Best Way to Launch unusual items?

[Dominick Ferone]

Originally Posted by Scott Kozutsky

Have you tried more power? You can relatively easily get catapults to launch really far and really fast. You can flatten the exit angle for it to go faster but not farther.

33 using a catapult to full court shoot in 2014
https://www.youtube.com/watch?v=1-oWAS9_Rl4

Don't know if that works for a hockey puck trying to reach 100+ mph

[Chris is me]

Originally Posted by Dominick Ferone

Don't know if that works for a hockey puck trying to reach 100+ mph

Building a generic shooter that can accept objects of basically any size or shape and propel them to 100 mph is a really, really, really tall order. A catapult launching generic objects at even as fast as 40-50 feet per second is a lot more possible.

[The other Gabe]

I'd suggest a linear punch shooter with an adjustable holder (like 1986, pink team, 948, etc in 2014). Really easy especially if the accuracy doesnt have to be that great

[carpedav000]

If you want 100+ mph, I would suggest air cannon. Just use a sebot to account for the various object shapes/sized.

[Dominick Ferone]

Originally Posted by carpedav000

If you want 100+ mph, I would suggest air cannon. Just use a sebot to account for the various object shapes/sized.

what is a sebot? and one of the small scale prototypes we will test tomorrow is an air cannon.

[Chris_Ely]

Originally Posted by Dominick Ferone

what is a sebot? and one of the small scale prototypes we will test tomorrow is an air cannon.

A sabot is a device used in a firearm or cannon to fire a projectile, such as a bullet, that is smaller than the bore diameter, or which must be held in a precise position.

Please remember to consider the safety of whatever mechanism you choose. You will be dealing with a lot of energy with whatever device you use.
In the case of air cannons, do not use PVC pipe to store gases under pressure.

[Dominick Ferone]

Originally Posted by Chris_Ely

A sabot is a device used in a firearm or cannon to fire a projectile, such as a bullet, that is smaller than the bore diameter, or which must be held in a precise position.

Please remember to consider the safety of whatever mechanism you choose. You will be dealing with a lot of energy with whatever device you use.
In the case of air cannons, do not use PVC pipe to store gases under pressure.

Yea we knew not to use PVC since we were looking into making a T-Shirt launcher.
And with whatever design we do were making a two stage safety where two buttons will constantly have to be pressed otherwise it wont shoot or shoot very slowly.

[asid61]

Footballs can be launched with just two spinning wheels for very large distances. Look up "football launcher" on youtube for a demonstration.

[GeeTwo]

Originally Posted by Dominick Ferone

The argument was started while going over the design of our hockey puck shooter on whether we should go:

air powered(paintball gun like)
inline wheel shooter(ultimate ascent)
top wheel shooter with roller bottom(baseball shooter)
Spring loaded plate(pinball)
Slingshot

or is there something else more efficient to shoot the put accurately at really high speeds for our challenge.

For hockey pucks, I would constrain the puck to be flat with some sort of chute and use a dual-wheel (left/right) shooter. If you want them to fly rather than slide along the ground, you could make one of the wheels rotate a bit faster than the other to give some gyroscopic stabilization.

Also, I recall a linear accelerator that I rough-designed for Aerial Assist that used a motor and gearbox, two linkage arms, and an angled glide with a basket that the ball sat in. The point of the two-armed linkage was to match impedance between a motor/gearbox running at a constant speed and a load which was accelerating. If you aren't interested in picking up strange pieces (or will do this separately), this would be a useful design. We didn't build it (used a hammer for AA), but I am pretty sure we could have gotten the yoga ball high enough and far enough to make clear truss and high goal within the required frame perimeter. By making the travel longer and gearing the motor lower, any reasonable speed should be feasible. I'll look for the drawings and spreadsheet this evening.

[GeeTwo]

Originally Posted by carpedav000

If you want 100+ mph, I would suggest air cannon. Just use a sebot to account for the various object shapes/sized.

You can do the same with pneumatics, using the piston as the sabot. Remember that for this project you can use larger diameter solenoid valves and fittings than the FRC standards!

[GeeTwo]

Originally Posted by GeeTwo

Also, I recall a linear accelerator that I rough-designed for Aerial Assist that used a motor and gearbox, two linkage arms, and an angled glide with a basket that the ball sat in. The point of the two-armed linkage was to match impedance between a motor/gearbox running at a constant speed and a load which was accelerating. ..... I'll look for the drawings and spreadsheet this evening.

I was unable to locate the original drawings and spreadsheet, but I have re-generated them in the attached document.

The linkage arm connected to the motor/gearbox is 7 units long, the shaft of the motor/gearbox is 11 units perpendicularly away from the glide rail, and the secondary linkage arm is 20 units long. The "layout diagram" shows what the linkage arms will look like for each 20 degrees of rotation of the gearbox. The position-speed-accel-power shows these curves as a function of angle through the cycle, assuming that the motor/gearbox is rotating at constant speed, the mass of the load is constant, and motion is horizontal. The goal in designing the offset and linkage lengths was to produce a flat "power" curve that covered as many degrees as possible without requiring too much acceleration (and likelihood of jamming) on the return stroke. This one gives a nearly flat power curve from 40 degrees to 160 degrees. Note that the "power" for the braking and return stroke are shown as being about six times as high as the main forward stroke; however, I am assuming that the load being thrown will be several times heavier than the carriage, and that the rail will be angled (presumably at somewhere near 45 degrees), so the actual power output is much closer to a constant. If you use encoders, you can select the throwing range simply by setting a constant rotation speed for a PID loop, so it should easily scale down to lighter loads than its peak.

If you provide a desired maximum load and launch speed/throw range, I'd be happy to figure out the appropriate motor requirement, gear ratios, and arm lengths to achieve it.

And to repeat - I have not built this yet.

[Dominick Ferone]

Originally Posted by GeeTwo

I was unable to locate the original drawings and spreadsheet, but I have re-generated them in the attached document.

I must say I am confused by the spreadsheet, then again I am not to sure what I was looking at.

[GeeTwo]

Originally Posted by Dominick Ferone

I must say I am confused by the spreadsheet, then again I am not to sure what I was looking at.

This is a two-linkage-plus-glide configuration to launch projectiles using a motor. Launching with a motor is usually inefficient, as the load begins at rest and ends at maximum speed. A motor produces no power at rest or its maximum speed, but produces maximum power at approximately half speed. This setup allows a motor spinning at a constant speed to apply essentially constant power to the load as it accelerates from zero to top speed.

First, set up some sort of glide with a carriage to support the load. I was thinking of bearings on a rail topped by a light basket, but other configurations can be used. For purposes of the calculations on the spreadsheet, I set this rail along the y=0 axis. (which is the top of the frame in the layout diagram). In practice, you would probably rotate the entire assembly about forty-five degrees counter-clockwise.
I then placed the output shaft of the motor/gearbox at y=-11 units and x=7 units; these numbers are the first row in the spreadsheet and can be seen as the "hub" towards the lower right in the layout.
Attached to the motor/gearbox shaft is a linkage 7 units long; this is given in the second row, and is seen as the "star" around the hub in the layout.
Finally, connecting the motor linkage to the glide carriage is another linkage 20 units long. Ths length is specified in the third row. This is constrained to start at the end of the motor linkage and end along the y=0 axis. I calculate the left end point using the end of the motor linkage and the pythagorean theorem.

Spreadsheet, main data table:
Column A is the motor linkage rotation angle in degrees. I started at -10 so I could calculate acceleration at zero, and proceeded in five degree steps through 360. As I plotted this in the layout diagram, zero is when the motor linkage is horizontal to the left, positive angles rotate the motor linkage clockwise.
Column B is rotation angle in radians, used for sine and cosine calculations.
Columns C and D are the x and y coordinates of the motor/gearbox hub, repeated from row 1.
Columns E and F are the x and y coordinates of the end of the motor linkage. These are calculated from columns C and D plus the sin and cosine of the rotation angle times the motor arm length.
Columns G and H are the x and y coordinates of the glide end of the carriage linkage, calculated from the coordinates of the motor end, the constraint that y=0, and the pythagorean theorem.
Column I is the speed. Since I'm assuming constant rotation speed, this is simply the change in Xpos. I've taken the liberty of scaling up by 10 so that the graph looks nice.
Column J is the acceleration. Since I'm assuming constant rotation speed, this is simply the change in speed from the previous angle. Again, I scaled by 10.
Column K is the power, assuming a constant inertial load. Since power is force times speed, and force is mass times acceleration, this is simply speed times acceleration. Negative power implies that the load is being decelerated. Again, I scaled, this time by 1/5 to make a pretty graph.

The Position-Speed-Accel-Power plot simply plots columns G, I, J, and K vs column A.

The main power stroke is between about 37 degrees, where the load is stationary, and 170 degrees, where the load is at maximum speed. The carriage then stops by 250 degrees (resulting in the load flying off the end), and the portion from 250 degrees back to 37 degrees is the return stroke.

Note that the y location of the motor hub and the linkage lengths were selected by a manual tuning process to produce a flat power curve during the main stroke. x=7 was selected simply so hat the x output position starts at zero at the beginning of the power stroke after I had put in the other three numbers. You can change these numbers in rows 1, 2, and 3 to try different configurations, though you will need to turn off some of the axes limits on the graphs for them to be visible.
If you find graphs with sharp kinks in them, this is a sign that you are asking for something that is not going to fit, or require excessive forces. For example, set the carriage arm length to 18 units - you will get a large spike in acceleration and power at the very low angles. If you look at the layout diagram, you can see that this is because the carriage arm becomes vertical and must reverse direction very quickly. If you were to actually build this device, the carriage would probably overshoot this point or jam.

If you get lost, the original values were: -7, 11; 7; 20.

The secondary data table (columns M through AV) are tabulations of the x and y coordinates in a format more convenient to produce the layout diagram.

[Deneb]

Typically, the best way is to find how humans thrown them most effectively and mimic that motion. There is freedom with throwing spherical objects such as catapults, "punches", or wheels. Frisbees offered a fairly simple solution where nearly everyone agreed that wheels would be the best for giving a spinning motion to allow the disk to fly. However an object such as a(n American) football, would (theoretically) need the spiral motion applied then a human throws in addition to a forward thrust as well. Personally, I don't see the spiral being enough without additional thrust being applied.

It would also be important to explore other aspects unlike the human process. Humans don't throw T-shirts with pneumatics--robots do.

[DarthCoder]

We used a hammer last year for throwing the big exercise balls. It was pretty powerful and worked pretty well, but you have to be really careful your holder arm doesn't get hit by it like ours always did.