coolrobot

wow cool robot

@coolrobot

  • she/her

Posts about my combat robotics adventures.


Vac0om is popped up onto the house-bot
In the last few seconds, my back motor falls out!

This was my third and final fight of the day, and was the one that I was least happy with how it went. Vac0om is a (lightly modified I think?) Jolt kit, which is an undercutting horizontal spinner with a high moment-of-inertia single tooth disc. This is the second one I've fought, and they're tough and can dish out impacts, but flop around the box like a fish the whole time.

Fight Recap:

Both of us take a cautious approach to the box rush. We jockey for position a little, and I get a little unlucky that his first hit rides up the wedge in the right way to pop one of the front wheels (see design considerations below though)

He gets inverted a few times, which I think actually helps him with getting hits off and not just bouncing off my wedge. Decent inverted driving from him too, since the controls are backwards when inverted. Something to keep an eye on for me, that overcutters can get hits more directly on the weapon uprights and spinner itself.

In like, the last 30 seconds one of the rear motors starts to work loose, and I have to be a lot more cautious with my driving to keep it in the bot at all, and hope it doesn't cause electrical damage from spinning the wires. Literally after the countdown ends it finally fully falls out, but that is, as they say, "a bad look" for the judges. They didn't post the score cards as far as I know, but I assume I lost damage 4-2. I think maybe I win that fight losing one wheel on control, but not losing 3.

What went well:

I had function of all systems for most of the fight, really only significantly losing drive ability in the last few seconds. Electronics stayed fit, weapon was running, and the wedge held firm. I was also winning the weapon-on-weapon hits, popping him up without getting jolted too much myself. If the wheels stay on (big if lol) I think I win this one on control. Other than allowing the front wheels to get hit, the wedge also worked really well I think. The TPU is stiff enough that it was deflecting the horizontal hits rather than just absorbing them. Part of that is the ways it's printed, face down on the bed with 7 bottom layers and 35% infill. It took some ablative damage so I would have replaced it after that fight, but it was all within spec.

What went wrong:

Most obviously, the wheels fell off, and one of the motors came loose. The front wheels are printed glass fiber (GF) nylon on some short 3mm diameter motor shafts, and clearly they aren't cut out for hits from a big 3lb horizontal.

Some things in the weapon system also loosened up in a way that at least partially impacted function. The pulley on the weapon motor I think became loose on one of the big weapon-to-weapon hits, which I think was when it stopped spinning for a little bit. For better or worse, it's set so that spinning up tightens the pulley, but that means that big hits can loosen it. One of the three bolts holding the spinner together also stripped off, and I think another was damaged enough to slightly rub the sides of the uprights. Not disabling hits, but all stuff I would have needed to deal with if I'd had another fight that day.

Design Considerations:

Obviously the wheels are the big issue here. My goal for next year's version is to go to either fully indirect or semi-direct drive with urethane wheels, if I can get the weight together. As far as the motor falling out goes, currently they're held in place by clamping between the top and bottom plates compressing the TPU that mounts the motors, with some hot glue on the inside to help hold it, and some nylon braces to prevent spinning. For the BotKits motors this works ok (though they're more likely to spin) but the DartBox nerf motors seem to get hot enough to soften the TPU and the glue, and eventually work themselves out after the full 3 minutes (plus a little help from the enemy's spinner). For the v2.1 I'll bring to NJ, I'm going to cave and add face-mounts as well, which hopefully should be enough to keep things in place. For the front wheels I might go to the Repeat brushed drive, which has a longer shaft and a full planetary gearbox. Currently they look like the most competitive 16mm brushed option for beetles.

I really like the power-to-weight ratio of the bell style drone motors (in this case an EMax Eco II 2807 1700KV, though I like the XING 2806.5 1800KV a little better), but I clearly need to mount it a little better than just "crank the Nylock down real hard".

I might also fiddle with the wedge design a little to try to protect wheels from horizontals. The geometry is supposed to try to deflect sideways rather than up at the wheels, but that only sorta worked. Maybe some bolt heads? Not sure, I'll probably fiddle with it.

Photo credits: both from the NHRL stream link




Vac0om is popped up onto the house-bot
In the last few seconds, my back motor falls out!

This was my third and final fight of the day, and was the one that I was least happy with how it went. Vac0om is a (lightly modified I think?) Jolt kit, which is an undercutting horizontal spinner with a high moment-of-inertia single tooth disc. This is the second one I've fought, and they're tough and can dish out impacts, but flop around the box like a fish the whole time.

Fight Recap:

Both of us take a cautious approach to the box rush. We jockey for position a little, and I get a little unlucky that his first hit rides up the wedge in the right way to pop one of the front wheels (see design considerations below though)

He gets inverted a few times, which I think actually helps him with getting hits off and not just bouncing off my wedge. Decent inverted driving from him too, since the controls are backwards when inverted. Something to keep an eye on for me, that overcutters can get hits more directly on the weapon uprights and spinner itself.

In like, the last 30 seconds one of the rear motors starts to work loose, and I have to be a lot more cautious with my driving to keep it in the bot at all, and hope it doesn't cause electrical damage from spinning the wires. Literally after the countdown ends it finally fully falls out, but that is, as they say, "a bad look" for the judges. They didn't post the score cards as far as I know, but I assume I lost damage 4-2. I think maybe I win that fight losing one wheel on control, but not losing 3.

What went well:

I had function of all systems for most of the fight, really only significantly losing drive ability in the last few seconds. Electronics stayed fit, weapon was running, and the wedge held firm. I was also winning the weapon-on-weapon hits, popping him up without getting jolted too much myself. If the wheels stay on (big if lol) I think I win this one on control. Other than allowing the front wheels to get hit, the wedge also worked really well I think. The TPU is stiff enough that it was deflecting the horizontal hits rather than just absorbing them. Part of that is the ways it's printed, face down on the bed with 7 bottom layers and 35% infill. It took some ablative damage so I would have replaced it after that fight, but it was all within spec.

What went wrong:

Most obviously, the wheels fell off, and one of the motors came loose. The front wheels are printed glass fiber (GF) nylon on some short 3mm diameter motor shafts, and clearly they aren't cut out for hits from a big 3lb horizontal.

Some things in the weapon system also loosened up in a way that at least partially impacted function. The pulley on the weapon motor I think became loose on one of the big weapon-to-weapon hits, which I think was when it stopped spinning for a little bit. For better or worse, it's set so that spinning up tightens the pulley, but that means that big hits can loosen it. One of the three bolts holding the spinner together also stripped off, and I think another was damaged enough to slightly rub the sides of the uprights. Not disabling hits, but all stuff I would have needed to deal with if I'd had another fight that day.

Design Considerations:

Obviously the wheels are the big issue here. My goal for next year's version is to go to either fully indirect or semi-direct drive with urethane wheels, if I can get the weight together. As far as the motor falling out goes, currently they're held in place by clamping between the top and bottom plates compressing the TPU that mounts the motors, with some hot glue on the inside to help hold it, and some nylon braces to prevent spinning. For the BotKits motors this works ok (though they're more likely to spin) but the DartBox nerf motors seem to get hot enough to soften the TPU and the glue, and eventually work themselves out after the full 3 minutes (plus a little help from the enemy's spinner). For the v2.1 I'll bring to NJ, I'm going to cave and add face-mounts as well, which hopefully should be enough to keep things in place. For the front wheels I might go to the Repeat brushed drive, which has a longer shaft and a full planetary gearbox. Currently they look like the most competitive 16mm brushed option for beetles.

I really like the power-to-weight ratio of the bell style drone motors (in this case an EMax Eco II 2807 1700KV, though I like the XING 2806.5 1800KV a little better), but I clearly need to mount it a little better than just "crank the Nylock down real hard".

I might also fiddle with the wedge design a little to try to protect wheels from horizontals. The geometry is supposed to try to deflect sideways rather than up at the wheels, but that only sorta worked. Maybe some bolt heads? Not sure, I'll probably fiddle with it.

Photo credits: both from the NHRL stream link



Argon Lights with Impact Crater's belt stuck in it's weapon system
Jamie (me) getting ready to drive

I'm actually going to start with this fight, because I'm already at 1000+ words about fireproofing and the Mixtape fight, and I want to deconstruct the (spoilers) melted version of the robot first.

I had met Impact Crater's builder Alex at a prior event in Jersey, and he was pitted near me. Cool guy, so this was a friendly "like vs like" match of two 4wd verts.

Stream Link- https://www.youtube.com/watch?v=HXstJoqMM-c&t=11344s

Fight Recap:

We both charge at each other, do a little dance for positioning, and IC manages to get an angle on me and get under my wheels and get a good hit off. In the process though I believe he sheers his weapon pulley off, and somehow we both end up inverted. I'm able to get into inverted driving position, but he's stuck "doing the thing" since his weapon belt is thrown. I go in for hits, and am able to self right in the process, but don't do much damage. Going in for a second hit though, and I manage to suck in his weapon belt, it looks like it gets wrapped around my weapon axle.

In retrospect I should have turned my weapon throttle off right here, before the attempted Brett unstick, but I didn't realize the belt was wrapped around the weapon, at the time I was actually thinking that my blade itself was caught in the gap between his titanium and his UHMW.

Anyway, the ref does the unstick a bit more gently than Brett (thanks Nik!) and gets us apart, and it's two roughed up bots crab-walking around trying to finish the job. He's got his weapon system down and one of his front wheels is locked up (hard to tell from what, maybe a bent in wedge?), and one of my back wheels is missing (which provide something like 70% of the drive force). But we're able to dance around a bit and I get off enough of a hit to invert him again, and he taps. My geometry means that big hits are hard without some drive speed behind them (so sorry Ricky, no roofings), but this one does the job.

What went well:

Well, I won, and in general was able to win the ground game. Inverted driving and self-righting went exactly as intended. The one big hit he did get off on me was absorbed by the bottom TPU armor, and just barely nicked the CF underneath, no delams or anything.

What went wrong:

A wheel fell off (what else is new). This time I think it wouldn't have been held as strongly against me in a JD since it was caused from an attempted unstick by the house bot? BUT, it should never have happened. When I got back to the pits, that wheel was missing a screw! Not sure if it was missing when I installed it (bad!) or if it vibrated loose and I should have had loctite on it (also bad!). Anyway, glad that it didn't do me in here.

Design Thoughts:

For a 4 wheel drive robot, I'm not thrilled with the redundancy being offered when I loose my back wheels. My major "offseason" modifications are probably going to be about revamping the drive train to be either indirect or semi-direct drive with urethane wheels, but only 2 motors rather than 4. So far it's been wheels that have needed the redundancy and not motors.

Image credits:

1st image from the NHRL video linked above

2nd image by Kim Flynn (used with permission), https://www.flickr.com/photos/[email protected]/52368248293/in/album-72177720302196436/



The builds, tests, and fights I'm posting here are taken with appropriate safety precautions. If any of this inspires anyone but you don't know how to do it safely, hit me up, and I can send you resources and connect you with people who can help you do this without seriously injuring yourself.

Also when I refer to "weapons" I'm talking about the active weapon on the robot, generally a spinning steel disc that's used to impact another robot. These carry a lot of kinetic energy, but aren't anything illegal or regulated. @FBI please don't put me on any watchlists.