Team 5406 2022 Robot Reveal

Team 5406, Celt-X is proud to present our 2022 robot “Tails”.

We’ll be competing at the University of Waterloo Day 2 event, and the University of Windsor Day 2 event.

Meet Tails.pdf (2.7 MB)

Catch us this weekend at Waterloo Day 2! Feel free to ask us questions below.

26 Likes

Here is a photo of Tails

9 Likes

Beautiful machine! How exactly is the climb automated, and are there ways you’ve thought of to make it even faster?

Everything is a timed sequence of rotations and cylinder actuations right now. The original plan was to put sensors in the arms to detect when the “pizza latches” engage a bar, but that got dropped for weight, and after we realized the"sensorless" climb was already pretty close to our goal (we’re at~7 seconds from latch now)

The speed and accelerations could be bumped more, but doing so causes more swinging, with the rare result that out bumper gets “hooked” under the mid bar.

2 Likes

Do you quite literally mean timed here? As in rotate the climber arm(s) for X number of seconds and then stop. Or are you using position feedback on the arms (ie encoders or something similar)?

Very nice robot!

How has 3D printing worked out for you for the turret and shooter? Would you go this route again? What made you decide to 3d print instead of using metal? Have any parts been particularly tricky?

While we manufactured our current turret gear out of metal, we considered making a 3d printed gear last offseason but didn’t due to time and lack of experience with large prints and nylon. I hope we do experiment with it at some point because we could most likely get a higher DP for our gears and thus higher precision.

Thanks

@below, ah, oops

I’ll let the 5406 confirm, but those parts look like they were cut on a CNC from Plastic (HDPE or Delrin maybe), not 3D printed. High Density Plastic Sheet is a reasonably viable material to use in a lot of FRC applications. It has a good amount of strength as well as impact resistance and offers a decent weight advantage over Aluminum. In addition, it is a lot easier to cut on a CNC. Our FTC teams have been making drivetrain plates this way for the past several years and our FRC is just starting to think about this material instead of aluminum for certain parts on the robot.

2 Likes

Sort of a mix. The arm rotations go to set angles, but the hook releases have some timed delays for opening and closing.

2 Likes

Laser cut from Delrin. For something like a turret, it’s a lightweight solution that most importantly is fast to manufacture.

5 Likes

I’d be really curious to know your speed and power settings for the delrin if you don’t mind sharing. I’ve got a 100w CO2 laser that I started cutting parts for the team on this year. Would love to try messing around with some delrin too, considering the advantages it can have in certain applications.

1 Like

Laser cutting PETG/Delrin has been my favorite “hack” the past few seasons. For certain mechanisms (say, big, internal to the FP turret gears/sprockets or other plates requiring precision), it is just so much faster than other manufacturing options with acceptable enough (for you to determine) tolerances. We run a 150W laser and our settings for acrylic (100% power, 3% speed) work well for both Delrin/PETG, but leaves an ugly finish on PETG, so we leave PETG for prototyping. Never bothered tweaking the settings so I’m sure they can be improved.

Edit: For greater clarity on “3%” speed, we are running this laser. All materials were 0.25” thick.

4 Likes

Don’t mind sharing at all! Our CO2 laser is also about 100W. We cut 1/4"/6mm Delrin at 92% power, 400 mm/min (with air blast of course). We prototype in 6mm Baltic Birch plywood, which we can run a lot faster.

I like how the two pneumatic cylinders on the climber are plumbed together with the base end of one cylinder plumbed to the base end of the other. Saves a bit of tubing.

I was assuming that this was done such that when one cylinder is extended, the other would retract. But, from watching the reveal video, I am not actually seeing this happening. In the picture, it does not look like the two locks that are actuated by those cylinders are at opposite ends of their travel. Nor does it look like you are actually using the full travel of those cylinders.

Can you explain how that system works?

I don’t think you have it quite right. There are four cylinders (one for each hook), but only the hooks on the same “end” of the windmill are plumbed together (on either side of the robot). For instance, you might have noticed the bit of green electrical tape on one end of the windmill arms. When we climb, one solenoid releases the “green” hooks (which are connected in parallel) and another solenoid releases the “not-green” hooks. All of the plumbing runs through the center of the axle tube (and out slots in the middle of the axle), so it may look like there are more interconnections than there are.

The climber arms had to meet some tricky packaging requirements:

  • The turret should shoot in any direction, with no “dead spots” caused by the climber or superstructure
  • The climber should climb no matter what position the turret was in.
  • Our CG should be as low as possible, and the robot should fit under the Low Bar
  • The hooks had to accommodate a +/- 0.5" allowable difference in rung spacing

That meant that the turret had fit into as small a diameter envelope as possible (~20" diameter), and the arms had to be as thin as possible and be pushed right to the edge of the frame perimeter. The whole climber arm and gearbox also pops up on a neat little custom linear slide. It took a few iterations, but I’m pretty happy that we were able to come up with something that met these requirements and also performs really well.

1 Like

Here are some renderings of the robot. The full CAD and code will be released after competition season when we have a few moments to catch our breath and finalize the model (we’ll likely be making design improvements all the way up to the end anyway).

14 Likes

The cylinders currently push roller bearings, which are held by aluminum “straps”, over the back side of the hooks, which holds the hooks in place against the climbing loads. When the bearings/straps are pulled back, the hooks are free to pivot and release the rung.

We had an issue at Waterloo where the point-loads caused by the bearings was causing the back of the hooks to indent a bit, making it harder for the cylinders to pull the straps back. In fact, a hook failed to release during one of our early qual matches. We are actually looking right now at improving this by possibly removing the bearings altogether and replacing them with a linkage, similar to these toggle clamps

The robot looks great!
We plan to look into a laser cutter for future FRC construction of parts.

The only thing missing is your swerve you designed.:grin::call_me_hand:. Been working well for us this season.

3 Likes

COVID kept us locked out of our build facility for all of 2020 and most of 2021, so we didn’t find the time to try swerve in the off season. Without that experience we weren’t ready to try it this year in competition. 2023 though… :slight_smile:

4 Likes

Looking forward to it! I’d love to see you guys with a swerve drive. (I also like the wire routing to the turret)

2 Likes

Thanks! That was a pet project for one of our grade 11 students and a mentor. I agree the solution they came up with is really light and effective.

Here’s an idea that didn’t make the cut (but was fun to develop):

8 Likes