mid air collision

[Ether]

Here's a fun little spreadsheet.

Put your launch speed, launch angle, and firing rate into cells L2, L3, and L4.

Column D will show the gap, in inches, between the balls.

If the gap is negative, the cell will be red.

This model just uses parabola equations for simplicity. Air drag is ignored.

You can easily select a relevant portion of columns B and C and graph it to see the balls appropriately spaced in mid-flight.

[jkelleyrtp]

Interesting tool, very useful for determining potential issues with current prototypes.


I turned it into a Google Sheet and added a y_0. I guess it would be fairly easy to expand this to have drag factored in which probably has a significant impact on midair collisions.

Link to Sheet

Just click "Save as Copy" to edit it yourself.

[SenorZ]

Whoa whoa whoa.

"Air drag is ignored." - Ether (2017)

[Ether]

Originally Posted by Ether

Here's a fun little spreadsheet...

... and here's a fun little gnuplot script that plots 5 inch diameter circles with spacing corresponding to the launch speed and firing rate.

1) download and install gnuplot

2) run the script !

[efoote868]

Originally Posted by SenorZ

Whoa whoa whoa.

"Air drag is ignored." - Ether (2017)

"All models are wrong but some are useful."
-George Box

If there isn't enough spacing without air drag, there certainly won't be enough spacing with air drag

[Ether]

Originally Posted by efoote868

"All models are wrong but some are useful."
-George Box

Indeed.

And some models are useful for engineering, while others are useful for teaching/learning.

the spreadsheet shows how to use parametric equations to get an XY plot with points a constant delta time apart.

[MrRiedemanJACC]

Love this tool. Thanks Ether. I was playing around with it last night and this is pretty curious.

# User Parameters:
V = 35.0 # launch speed ft/sec
q = 82.0 # launch angle degrees
bps = 12 # balls per second

35 feet per second is approximately what a cim motor on a 4" wheel and efficiency losses included. If you can feed at 12 balls per second (ludicrous I know and keep the speed up on the motor (2 cims maybe for the torque and a flywheel?) They actually won't hit each other. Now that means the balls are going practically vertical, so the affect of the air on it's downward path will be interesting. It make not take a very good trajectory on the downward side. Once we are down with our production gear loading / unloading, and the climber, this is our third priority, so we'll know with testing.

But until then it makes me think will a flatter path or a more upward path work better?

Any thoughts? Ether? Richard?

[Richard Wallace]

Originally Posted by MrRiedemanJACC

Love this tool. Thanks Ether. I was playing around with it last night and this is pretty curious.

# User Parameters:
V = 35.0 # launch speed ft/sec
q = 82.0 # launch angle degrees
bps = 12 # balls per second

...

Any thoughts? Ether? Richard?

What goes up must come down. In this example, the FUELs will peak ~12 ft. above the 21" diameter HEG. As you observed, they might not all come down by the desired path.

[billbo911]

Our testing, so far, confirms and alingns accuratly with this tool. Longer, flatter shots do indeed have less tendency to interact with each other.

That said, there is a variable that is difficult to compensate for, the fuel have holes. Getting consistent, repeatable contact with the same part of the fuel is next to impossible to do at a rate that is reasonable to be competitive.

The fuel fly fairly straight when fired with backspin, but the apex of the parabolic path can be different from shot to shot because of the non repeatability of fuel contact. Thus the trade offs in choosing which shot to take come down to:
Can you aim reliably enough at a longer distance and repeat the distance of the shot enough to successfully make shots at a higher rate than shooting close and dealing with fuel collision will produce?

[Ether]

Attached is the gnuplot graph and script for 35 82 12.

As you can see, the balls are pretty close to each other at the top.

I agree, at that speed and angle (and length of flight path) air drag will be a factor.

This is mostly a teaching and visual-aid tool. You really need to prototype and test !

[billbo911]

Originally Posted by Ether

... This is mostly a teaching and visual-aid tool. You really need to prototype and test !

^ Absolute truth! ^

[IKE]

I would recommend tweaking the tool timing down to 0.125 seconds. I have found with realistic values, that the smaller time window gives a lot of insight.

[MrRiedemanJACC]

Originally Posted by Ether

Attached is the gnuplot graph and script for 35 82 12.

As you can see, the balls are pretty close to each other at the top.

I agree, at that speed and angle (and length of flight path) air drag will be a factor.

This is mostly a teaching and visual-aid tool. You really need to prototype and test !

Yes absolutely agreed on the testing. We did some with a wooden mock up to prove to ourselves we are in the ball park with a mini cim and weighted flywheel style shooter. Now we are finalizing our gear and climber (our priority), then I have two students working on the shooter hard design. We didn't use an encoder for speed yet, so we need to get that going and then we can get some usable data on the shot. The plotted data will give us something to go off of to start the testing and help set our ranges. Then we will see from there!

Thanks for the input everyone!

[Skyehawk]

Thoughts on what would be involved to accommodate for a 2,3,4,5? wide shooter? Assuming the balls do not significantly interact between 'layers' I suppose you could used a constant feed per 'layer' or a constant feed with a random component per 'layer' to simulate rapid evacuation of a hopper. So it would essentially be running one simulation per 'layer' and plotting them all at the same time.