Ornithopter R&D Thread

[Balto-the-Wolf-Dog]

Given I haven't seen an operable one yet, I thought it might be worth a thread to share ideas and progress in developing a working ornithopter. I'm about three hours in myself, and so far my main prototype has managed about 11.5 m/s before the batteries die. It doesn't seem to be very short of rotation speed at that point, but isn't quite capable of it. Some of my more adventurous designs have pushed 14 m/s, but weren't very stable (and were hideous). All so far have taken a relatively long period of time to accelerate.

My current configuration features dragonfly-style single-joint wings with opposing strokes. The strokes themselves are biased toward the upstroke: the angle of incidence rotates more steeply up on the upstroke than it does down on the downstroke, and the bulk of the dihedral range occurs above the neutral position for gear clearance reasons (60 degrees above 90, ten degrees below 90). Balanced strokes would probably be more effective, but haven't tested that yet. Wing area seems to relate fairly directly to acceleration (assuming the stroke remains the same), but the 14 m/s cap I've been seeing seems to result primarily from the stroke topping out and not beating fast enough for its net thrust to overcome drag. Basically a prop-pitch issue, but for ornithopters. The most effective workaround I've found so far is to offset the wings away from the fuselage so they're forced to travel a greater distance over the course of the stroke, but so far it hasn't been enough to fly, and it has a habit of baiting the kraken.  Increasing the range of the wing incidence curve would probably help too, but I haven't found a good way to do so without outstripping the movement speed limits. 

I'd wager by now somebody has a much better design than this one, but I thought I'd put it out there

Spoiler

Edited June 13, 2019 by Balto-the-Wolf-Dog

[Geschosskopf]

Fascinating subject and thanks for all the good intel

I've avoided messing with flapping wings due to early experiences with trying to implement tilting autogyro rotor disks.  I quickly discovered that hinges simply lack the strength to resist the aerodynamic forces imposed by the rotor disk, so just get blown around uncontrollably.  I would thus expect similar problems with flapping wing joints, which would severely limit both on the size of the wings (and thus the vehicle), plus the speed of both the wings and the vehicle as a whole.  Thus, I expect working ornithopters would be in the ultra-light size range.

[KerikBalm]

On 6/13/2019 at 3:17 PM, Geschosskopf said:

Fascinating subject and thanks for all the good intel

I've avoided messing with flapping wings due to early experiences with trying to implement tilting autogyro rotor disks.  I quickly discovered that hinges simply lack the strength to resist the aerodynamic forces imposed by the rotor disk, so just get blown around uncontrollably.  I would thus expect similar problems with flapping wing joints, which would severely limit both on the size of the wings (and thus the vehicle), plus the speed of both the wings and the vehicle as a whole.  Thus, I expect working ornithopters would be in the ultra-light size range.

I had similar problems with tilt-wing VTOLs, and I now stick to tilt-engine VTOLs. Its a shame, I'd like to be able to tilt my wings up as well as I try to slow down for landing.

Maybe something more conservative like the wing hinge/incidence change on an F-8 is doable.

[Daniel Prates]

Hmm.... 

RL wings generate speed when air flows through them. KSP mimics that relating speed to lift, as it should, but I doubt the under-the-hood calculations will consider that moving wings up and down generates lift....

.... or does it? Has anybody made an ornithopter fly yet? Like REALLY fly? KSP physics can be weird, if an ornithopter is gaining flight for a few seconds that may be just that it went all bouncy, not that the wings are actually generating lift.

[HebaruSan]

See:

 

[Daniel Prates]

4 hours ago, HebaruSan said:

See:

 

Sweet Jesus, I didn't think this would be possible with KSP physics. 

Well, that's that then. Time to fly to a desert sietch and recruit some Fremen. Long live the fighters!

[Geschosskopf]

9 hours ago, Daniel Prates said:

Hmm.... 

RL wings generate speed when air flows through them. KSP mimics that relating speed to lift, as it should, but I doubt the under-the-hood calculations will consider that moving wings up and down generates lift....

The key to understanding robotic-based non-fixed wings in KSP (as opposed to parts with a Module_Engine thrust-generator), whether using them for rotors, props, or flapping them, is that the KSP "air" has no substance.  Thus, nothing from the real world that relies on pushing air 1 way to go the other way works in KSP.  The only thing that works in KSP is moving wing surfaces in such a way that the aerodynamic lift they generate simply by moving through the "air" is of benefit.

[Daniel Prates]

It just crossed my mind that, in that case, a paddle-wheel type of thing could very well work like a gyroscope of sorts and cheat gravity!

[GameTourist]

On 6/23/2019 at 4:39 PM, Daniel Prates said:

Well, that's that then. Time to fly to a desert sietch and recruit some Fremen. Long live the fighters!

You've worn a stillsuit before?
Your desert boots are fitted slip-fashion at the ankles.
Who told you to do that?