Bone Dead Robotics

And now for a little trip! About two weeks after a run at NHRL it was time to turn around and take the 3lb robots south to Florida. Normally I have a rhythm of having bots done early and tested beforehand so the expectation is that when I arrive I have confidence in the machine to get through safety and be ready for its first fight with the only exception being configuration changes. In short, past me works so future me has no concerns. This time, there wasn’t time for this event prep due to the previous event’s prep. All of the difficulties to get Mind Flayer ready (detailed here) meant that Phenomena had the least amount of work going into it of bots in the past two years. Add to that Cattywampus’ damage from the previous event which had to be fixed and you end up with a lot of tasks in a short amount of time. It was going to be tight, but every now and then you have to push your limits to keep your creativity active.

Production began on Phenomena in earnest roughly a day after October NHRL ended; at the time of writing this post Mind Flayer still has not been torn down for evaluation but simply stored on the shelf. Instead, the design was given a quick once-over then was sent to the 3D printers. The beauty of 3lbers over 30lbers is how quickly modern printers can make the parts and those parts can be immediately useful. After an overnight TPU print, the billet chassis of the new Phenomena was printed and after a very short run on the router the last metal parts were made. Fortunately all of the guts from the machine that went to SWORD back in August were unaffected (including the hubmotor) so it was an easy transplant. Unfortunately, my assumptions about the robots proportions were off compared to what I was hoping for; the TPU was too floppy, and the robot far heavier than expected. While this machine was designed to use a servo to mirror its progenitor it was in fact way overweight even without it. Perhaps the worst realization was the center of mass was too far forward so the robot drove very poorly compared to Cattywampus who had the same drive system. It was now just over a week to competition so the evenings quickly became test sessions and modeling a new machine.

One of the obvious limiting factors of the weight and balance of the machine is the steel forks. Once again, for the sake of modeling the larger one, I made steel forks to see how well that design scales between classes. The full details will be further elaborated on in the end of year update for each machine but in general I knew they would be a really reliable feature to take a beating in 3lbs if they work and they needed real testing to know for sure. Unfortunately, having the heaviest components far away from the drive axis meant that the wheels weren’t receiving nearly enough downforce for good drive. This simply would not be acceptable. I did a bit more testing to see what would be required to get the current chassis to work but the results weren’t promising. I debated re-machining the inner rails to create better support over the front, but I still didn’t like how narrow the bot was versus how long the forks are. While I wouldn’t throw away this setup just in case a better idea came together in time, I decided it was time to refocus and design a new shape to make certain the bot had a fighting chance.

The new design is a more extreme application of my wide, shallow robot shapes. The design ethos is best explained in Vorion’s and Phenomenon’s pages but for this machine there would be a few more nuances. This robot is designed with only a few stiff elements (the back rail and central frame rails machined from UHMW) with the steel providing the stiffness for the otherwise very floppy TPU. The steel does not connect to the UHMW so the machine can still flex when struck with only the shoulder bolt through the weapon and ~ 1/4″ UHMW backplate as the only linking factors. There is no joining baseplate so in theory someone could cut through the back rail, something many dangerous 3lb spinners are able to do these days, and the machine would be almost bisected. It’s a very all-or-nothing failure mode that I like to resist for redundancy but desperate times, desperate measures.

Now that the final shape was sorted, it was time to solidify it in the final materials. This machine is a further experiment as only the center rails and back rails are machined from 3/8″ UHMW and serve as the stiffness elements with the rest of the chassis being printed. The main bulkheads are TPU where the steel fork mounts lay on top of it. Though the steel is very strong, its sitting on a rubbery base so it can still float and absorb energy while resisting bending; it should be a fun material properties experiment! The top and bottom plates were originally designed to be 1.5mm G10 Garolite but in an effort to minimize time and avoid having to machine composite materials (their abrasive nature wears down endmills and creates dust that is harmful to your lungs) they were instead printed in Carbon Fiber Nylon. These proved to be very fast to make and later a good balance of stiffness and ductility so I may try this idea again (probably in events without flame bots). After that, both bots were packed for their trip to Florida!

I haven’t been to Robot Ruckus since 2019 when the large heavyweight arena was there but after arriving on Friday night for safety I was impressed. In this newer configuration, there are 3x insectweight arenas for running the 1lb Gladiator (beginners), 1lb Plastic Ants, 1lb Antweights, and 3lb Beetleweights. With the extra space freed up there was ample room room builders, crowds to gather around the arena, and a whole announcer crew for each one. During the event they had pit runners as well who used a competitor match to let people know of the next 2-5 fights. Its a similar idea used at other events like NHRL but I think the volunteer team at Ruckus did an exceptional job juggling all of these things so big kudos to them!

The downside of first arrival was with the forks on Cattywampus. I’m not sure if it was accumulated damage from NHRL that I didn’t catch in my post-event teardown or the TSA got very aggressive with it, but the right fork was torn out of its socket. While the steel body was good, the TPU was torn out of socket and, due to the tight fitting with the Titanium and the robot throwing itself around, would easily get stuck bent down during testing inhibiting drive and control. Since there wasn’t a spare chassis prepared and limited time until fights the next day, the gap was filled with super glue, clamped together, and left overnight.

Completely unrelated to my own machine, I also find this little guy whose name I had forgotten but was an incredibly cool self-body flipper. By overdriving the rear wheels at will, it allowed the entire robot to scissor, making it an entirely drive-related flipping mechanism. I’m not sure if it fought much but I appreciate the innovation.

Round 1

As the competition began, both bots lined up for their fights. Phenomena drew Boop the Snoot by Andrew Smith, a builder I’ve known for years whose family made the very popular Weta and Saifu kits. I know this machine was well-made with a very capable driver so it would be the perfect test for Phenomena’s durability. I moved the right fork inward to guarantee that the forks would touch BtS’s chassis before getting fed up to the blade while keeping the left fork out wide ensuring I could still coral the chassis it case it tried to get to the side. This fight also revealed a design weakness of Phenomena early on. Despite a hubmotor based around a Propdrive 3530 (which produces tons of power for beetleweights) it did not have enough torque to self-right the machine. In addition, the safety stickouts were a little longer than expected while the wheels, worn over the course of a few competitions, were no longer tall enough to reach the back leading to the robot easily ending up on its back.

The fight itself was very dynamic with Phenomena and BtS exchanging rood shots back and forth but Phenomena used its forks to get better position and rip off the front wedge of BtS. Ultimately BtS got stuck in the arena pit but the same shot inverted Phenomena and led to a double KO. Phenomena won the judges decision due to its damage and dominance earlier on. One match in a winners pog! Phenomena 1-0.

Cattywampus was not so lucky as it fought a very tall vertical spinner called Reign from Purple Fire Robotics of Florida Poly University. Originally I planned to fight with the head on but due to its drive being clunky (perhaps more funny business from its plane ride) I ended up postponing and using the opportunity to both fix drive and remove the head. This removed any obvious catch points and save the only working servo I had for Cattywampus.

During the fight Cattywampus was able to survive the big shots from Reign and tie with its forks but in a bot that demands getting under the opponent this still wasn’t good for Catty. While Reign did lose its belt in a shot, Catty was unable to effectively capitalize and the self-wronging tail throws looked to the judges more as a lack of control than trying to aggress due to the time I had to take to reinvert the machine and reset the tail after missing a launch. While Cattywampus definitely demands a lot of interpretation from judges, this is something that has to be minimized for the machine to be more competitive so I began tinkering with ways to make Catty’s chaos more orderly (as oxymoronic as that sounds). Ultimately Cattywampus lost the fight by judges decision. Cattywampus 0-1.

Round 2

Sadly the photos for the competitor rounds get a little sparse as the robots didn’t garner too much damage until the very end. Phenomena, by the random chance of the bracket, actually drew Reign which previously fought Cattywampus. Due to Reign’s enormous reach and size it was a tall order for Phenomena and that intuition was correct. After what felt like only a few seconds, Reign was able to directly impact the hubmotor body on the weapon and invert Phenomena which, due to the same design flaw shown in round 1, was unable to self-right. A quick loss but nothing too surprising, Fortunately the hubmotor still worked perfectly fine so nothing left to do but try to brainstorm solutions. Phenomena 1-1.

Cattywampus’ second fight was against an SSP by the name of Strange. Lillian, driver of Strange, and her father were excellent sports and really enjoyed driving against Cattywampus. It was nice to have a fight where your opponent can’t destroy you and Cattywampus’ movements always seem to spark some joy from those who see it. Strange did a good job controlling against Cattywampus as it could self-right much faster and Catty’s fork being bent down again led to the robot having a harder time getting back into control mode. The machine was certainly fun to watch but ultimately lost, propelling Catty down to a rough position for making it to the second day of competition. Cattywampus 0-2.

Round 3

Phenomena’s third match was against a narrow drisk robot called Dreign, a teammate to Reign. Once again his weapon outreached my own so the odds of the same thing from the previous match happening were high. I attempted to ground down the safety hoops and jankily drill a hole in the frame rail to use an allen key as a lock but unfortunately the wheels on the back were just worn down enough that they couldn’t get consistent traction on the back and couldn’t move well. The fight itself lasted all of five seconds as Dreign ensured a nose-to-nose shot that struck the hubmotor and inverted the machine again. While the hubmotor was still fine and the bot functional it did not Phenomena out of contention to keep moving forward in the tournament. Phenomena 1-2.

Cattywampus’ last fight was against a very long and wide vertical spinner from Purple Fire Robotics called Systematic Chaos. The builder of the bot was busy with his other machine so he already warned that it may have issues during our fight. During the battle, Cattywampus was able to use its longer reach from its forks and durability to outlast Systematic Chaos. It wasn’t a very flashy battle, but it Cattywomped anyway. Cattywampus 1-2.

Rumbles & Grudges

Cattywampus was able to fight Double Destruction, a robot with a Fingertech beater bar and horizontal spinner kit. Once again, the grabber head from the top of the machine was removed to avoid horizontal shots. The fight was short, but Catty managed to to Cattywomp anyway.

After Phenomena and Cattywampus’s fights ended there was no longer a time crunch, just time to think and plan. I had already begun CAD’ing a new Cattywampus on the computer so as an offset to that I figured I would experiment with Phenomena to see how close I can get it to functioning correctly. I switched in a Sequre 70A with AM32 in place of the iFlight 80A and started using 1in long 6-19 plastites to simulate new supports. Using a common trick by other 3lbers where the blade is allowed to stick out a tiny amount beyond the support. That tiny difference gives the blade enough leverage to get going but the screw supports the robot enough so the blade doesn’t get too much leverage against it before the tooth comes around again. Between these two things I was able to get the robot self-righting reasonably well and prepped for the 3lb rumble.

The rumble was a very good testing for Phenomena. In a rumble of ~10 machines including heavy hitters like Weta it did a phenomenal job taking the hits. The carbon fiber Nylon plates ablated some but never fully shattered even with big scary spinners wearing on it. The TPU front wedge (a last minute addition due to having very little weight remaining) was destroyed by a horizontal early in the rumble, but I’m pleasantly surprised with how well the AR500 mounts and Titanium forks held up. I can see some warping in the steel due to the shots, but they were still very functional at the end. All of the core functions like drive and weapon were completely functional with all other key structures together too. While the geometry needed some work, its very clear the heart of the machine was still more than ready to keep going. This machine will definitely be returning in 2025, perhaps at an NHRL alongside Vorion.

Conclusion

Though neither machine won a podium, Cattywampus did win Most Entertaining Beetleweight (perhaps the most appropriate award it could have won) alongside a $75 CAD giftcard to Fingertech! Both machines also leave with a winners pog and only some damage to worry about. All of the core components are still in fine working order and it gives new ideas for the future!