paper: Ballistic trajectory with air friction drag and Magnus

[Michael Hill]

Pretty interesting. Unfortunately it looks like just a 2D simulation. The 3rd dimension can really play into the aerodynamics, for example, the surface velocity of a spinning ball spins relatively slower the closer you get to the "poles" of the sphere. This would be valid for an infinitely long, smooth cylinder. Not to say this doesn't provide valuable insight into how the Magnus effect affects rotating bodies in air.

[Ether]

Originally Posted by Michael Hill

This would be valid for an infinitely long, smooth cylinder.

It's still valid for a sphere, provided you use the appropriate coefficients.

Not to say this doesn't provide valuable insight into how the Magnus effect affects rotating bodies in air.

Yes, insight. This was the purpose of the paper, to use Rebound Rumble as an "excuse" to get students interested in the physics, math, and computer simulation aspects of ballistic trajectories using a hopefully-not-too-difficult simplified model.

For Rebound Rumble, a simulation was not necessary and perhaps not even all that helpful for the purpose of building a winning robot. This is because, as many folks have already pointed out here on CD, the major design challenge of the shooter for Rebound Rumble was to get a shooter that was consistent from shot to shot. This was a big challenge due many factors, a key one being the variation in the ball due to manufacturing tolerances and changes in the ball's physical characteristics as it got worn. Once you've built the most consistent shooter you can, it's a simple matter to test-fire it to determine its performance.

[Michael Hill]

Hmm, perhaps you're right (capital D instead of lowercase D in CD). However, finding the correct constants aren't so trivial.

[Ether]

Originally Posted by Michael Hill

Hmm, perhaps you're right (capital D instead of lowercase D in CD). However, finding the correct constants aren't so trivial.

If I had a shooter that could launch a ball with the same (known) launch speed and launch angle with and without spin¹, I would try the following:

Set um=0 and adjust ud to match the trajectory¹ with no spin.

Leaving ud at that setting, adjust um to match the trajectory with the expected spin. (repeat to get um's for various spins if needed).

Then I'd launch the ball at various angles (and speeds) and see how well the simulations matched the actual trajectories¹.

It would make an interesting science project.

¹Probably need a high-speed frame-timestamped video camera to determine launch speed, angle, spin, and trajectory