Quietly anticipating videos of that arm being tested!
3 Likes
I always love to read the math and in depth engineering that goes on with your team. Moving into the back half of the season, game strategy, lessons learned from first tests and practice matches, how the drivers are practicing cycles and what drills (if any) do you use? The CG math is great, thanks for posting that in particular and everything!
8 Likes
Me too! It bears a strong resemblance to our 2015 arm, but will be much faster. The composite construction is totally new to us as an FRC team, hopefully we didn’t bite off too much.
Thanks!
Drills depend on the game, naturally, but a few types of drills get used over and over.
- Unobstructed cycles to HP station
- Unobstructed cycles to randomly distributed game pieces on the ground
- Cycles with random obstructions placed in the way (old robots, trash cans, whatever)
- Cycles against an active defender, sometimes driven by me, sometimes driven by another student, sometimes driven by our chosen chassis driver (understanding how to play defense effectively helps a driver understand how to deal with it)
- Reaction and obstacle courses of various layouts
- Boop the cone - we put a cone in the middle of the drive area and one robot needs to touch it while another robot needs to defend it, helping to practice area denial defense
We may also add a stressor to some or all of these drills. Loud music, shouting, shooting rubber bands, and the like. I aim to make training harder than the event, which seems to pay dividends on the field.
While I have posted some calcs, and Wes posted some code-based solutions, this does bear going into some more detail (and some bloody scar tissue).
Let’s talk about CG location and tipping!
This is a little painful for me, but here goes.
In 2018 we did a lot of work to make our robot effectively impossible to tip over. Acceleration limits vs arm position, long wheelbase, low CG, and extensive testing.
And yet…
Oof. The one and only tip-over we had in 2018 was during F2 at Hartford. Figures. What happened? Our bumper clipped the wall around the switch. Not really visible in the video, but I recall it vividly.
Acceleration limits are great, they can help stop tipping*, but velocity limits will be important as well when there are defenders and solid objects you may collide with.
Test thoroughly. Test hitting a stationary object with your chassis while your mechanism(s) is(are) deployed. Test different self-righting techniques that may be available to you.
What math can we do?
This post covers drive-limiting acceleration and computation.
The math is different when you get hit, or hit something stationary.
The dynamics of what happen in a robot collision are complex, so I did not try to understand them completely. What I did try to understand was how CG height matters and how kinetic energy matters.
What I took away is that:
- If your CG is inside the bumper zone it is much harder to get flipped during a collision. Maybe it makes sense to put your bumpers at max height if you can?
- The energy needed to tip up the robot is stored in its kinetic energy from driving and is proportional to v². So a small velocity limit will have a big effect on tipping likelihood.
- The X/Y position of your CG will matter, it’ll be easier to tip to the side of the robot where mass is concentrated (perhaps obviously).
- Definitely want to test this. Will test all axes with swerve drive too.
How can you determine your robot’s CG? The math is all different derivations of this:
- CAD, but requires a lot of detail in mass and components
- XY can be determined with 3-4 scales under the robot and math
- Z can be determined by laying the robot on its side on top of two scales and doing math
It can also be constructed geometrically by balancing/hanging your robot in different positions and geometrically constructing it from the angles and balance points you measure.
i.e. find two or more states that match with (C) and your CG is where the balance lines (dotted lines in the diagram) from multiple balance points cross.
There are a zillion variants of these approaches, and other approaches, so I am not going to try to cover them all. Whatever else you do: measure carefully, evaluate your tipping risks, plan mitigations, and test them. We’d all be thrilled to have zero tipped robots this year, so do your part to avoid it!
*Braking/slowing acceleration needs limiting too! Don’t just #yolo with dynamic braking on.
14 Likes
I’ve been asked on more than one occasion by rookie mentors what are some great examples of applied engineering principles they should show their students. Ever time i point to a 95 thread for the outstanding principle walk-throughs.
(Also a great way to make sure I’m remembering my principle theories from undergrad correctly 
)
6 Likes
If I’m being honest with myself a non-trivial reason I do these kind of writeups is so that I can use a notebook and FBDs again! The type of modeling and design work I do at my day job these years does not require it much.
15 Likes
So, uh, turns out you can heat-stake 3D prints with Rivnuts pretty okay…
For a magical main breaker protector.
Model here, compliments of 2220.
Some intake testing videos!
Just playing around and trying some different things here
Cone-only intake mode
Cone-only then cube-only intake modes
Success at tipping and grabbing a cone! We need to do some work to make that more consistent though.
Darn weeble cones…
We’re still digesting this testing and planning what to try next, but this ‘on a real chassis’ testing is quite encouraging.
27 Likes
Man that carpet is really showing it’s age 
10 Likes
Luckily we’re getting more on Friday!
2 Likes
For your gripper, what material are you using to make sure it actually gets picked up?
2 Likes
@StephenWitwick created a sandwich of materials for this:
- 3DP curved edge piece to avoid sharp edges
- Adhesive foam strip stuck to the 3DP part to give some extra grip compliance
- 3M grip tape (from an old KoP) on the outside of the foam for maximum traction
We are likely to try additional materials as we are able and inspired to.
5 Likes
For your front intake roller in the intake testing videos, what material are you using?
2 Likes
A drawer liner from home Depot.
I looked through the Home Depot website and found this : https://www.homedepot.com/p/Con-Tact-Grip-Black-Shelf-Drawer-Liner-04F-C6U51-06/100388906
Would you be able to comment on if this is what you all are using?
From what you have seen so far, do you think this will survive beyond initial prototyping? Or do you anticipate replacing with surgical tubing over a roller?
2 Likes
That looks remarkably similar. PNs on these kinds of products change regularly, so I can’t match it to a moral certainty.
We have used carpet grip/drawer liner materials for intakes on many competition robots, going back to at least 2002. It is a wear item, but totally workable on a competition bot. I can’t say with certainty that we won’t wind up with another material though, we are not done testing different options.
4 Likes
Minor update:
Scouting tablets are here.
Our scouts have been working from PWNAGE’s scouting app, customizing it for our usage. We have been quite happy with the QR-based communication and are using this scanner with reasonable success.
We got new (to us) carpet! Woo! Our driver is super excited about it too.
11 Likes
Today was one of those days where things just broke.
Installed a Pigeon 2.0 and started getting CAN errors at seemingly random intervals, but only when the robot was moving. After much derping around we found an uninsulated CAN lead that was occasionally shorting against a bit of metal and killing the CAN network.
We got the intake all together, even stuck a 3DP gear on the linear drive (ordering issues for the metal one… woo…) and had some okay luck testing it for a few minutes. Then a poly belt welded itself to a pulley. We swapped the poly belt out for chain only to find out that the NEO550 that had been stalled for some time by the welded poly cord was damaged.
Getting there, slowly, ironing out bugs mostly at the moment.
We also invested time in properly installing our new carpet with tape. Old robots made for great dead-weight to hold the carpet in place while we fussed around.
The scouts named all of their new tablets and are setting up software.
The tablet caddy will be named Appa, and the two QR scanners will be Momo and Hawky.
Our 2023 robot is small.
Cleaned up practice space looks pretty nice now at least!
I hope we can get some decent intake testing videos on Monday.
We are still waiting on ordering/purchasing processes from the school district to get our carbon fiber tubes for the shoulder and arm… but we are planning out and pre-working as many things as we can to make installation quick.
24 Likes
Ironic.
16 Likes
So glad you’re enjoying ScoutingPASS. Have you checked out the new features for this year? We improved our clickable images, added timers and cycle timers, and other various bug fixes and improvements.
Love the tablet and scanner names!!
5 Likes
I’ll PASS this along to our scouts, thanks!
5 Likes
Hey, scout from 95 here! ScoutingPASS has been great for us, although we modified and gutted our fork of it a lot to get it to work how we wanted on our old 32-bit iPads. Now that we’ve got nice tablets, we’re working on porting it to an Android app, which we think will work better for our purposes.
I’ve seen some of the new features, and they look neat! We might pull them into our fork once we get some time.
We really appreciate the fact that ScoutingPASS exists! Our scouting was basically nonexistent until last year, and having a pre-made app ready to use was really helpful for getting a scouting program going.
7 Likes