The paper on Ingenuity talks about the test setup of ingenuity

Associated metrology in the form of a force/torque sensor, a Vicon motion tracking system, stroboscopic lighting, a thermal camera, and temperature sensor monitors were developed to support operation in the chamber. The helicopter and test stand were instrumented with monitoring accelerometers and thermocouples. An infrared camera provided another means for monitoring the helicopter’s performance.

[emphasis mine]

I understand the stroboscopic lightning as the type of closed loop lightning system that would make the rotor appear still.

Why would such a system be required? Why would one want to see the rotor as still?

  • $\begingroup$ At a wild guess, it allows you to see flexing and twisting modes of the rotor blades. You probably don't want too much of that, or at least too much of the wrong kind of motion. $\endgroup$ – Starfish Prime Feb 26 at 10:48
  • $\begingroup$ Given that those motion have frequency much lower than rotor frequency [ which will be equivalent to stroboscopic lighnting], else those flexing motion will be aliased $\endgroup$ – Prakhar Feb 26 at 10:49
  • 1
    $\begingroup$ Stroboscopic lighting can be used to test the rotational speed of a device. The frequency of the stroboscopic light can be set up be the same as what the rotational speed of the device should be. If the device appears to be motionless all is well, if not variation in the rotational speed can be detected as either an advanced or lagged rotation. $\endgroup$ – Fred Feb 26 at 10:52
  • $\begingroup$ @Prakhar if you were interested in those motions then you'd surely use a strobe with the appropriate frequency to image them correctly. $\endgroup$ – Starfish Prime Feb 26 at 10:56
  • $\begingroup$ @fred I think surely ingenuity will be having some inbuilt way to measure that, a simple IR sensor or encoder shaft will do the trick. But is thay the sole reason? or as pointed out it is for analyzing flexing modes also? $\endgroup$ – Prakhar Feb 26 at 11:00

If you illuminate the rotors with very brief, very bright, flashes of light then you can use a camera with a much longer exposure suitably synchronised to the strobe flashes to take pictures of them, and they will appear stationary. This means you can get a good idea of what deformations the blades are undergoing. This is a common technique for looking at rotating systems of any kind where you want to get a good understanding of what's happening.

It's much technically easier and cheaper to make very brief, very bright flashes of light than it is to build camera systems which will take very, very short exposures (which will in any case require an enormously bright light source), and if you have a setup where you can conveniently do that then it's generally a better solution.

You want to synchronise the flashes of light with the rotation for two reasons. Firstly you will be very interested in what's happening with the blades at very specific angles, such as when the blades pass over each other, when they are at right angles to each other, when they pass over bits of the body of the vehicle and so on. Secondly by running the strobe synchronously with the rotors you can film successive passes of the blades through the same position, and thus catch slower changes in them: for instance, there might be periodic deformations of them which happen at multiples of the rotation period, or deformations which are slower than the rotation period but unsynchronised with it, for instance, dependent on the length of the blades as waves of deformation of various kinds pass up and down them and so on.

Note that these slower-than-rotation changes can happen even if the blades have characteristic frequencies much higher than the rotation speed, for at least two reasons.

Firstly the system is not just the blade: it's the blade, the shaft, the bearings the shaft runs in, the motor, and the joint between the shaft & blade. That joint can, for instance, deform under various loads, and may end up having play in it (you hope not I imagine, but that's why you're doing experiments). In any case the resonances associated with this whole system could be very complicated, and could change significantly under load.

Secondly even if the blade has characteristic frequencies much higher than the rotation frequency there can still be beat frequencies: if the blade has some characteristic frequency $f_B$ and the rotation frequency is $f_R$ then $f_B - n f_R$ is a possible beat frequency for any $n \in \mathbb{N^+}$. Because the system will be, probably, fairly nonlinear these beat frequencies can be real things, and you want to know if they matter.

Finally by running the strobe very slightly faster or slower than the rotation period you can use it to make, essentially, movies of the blades rotating in slow motion, which you can then look at to see if interesting things are happening.

Something I forgot in an earlier version of this answer is that one of the very interesting things to look at is the pitch control of the blades, especially the cyclic pitch control.

  • $\begingroup$ So flexing motions are generally subharmonics? $\endgroup$ – Prakhar Feb 26 at 11:02
  • $\begingroup$ @Prakhar: That was a mistake, sorry: I'll fix it. I meant 'changes which are slower than the rotation period' not 'changes which are an exact multiple of it, but then I forgot that's what I meant. $\endgroup$ – tfb Feb 26 at 11:13
  • $\begingroup$ Later in the paper they mentioned that blades were designed to be very stiff resulting in flapping motion of much much higher frequency than rotation. I think there might be some other alternative explanation also $\endgroup$ – Prakhar Feb 26 at 11:15
  • $\begingroup$ @Prakhar: the blades may be designed to be very stiff: you still want to know if they are. Additionally it's not just the blades: it's the whole system of shaft/motor/bearings/joint/pitch-control/blade. I've added some more text about that. It's worth noting (which I now have) that this is a completely standard approach to looking at rotating systems: it would be really surprising if they didn't do this. $\endgroup$ – tfb Feb 26 at 12:07

Supplemental answer:

Let's ask ourselves

  • How does Ingenuity start moving forward?
  • How does Ingenuity steer?

Then answer:

  • The same way a normal terrestrial helicopter does and not the same way a quadcopter does. From this (frighteningly thorough) answer to Why Ingenuity drone has 2 coaxial rotors, not quad-rotor:

The way a helicopter works (and the way Ingenuity will work, as you can see if you look at NASA's 3D model of it is that the pitch of the rotors gets adjusted dynamically to control the lift (and also to create forward motion &c), while the angular velocity of the blades stays more-or-less constant.

It runs at about 2400 RPM or 40 revolutions per second, and whenever it wants to have a sideways force the pitch of the rotors oscillates at 40 Hz so that there is more lift on one side and less on the other, the same way a helicopter seems to "lean forward" before starting to move forward. The tilting takes some of the upward thrust and directs it forward.

So in addition to checking for vibrations, order to check that the pitch changes smoothly and correctly during one complete 25 millisecond trip around the hub and get good depth of focus and no blur, they need to photograph with both a very bright light and a very short exposure. Strobes provide both; a means for a short exposure (as pointed out in the other answer) and a very bright light for the short time that the strobe is turned on.

I was going to say the short time that the capacitor discharges through the xenon tube but I don't know if solid state strobes are used now for this particular kind of metrological photogrammetry.

screen shot from Veritasium's "This Helicopter Is Now On Mars!" screen shot from Veritasium's "This Helicopter Is Now On Mars!"

screen shot from Veritasium's "This Helicopter Is Now On Mars!"

cued at 05:58 for the discussion of "collective and cyclic" pitch modulation:


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.