going to take the time to design a stiff frame and adequate ground strike protection for the rotors. Finally, I am including a 3x1.5mm carbon fiber tube as a spar in the blades. This should stiffen everything up slightly, but the real benefit is that I will be able to screw the bearings directly into that instead of needing to epoxy plastic plates onto the foam. Every single structural failure that I've had was because of the foam ripping near the plastic plates, so this should significantly increase robustness.
1721671425547.png
All together, I'm hoping for a significantly stiffer rotor. The last design warped too much when I tried to apply higher voltages (3s -> 6s, got like an extra 10% thrust), so this will hopefully be able to handle that. Oh - and I'm increasing both the wingspan and diameter from 150mm to 200mm. Looking at the non-dimensional coefficient of thrust I should theoretically get ~1080g thrust per rotor instead of the previous ~320. However, it's somewhat thought that the unsteady aerodynamics get less favorable as you increase size, so we're looking at a realistic 750g? Honestly no clue but it's what I'm shooting for. The same 4.7:1 thrust to weight from the old rotor would mean that I really just need to get over ~450g thrust for this to be an improvement in that aspect. I'm also looking at doing an experiment on the aerodynamics side of things that reduced power consumption by ~40% in a CFD, which would be pretty crazy to implement too. Not sure how good it is experimentally though.

tldr; Basically I have new things to try that should hopefully double the thrust to weight? Also sorry for the long post, I'm trying to step up documentation efforts and figured I'd just do that in here because I kinda already stole this thread from Nick.

EDIT - controls: forgot to mention this but the reason I chose the configuration of the 4 rotor was a) to stiffen everything up with as much support as possible and b) to benefit from the change in rpms when leveling. I noticed a significant difference in pitch vs roll stability on the last copter due to increasing rpms leading to torque that levels the craft. On the controls side, I am planning to use only the RPMs to stay level, and use the thrust vectoring for translation. Flight controller/RPMs will keep it level, and the stick inputs will pass through to only translate the copter. This is to highlight the unique benefit of the cyclocopter's thrust vectoring - the ability to translate without changing RPMs or tilting - meaning that it should be much more maneuverable than a quadcopter. Anyway that is all, let me know if there are questions or thoughts on anything.
 

NickRehm

Member
I love the new configuration approach to help stiffen the rotors by supporting them on both sides! Only problem I can think of is there is no means to control yaw with this configuration. Any thought on solutions there?
 

OCPatch

New member
Great stuff. Long time read, first time post. A question: I'm trying to get an understanding of how much power loss is saved using metal ball bearings versus just printing/drilling/polishing nylon bushings. I expect the main shafts really require a proper bearing. But what about out there on the pivot and pitching mechanism? Does anyone have a sense of what the difference would be? Nylon would be lighter - and would possibly build and assemble faster - but would it be worth the extra friction?
 

2jujube7

Well-known member
I love the new configuration approach to help stiffen the rotors by supporting them on both sides! Only problem I can think of is there is no means to control yaw with this configuration. Any thought on solutions there?
I tried the both-side-support previously but didn't try hard enough to get it stiff. I know there's some other cyclocopters out there that have been successful though, so it should go alright.
...
Believe it or not I somehow never stopped to consider yaw 😅. Guess I'll be switching up the design a little bit lol. Might just have to throw another set of rotors onto one of the sides and make it a hexacopter or octocopter. Might be cool because I'd have a big open space at the cg for a payload. Either that or I'll have to go back to using a center shaft.

Great stuff. Long time read, first time post. A question: I'm trying to get an understanding of how much power loss is saved using metal ball bearings versus just printing/drilling/polishing nylon bushings. I expect the main shafts really require a proper bearing. But what about out there on the pivot and pitching mechanism? Does anyone have a sense of what the difference would be? Nylon would be lighter - and would possibly build and assemble faster - but would it be worth the extra friction?

That's a pretty good question. I think the problem is that there is lots of centrifugal force on all of the bearings engaged in the pitching. They're probably experiencing rotation rates that would work for plastic on plastic contact, but they wouldn't last long under the force. For the (12) bearings out on the airfoil I'm using very small and light 5x2x2.3mm bearings that only weigh 0.4g each. So having the bearings on there isn't adding a whole lot of weight. However, I am using plastic on plastic and CF on CF contact for the vector linkages that only rotate slightly when the vectoring direction/strength changes.
 

Scotto

Elite member
Cool progress. Its always interesting to see.
I am trying to imagine if it would be possible to build a big single free rotor cyclocopter driven by props like Nick's giant flying ceiling fan. Driving a fuselage and the linkages in pitch instead of yaw. Maybe drive that fuselage framework from both sides and twist the blades of the rotor for roll? Just dreaming...
 

NickRehm

Member
I think you could pull off a configuration like in your flight attempt video which works for control of roll/pitch/yaw, but also have the rotors braced on both sides. Kind of having the airframe wrap around the rotors like a cinewhoop.

You'd take a weight hit on the frame structure, but the added stiffness I think would let you push the cyclos to higher RPMs for more thrust with all your planned upgrades