Ornithopter. Things lock up at speed.

[Klapaucius]

Over the past couple of days I have been thinking about how to build an ornithopter. I've got a basic idea here, the problem is, I cannot refine it because when I try to get the wings flapping at speed, they just lock up.  Evidently the physics model does not like this.  As you can see in the video, in the SPH it works fine, but on the runway, not so much. Slow cycles are fine, but as soon as you crank up the speed the rotors stop doing their full turns. A bit faster, and the hinges just fibrillate, for lack of a better word.  Changing the servos to the larger ones did not change anything.  Any suggestions?

 

 

Edited June 3, 2019 by Klapaucius

[fourfa]

What are the speed and damper settings on your hinges and pivots?  What I see on the high speed test on the runway is the sequencer trying to drive them faster than they can respond.  I'm guessing that no matter what, there will be a delicate balance between driver power and speed, inertia of the driven parts, and aero drag of the driven parts in the flight scene.  None of these will be visible when simulating in the SPH/VAB.  Finding the right controller speed to max out these values will be highly iterative!

At some point it'll be interesting to look at how a whole bunch of small, lightweight "hummingbird wings" compare to a smaller number of large, slow "condor wings"

[Klapaucius]

2 hours ago, fourfa said:

What are the speed and damper settings on your hinges and pivots?  What I see on the high speed test on the runway is the sequencer trying to drive them faster than they can respond.  I'm guessing that no matter what, there will be a delicate balance between driver power and speed, inertia of the driven parts, and aero drag of the driven parts in the flight scene.  None of these will be visible when simulating in the SPH/VAB.  Finding the right controller speed to max out these values will be highly iterative!

At some point it'll be interesting to look at how a whole bunch of small, lightweight "hummingbird wings" compare to a smaller number of large, slow "condor wings"

After I posted, I realized that there is a speed setting on the servos as well which does not seem to matter in the SPH. Having said that, while that improves things, you cannot get anywhere near the speed I would like.  I've since tried doing this on Duna. Thinner atmosphere but far less gravity to overcome. I'm leaping off the ground, but honestly I am not sure whether flight is involved or whether it is just the physics of thrusting the wings down quickly causing a reaction. It always seems to return to the surface.  I wonder if a craft like this is actually possible in KSP or if the physics model does not account for it, in the same way that a paddle in the water does very little as well. 

Also, stability is a nightmare. I have not at all got a handle on that despite multiple reaction wheels.

Edited June 4, 2019 by Klapaucius

[RocketSquid]

Somewhere in the BG previews they posted a video of a (barely) working ornithopter. In general it seems like large wings are the way to go, since all the robotic parts are very limited in terms of maximum speed. In general, an ornithopter or prop plane is unlikely to get anywhere near the speed of a jet, and this is pretty reasonable. 

Oh, and speaking of paddles, they work... honestly a lot worse than ornithopters. Paddle wheels are somewhat effective, however. Screw drives work if you make them out of the right parts and make sure the boat doesn’t flip.

[Snark]

Moving to Breaking Ground Support, since this seems like a BG issue and the devs are more likely to see things that are located there.

[Klapaucius]

4 hours ago, RocketSquid said:

Somewhere in the BG previews they posted a video of a (barely) working ornithopter. In general it seems like large wings are the way to go, since all the robotic parts are very limited in terms of maximum speed. In general, an ornithopter or prop plane is unlikely to get anywhere near the speed of a jet, and this is pretty reasonable. 

Oh, and speaking of paddles, they work... honestly a lot worse than ornithopters. Paddle wheels are somewhat effective, however. Screw drives work if you make them out of the right parts and make sure the boat doesn’t flip.

So what do you use for screw drives?  I got a whopping 2m/s out of a twin screw with this throw together test:

19 hours ago, fourfa said:

What are the speed and damper settings on your hinges and pivots? 

What does the damper do on a rotor?  Is it some sort of bezier curve for rapid acceleration?

[RocketSquid]

54 minutes ago, Klapaucius said:

So what do you use for screw drives?  I got a whopping 2m/s out of a twin screw with this throw together test:

What does the damper do on a rotor?  Is it some sort of bezier curve for rapid acceleration?

My early tests used much smaller boats with the basic fins, three per rotor. They weren’t fast, but they were a bit faster than 2 m/s. The paddle wheel was actually faster, but it still wasn’t anywhere near jet boats and it produced a good bit of lag with the water. At the same time, though, both of these were made to be kinda fancy and were thus very unoptimized. I’m certain a more optimized boat could do better, both in terms of hull design and rotor pitch optimization. I feel like in KSP it might be the opposite of in real life—a longer screw may be more efficient.