Time 400 Crab! Under the Sunk Cost falliSea

moop · 8 July 2024 21:05

I’ve committed myself to a Featherweight version of my antweight Time for Crab! so its time for build log.

First up, my excuse for the name: It’s (rounding up) 100x the weight of Time for Crab! and I’ve sunk cost fallacied myself into building it after picking up a couple of second hand drill motors.

The current plan is to make this as close as possible to a scaled up version of the antweight, but we’ll see if that survives through to the final robot. I probably won’t make a giant transmitter and tell everyone around me it’s us who have shrunk, but the temptation is there.

I’m in love with the idea of scaling up the 34mm wheels to featherweight sizes. These will either have a rubber overmold or if time and cost get in the way have some bike tyre screwed to them. There will be an imitation D shaft sticking out somehow.

Inside there’s the aforementioned drill motors, with the plan being to add a belt drive and some reduction so they have a chance against the giant wheels.

The claw mechanism is currently based around a big linear actuator that @McMullet had lying around. It’s rather slow (the full 20cm stroke takes about a minute, though I was giving it 18v and it wants 24), but might be okay if I mount it close to the claw pivots and use a limited amount of range.

The backup plan if the speed becomes a problem is to add a second motor that rotates the ram of the linac. Since that acts as the ‘nut’ on its leadscrew you can spin it and it screws itself along the leadscrew, normally it relies on the mounting bolt to stop it from just spinning freely. Hopefully that will give me a fast mode, but I might need some way to stop it spinning when more force is needed.

I’m still trying to calibrate myself to featherweights so right now there’s a lot of wrestling with material thicknesses and how big things should be.

The current version is scaled up 5x from the antweight, making it about 550mm wide and 500 long, with about 150mm up to 650mm gap between the claws.

Kraken4me · 8 July 2024 21:58

Nice! You don’t see many FWs on the forum, so this is a change. Also questioning why you posted this at 10 PM…

hhills · 9 July 2024 07:05

We love a good crab.

Good plan trying to save the drill drives. Honestly a lot of what kills them is the axial load too so taking that away is a solid idea. My concern is the wheels are awfully far back without much weight over them. Doesn’t strike me as handling very well at.

moop · 9 July 2024 10:15

The wheels are pretty much copied from the antweight right now. On the antweight the claws are very light so it tends to get flipped over, and I tried to make sure the wheels could be in contact all the way around.

I might need to revisit for this version, but I’m aiming for the claws to be fairly light again. I haven’t modelled it in the CAD, but there will almost certainly be some kind of casters on the claws to get it driving well, especially in the BEVs arena.

Antweight crab has some tricky curves in the claw profile so when it grabs an opponent the side nearest me gets lifted, which means the crabs wheels stay on the ground. I’m wondering if combined with the casters in the claws I can come up with a clever mechanism so this works both ways up.

Joe · 9 July 2024 11:42

As my red-featherweight-horizontal-grabber-crab is very much retired and on a wall as if it’s been hunted or something (pictured below), I very much welcome this excellent addition to the scene and I’m very excited to see this fella in action at August BEVs!

Definitely look at swapping out the linac motor for a faster/beefier brushed or brushless one down the line if you’re looking for more speed

morgan_ls · 9 July 2024 15:52

the world needs more horizontal grabbers! really loving the upscale design with the little D shaft additions will be a very fun visual

i think that claw gap should be about right too. most feathers are between 300-500mm wide. anymore than a meter and its almost too wide

on timingila i have little side forks to help pick other bots off the floor when i grab them, might be worth looking into for you bot. helps a huge amount with winning a shoving match

looking forward to a grab off with timingila!

moop · 31 July 2024 21:11

Quick update since I got the PoM video sorted today. It’s ponderous and skiddy without any tyres on the wheels, but it moves.

The main chassis is cut and assembled (wood screws - hopefully temporary) with 20mm sides and 10mm rear and base.

The drive system is based around some chonky 775 size drill motors I picked up after realising the drill motors I have were mismatched. I was considering bigger motors anyway since the wheels on this thing are huge.

I was initially planning on belt drive, but I’m currently using some HDPE gears screwed to the wheels,in a (maybe futile) effort to keep things simple.

The HDPE gears on the motors will be replaced with something more substantial soon since I don’t expect them to last in a fight.

The 34mm N20 style wheels came out nicely and I designed them so the waste material can form the bulk of the mould for the tyres. The waste from the two tyres bolts together through some alignment holes, and has some holes drilled with a ball nose cutter for tread. I need to make some spacers to position the wheel in the mould properly since the tyre is thicker than the wheel and locks into a bunch of holes drilled near the outside of the wheel.

The next steps are implementing a better bearing arrangement that doesn’t involve the chassis sides and the axle supports being captive between two big 37mm OD bearings, and that will sit alongside the motors nicely. I think with a smaller bearing on the inboard end of the axle and a thrust bearing I can combine the motor mount and shaft support while allowing the whole wheel and axle to pull out of the side of the robot by removing a shaft collar and thrust bearing, which should be much better. It also sits alonside the motor nicely letting me cut the axles to a sensible length.

After that I need to get the parts for the claws cut and draw the rest of the crab!

moop · 11 August 2024 17:35

Carcinisation continues at pace.

Using the offcuts from the wheels as moulds for the tires worked pretty well, though I switched up my machining strategy between parts so the tread on one half of the wheels is deeper than the other from different tool deflection.

Clamping the two together, backed by a lasered perspex sheet and a couple of spacers worked well. The back of the wheel was open to the air so there’s some visible surface tension artefacts, but you wouldn’t notice if you weren’t looking for it.

After a couple of machining sessions and fighting Z axis issues with the Hackspace’s CNC (I think some balls in the nut have gotten crunchy) I got the rest of the parts made.

First up, real versions of the motor and axle mounts.

These fit nicely and I could finally cut down the axles. There’s a minor hiccup in that the leftover 12mm shaft isn’t quite enough to hinge the claws, but @McMullet’'s kindly lent me some he had spare so I’m slightly less dependent on the extra I ordered arriving before next weekend.

I also got the claws mostly built yesterday, so it finally looks like a crab!

More importantly, I 3D printed some mounts for the all-important googly eyes.

And finally, I machined some aluminium drive gears this afternoon to replace the HDPE ones I used for Proof of Motion. These should work at least a bit longer.

moop · 14 August 2024 08:48

Quick update, after a frantic day of work yesterday it looks like a crab, and has a cradle, locking bar, link, safety light, and all that good stuff.

I do need to take it apart and pocket the heck out of it to make weight, and cut the various shafts in the claws to length, but otherwise I think its done.

moop · 21 August 2024 11:41

Time 400 Crab! had a great first outing. She won all her 1v1v1 and 1v1 fights, went 1 for 1 in the 2v2 fights, and seemed to be a crowd favourite. Looking at the scoreboard (not that anyone was keeping score), if the result of the final fight had been recorded Crab would have been the winner.

I did a quick test-drive in the arena before the fights started and I was pleasantly surprised by how fast and punchy the drive was, and by how well it handled. This was frankly amazing given the location of the wheels and the fact that it’s a replica of an antweight with fairly poor handling.

The only really issue was the speed of the linac, which I knew about going in. She’s an extremely firm hugger but patience is required! I worked around this by opening the claws as much as possible pre-fight to get in the ballpark for the size of the opponent and increase the chance of benefiting from any grab opportunities.

The first fight of the day was a 1v1v1 vs The Hog and Malachite. The fight started in a less than ideal way with Malachite flipping the crab and my claws catching on the arena kickboard. Fortunately I was able to get free fairly quickly and I’m very used to driving the antweight crab upside down - it flips very easily. Once free I was able to grab Malachite and spend some time pushing it around before it managed to flip me between The Hog’s chassis and its flipper for a while before Malachite flipped us again and left the crab effectively suplexing The Hog, allowing me to get right side up again. After this I was able to push The Hog towards the pit, though it ended up stuck down the side rather than pitted. The fight ended with a classic Time for Crab! situation with me grabbing Malachite and controlling forwards and backwards while they controlled the steering with the two wheels they had on the ground.

Fortunately for me the crowd liked the crab so she took her first victory!

After this fight I was a little worried the day might be over - I didn’t seem to have drive on one wheel (or any spare motors) and couldn’t drive back to the arena doors.

After getting back to the pits it became obvious that one of the left hand threaded screws locking the drive gears onto the motors had backed out, so the gear was unscrewing for a few turns before it started driving the wheel. The washers that were adapting the screws to the size of the 3/8" hole in the gear had deformed under load so I decided to replace these with a stack of washers and just tightened them after each fight.

The second fight of the day was crustacean vs crustacean against Lobster.

Unfortunately for it Lobster was having a bad day and threw its chain on the second hit.

The fight turned into a bit of a pushing match, with Lobster getting its tail stuck under the kick-plate several times, eventually into a situation where I couldn’t get it free.

Second victory for Time 400 Crab!

The third solo fight was against Need 4 Tweed which turned into another pushing match, with one bot or the other being dominant depending on how they made contact.

The fight ended with Need 4 Tweed seeming to lose drive and Time 400 Crab grappling it but failing to drag it around.

Fortunately the crowd came through for the crab again.

After this fight I was again unable to move to the arena doors for unloading, and on closer inspection it was obvious why.

The pushing match had taken its toll on the HDPE driven gear screwed to my wheels. I’m very glad I spent some time I didn’t really have a few days before the event making a spare set of these gears!

Next up was the doubles rounds, where I was paired up with @McMullet 's terrifying chainsaw bot Captain Kathryn CHAINway.

The first fight was against Malachite again, and Variable - a roughly wedge shaped lifter with long forks. I’ve learnt from experience with my crab shaped robots that they don’t do well against wedges and forks.

The fight started with CHAINway going for Malachite and me going for Variable, which it rode up before getting a bit stuck. Crab spent a lot of time up on top of of Variable getting pushed around while Malachite tried to flip it into the OOtA zone but not really succeeding. Unfortunately CHAINway got flipped at some point during this and was out of the fight leaving crab at a disadvantage tangled with Variable and getting harried by Malachite.

The fight ended with Variable driving us both into the pit. Afterwards we had to spend a fair bit of time untangling Variable’s forks from Time 400 Crab’s linac as they were very well stuck!

First loss of the day, but against strong opponents.

The final fight of the day was against Easy Over and Flipperpool.

This started off with Flipperpool flipping the crab over a few times while CHAINway went for Easy Over. I managed to grab and push around Flipperpool a bit, which was nice. Eventually Easy Over ended up on top of us and a charge from the crab got it tangled up in the OOtA zone where it was unable to get free.

Shortly after this Flipperpool’s drive seemed to fail (and it lost a wheel) and Crab was able to drag it away from the wall, and after a few attempts ram it into the pit while narrowly avoiding pitting itself.

All in all a great day out and a great result for the robot’s first outing with no real damage outside of the one mashed gear, some scratches and sticky stuff on the outside of the robot, and some slightly mangled googly eyes.

Due to logistics, Time 400 Crab! is currently living it up on holiday with @McMullet while I’m back at work.

I’ll probably make a few minor changes before the next event, but I’m really pleased with the drive system and general behaviour of the robot:

Replace the linac motor with something faster and more compact
- Current motor is really slow (and I’m running a 30v motor on 3S which doesn’t help)
- It’s also huge and I had to limit the travel of the right claw to avoid the motor getting smashed against the side of the robot in one direction or my electronics in the other
Maybe some forks (pedipalps?) between the claws
- Feels like it needs something to stop it riding up wedge shaped opponents

Finally, I’m not entirely surprised I struggled to keep this thing in the weight budget - it’s huge. Here’s Time 400 Crab! hanging out with King B for comparison:

James_Draper · 21 August 2024 14:32

The antweight concept worked well at FW, it seemed to have a lot of push!

Are you going to remake the drive gears from a new material?

moop · 21 August 2024 14:54

Yeah, I’m really pleased with its ability to shove things around.

I’m torn on whether to change up the material for these at the moment, I might see how it does in a few more events and go from there.

As a point of failure they’re kind of ideal so if these breaking saves the motors, the aluminium pinions, or the overmoulded wheels from breaking I think they’ve done their job.

I might try making some in Delrin for comparison and see if they behave better under catastrophic loads.

Kraken4me · 26 August 2024 18:25

You even made the 34mm antweight wheels bigger!