I'm working on a fairy costume and I'd like to build a pair of damselfly-type wings that realistically, randomly flutter when folded back and can also fully extend and flap. I found a few videos that give a clear view of the kind of motion I'm going for:
The rig needs to be lightweight, battery powered and as small as I can manage, as this will be a completely mobile costume, not a pneumatic rig or anything. I'm planning on running everything off of the Arduino platform, as I am already very familiar with it. The mechanism is really the part that's challenging me at the moment. Thoughts?
It's definitely doable. I've been working on several wing designs for people over the last month so I've been looking at wing mechanics/kinematics a lot lately. When thinking about the mechanics the first thing you need to consider is how light and strong you can make the wings. The weight of the wings will determine which servo motors you use.
It appears from the videos that you could get away with just four servos as the wing movement from the folded back position to forward flapping mostly happens in one plane. I don't think there's a real need to have wing twist as that would greatly complicate things. Basically you just need one servo mechanism for each wing. The key here is to think about how the servos and wings will be attached to your costume
The flapping isn't a problem but a fluttering motion will be difficult to achieve as most servos can't move that fast.
I'm glad to hear someone else thinks this is doable, lol. I'm planning on keeping the wings pretty light - a simple framework with cellophane as the membrane. I have a good amount of experience with servos, so I think I can manage those calculations. The part I'm stuck on is the physicals engineering of the mechanism to transfer the rotary motion of the servos to the actuation of the wings. I've considered several options - linkages, cams, pulleys, etc. - but this level of engineering is pretty new to me, so I'm not sure what the best way to proceed is.
As for the fluttering, I wasn't picturing it being very fast, more of the partial opening movements seen in the video when the damselfly is at rest.
I think you can do it strictly using the rotary motion of the servos. I don't think you really need to translate that motion into anything more complex. The trick is in mounting the servos in the correct orientation and making sure they can support the wings directly. You might check out these servo mounting blocks that have a support bearing and extended shaft for the servo output spline-
If you need something more low profile than that then you'll have to build a wing mount (driven by a separate servo) that has the proper amount of movement. It's certainly possible to combine multiple servos to drive a mount that will allow for more complex motions. I can help with the design of that.
Those servo blocks are pretty interestingly! The example they have of the pan/tilt servo combination seems promising. I am a little worried about it getting too bulky, though. I'd like to keep these as streamlined as possible in order to simplify integrating them to the costumes.
Yeah they are pretty big. Since your wings are going to be so light you could probably use mini servos like the Hitec HS-225MG analog servo or the HS-5245MG if you want a digital servo. Use one servo to rotate a swivel hub at the base of the wing. Then have a hinge mounted to the hub (make a small adapter plate) that allows the wing to flap and drive that using a second servo. You could mount the flapping servo to the base of the wing itself (simplest and most reliable, but would require a solid base for your cellophane wing) or have have the flapping servo actuate the wing flap using a cable (more complex but potentially cleaner since you could hide the servo.)
If your wings are super light you could have the flapping servo mounted to the swivel hub and then mount the wing base to the servo output shaft using a shaft coupler. As a general rule I don't advocate doing this as it as servos really aren't designed to have a side load applied to the output shaft but you can get away with it for short term use if the load is light enough and you use a ball bearing servo with metal gears. Also standard size servos can handle much higher side loads than mini servos.
Had to step away from this for a little while due to a serious case of life, but I'm digging back into it now. I'm thinking of trying something like this:
I figure this would allow me to hide all of the electronics down in the bustles of the skirts, letting me use larger servos and batteries, running the cables up the back to the pivot points of the wings. For the pivots, I was picturing a slightly off-angle assembly that would recreate the sweep of the wings with the pull of the cable.