Catapult issues with boulders

[Brandon Holley]

The left video (by my naked eye) appears to be sitting in the catapult cradle just a smidge longer, which to me would point to it shooting a tad lower against the tower.

Is this what you are seeing?

-Brando

[Darkseer54]

Originally Posted by Brandon Holley

The left video (by my naked eye) appears to be sitting in the catapult cradle just a smidge longer, which to me would point to it shooting a tad lower against the tower.

Is this what you are seeing?

-Brando

Yes, the left video is the new ball, while the right one is the old ball.

[Brandon Holley]

Originally Posted by Darkseer54

Yes, the left video is the new ball, while the right one is the old ball.

Catapults are finicky mechanisms. We went through a very intense and crazy design process with our 2014 catapult. When we started considering using one for this year's game (for consistency), we very quickly implemented two design rules based on our 2014 learnings.

#1. Pneumatic powered, no questions asked. This rule was easily accepted by having a nice small and lightweight game piece.

#2. Roll off the tips. This was a huge part of our 2014 design and something we jumped on right away for our 2016 version. By giving the ball a predictable spin we found it helped keep it on a more predictable trajectory.


When it came to diagnostics, slowmo video like you posted is mission critical. It's actually very encouraging to find a 'smoking gun' in the slowmo you provided. It gives you something to react to in your design aside from just adjusting things until it works right again.

This is my personal critique of what I see: Your catapult is stuck somewhere between a 'stay in the cradle' and a 'roll off the tips' design. It stays in the cradle for a while, but eventually does roll off, and the mechanism that is keeping it in the cradle for as long as it does appears to be the 'squishiness' factor of the ball.

If the boulders you are seeing on the field are consistently stiffer than what you originally tuned for, the solution is obvious - just retune your catapult to that ball. If you are seeing variation in the balls that are on the field between stiff and less stiff, then the real fix may be to adjust your cradle to either hold that ball in the cradle through the stroke - or to completely adapt 'rolling off the tips'.

Just my $0.02.

-Brando

[Kevin Ainsworth]

We did a catapult in 2014 and this year.

There are two reasons we found that the balls behave differently.
Your issue looks to be #2, but #1 might be contributing also.
Read below for more detailed information.
The new ball doesn't move as far up the cup as the old ball does.
Looks like your old ball slides up further, causing the ball to be further from the catapult fulcrum, causing more height/speed/etc. Sounds like the new balls are shooting lower, which is to be expected. They are not sliding up as far and shot from a slower/lower position on the catapult arm. We have a soft noodle behind the ball, mounted horizontally, and we do not allow the ball to move in relation to the catapult arm during the firing. This results in more consistent shot trajectory. A ball resting in the catapult where a huge chunk is missing can still cause issues, but not much you can do about that.
The higher spring back to the older balls also can add launch force and would add to the issues listed above, but to a lesser amount. The large surface area you have should reduce this effect to a minimal contribution, one benefit to the large cup method of containing the ball.

#1 The ball is "squished" when the catapult is fired and when the catapult comes to a dead stop the ball rebounds (springs) off the catapult. The different balls have different spring rates, for different reasons. Balls with large chunks missing tend to rebound back to their original shape faster since the foam can fill back up with air faster due to the large opening in the skin. Whereas the new balls expand back slower due to the single small vent in the skin requiring time for the air to reenter the ball. The amount of foam added also differs so some are denser than others. This requires you to slow the catapult arm down at the end of its travel to allow the different balls to rebound (return to their original shape) slightly before leaving the catapult. We found that allowing the surgical tubing to have minimal preload when the catapult arm was at firing position, allowed the catapult arm to slow down slightly before the limit straps stopped the catapult arm. This allowed the catapult to decelerate slightly and therefore allowed the ball to start returning to its normal shape before leaving the catapult.

Another cool catapult was 469's catapult from 2014. We were planning on using this type of device until we prototyped and found it was not necessary for the low forces required this year. 469 used a technique for reducing the stopping force distributed into the robot when the catapult reached the end of its travel. We were planning on using it to slow the initial arm speed so the ball moved a minimal amount with our long travel, tall catapult. 469 had an intermediate arm which they pulled on to load the catapult, it was tied to the main catapult arm with surgical tubing that acted as a damper. This allowed the main catapult arm to start off slower due to the rubberband effect. The intermediate arm pulled the catapult arm, and ball, via the surgical tubing. The intermediate arm also kept pulling after the catapult arm hit its stop reducing the force the catapult arm was exposed to and lessening the movement of the robot when the ball was released. Check it out, pretty cool idea. I would love to hear 469's explanation for their 2014 catapult, but that's our take on it.

#2 Depending on your "cup", "flinger", "cradle", etc. the way the ball leaves the catapult is also important. In 2014 we used manually adjustable flinger ramps that allowed us to vary the angle of release slightly to dial in our shot. The large balls would roll up the flinger fingers and the angle of release allowed us to adjust the height/release angle of the shot. That worked in a year when the balls "slide" along the flinger consistently. This year we found retaining the ball for the entire shot gave us more consistency. This years balls are tacky when new and more slippery when worn. So allowing the balls to slide or grab onto elements of the catapult can cause them to move up or down on the catapult arm (cup for you) and mess with the shot trajectory.

Hope this helps.

[JesseK]

Originally Posted by Brandon Holley

If the boulders you are seeing on the field are consistently stiffer than what you originally tuned for, the solution is obvious - just retune your catapult to that ball. If you are seeing variation in the balls that are on the field between stiff and less stiff, then the real fix may be to adjust your cradle to either hold that ball in the cradle through the stroke - or to completely adapt 'rolling off the tips'.

The boulders would start out stiff, and then they would get run over by a robot. A few boulders were completely flattened by robots that came over the CDF. In many matches there was a broad mix of new/used boulders, particularly on Saturday morning.

[wilsonmw04]

Data collection is under way...
https://goo.gl/photos/h2Hm3HTgxubcyZep7

[Clayton Summerall]

Have you tried using the exact same ball and shooting it a dozen times? We have a pnumatapult and have 0 issues with ball changes.

[Electronica1]

Originally Posted by Clayton Summerall

Have you tried using the exact same ball and shooting it a dozen times?

Ok, going to need you to explain this one a bit for me. We were relatively solid with getting the high goal from the defenses during our first two events, but then at our third started having this issue. Wouldn't it be better to make sure that we could shoot multiple balls consistently, rather than shooting the same exact ball multiple time which would not establish whether it could handle different ball quality/properties?

[Clayton Summerall]

Originally Posted by Electronica1

Ok, going to need you to explain this one a bit for me. We were relatively solid with getting the high goal from the defenses during our first two events, but then at our third started having this issue. Wouldn't it be better to make sure that we could shoot multiple balls consistently, rather than shooting the same exact ball multiple time which would not establish whether it could handle different ball quality/properties?

It's all about seeing if your mechanim has degraded or changed at all. If it can shoot the same ball to the exact same spot it shows a different issue.

[wilsonmw04]

Originally Posted by Clayton Summerall

It's all about seeing if your mechanim has degraded or changed at all. If it can shoot the same ball to the exact same spot it shows a different issue.

Let me explain the test. We have no issues shooting a single ball. Out of 20 shots, 19 are consistent. Our mechanics have not changed. What you see in our data points is the 30+ shots we took tonight under the watchful eye of a 240 FPS goPro. The yellow tape is the y-axis value of each shot. The paper records the variables of the shot. You can see that as balls wear our shooter gives them a higher trajectory. Why that is is TBA. We haven't looked at the video yet.

We did find out that we have enough adjustment on the pan to get the shot where we want it. The issue was interference with a secondary part. We are going to look at the film, do some analysis and come up with a few more tests.

We will figure this out. More to come. stay tuned.

[Clayton Summerall]

Originally Posted by wilsonmw04

Let me explain the test. We have no issues shooting a single ball. Out of 20 shots, 19 are consistent. Our mechanics have not changed. What you see in our data points is the 30+ shots we took tonight under the watchful eye of a 240 FPS goPro. The yellow tape is the y-axis value of each shot. The paper records the variables of the shot. You can see that as balls wear our shooter gives them a higher trajectory. Why that is is TBA. We haven't looked at the video yet.

We did find out that we have enough adjustment on the pan to get the shot where we want it. The issue was interference with a secondary part. We are going to look at the film, do some analysis and come up with a few more tests.

We will figure this out. More to come. stay tuned.

Do you have a mechanism on the bot that "packs" the ball?

[wilsonmw04]

Originally Posted by Clayton Summerall

Do you have a mechanism on the bot that "packs" the ball?

please explain what you mean by "pack"

[Clayton Summerall]

Originally Posted by wilsonmw04

please explain what you mean by "pack"

We have a device that compacts the ball by about 4" every shot. This seems to have been an integral part of our consistency.