Register To Comment
Page 3 of 4 FirstFirst 1234 LastLast
Results 21 to 30 of 32

Thread: Léim - Beetleweight Spring Flipper Build Diary

  1. #21
    Shooty's Avatar
    Roboteer

    Cheers for the support. I have many improvements planned for the next version of the bot: four-wheeled drive, brushless, an entirely new shape, and a far more powerful, hopefully actually viable flipper. Will update this post once I've happy with what I can show off.

  2. #22
    I hope you're going to keep at least one screw the same so you can say it's still the same robot.

    Look forward to seeing V2!

  3. #23
    Shooty's Avatar
    Roboteer

    Now that the rest of the Bugglebots specials I was in have come out, I can finally finish off this event report.

    FIGHT 3: DUNG BEETLE MELEE


    So even though I was out of the main competition, I was by no means done fighting. My next fight was the Dung Beetle Melee, where all ten competitors without a win in the main competition competed for a "best of the worst" style title. The only notable thing to point out before the fight was that as the melee was filmed early on the final day of filming, the team behind The Berg couldn't make it to the interview beforehand, so I offered to impersonate Charlie in his absence (minus the whole getting dragged out by Tim thing). I've gotten a lot of mileage out of other shirts this event!



    Some observations from this fight:


    • Just shoved some stuff around really before getting knocked upside down by The Berg, going out at 8th. Not really a lot to add here, just a sucker for consistent formatting.


    POST-FIGHT:

    The good news was that the machine could still drive, but the bad news is that the hit from The Berg actually shattered the servo arm that wound the flipper mechanism down, so what little use it had was now gone. Fun fight at least to watch, so I couldn't complain!



    So thanks to there being a Rest of the World special being hosted as well, I was given one more fight to compete in, and was drawn up against Zero Gravitas. Given how Team International had done so well up until this point, I was nervous going into the fight as now a loss would be disappointing for people other than me. Since the spring flipper mechanism was bust I used a bolt to hold the flipper blade down, and as such so long as I could drive around ZG the 500RPM wheels should allow me to push him. The rest of my strategy was covered in the pre-fight interview, so all that was left to do was to hope for a win in a pushing match.

    (Skip to 15:45)

    Some observations from this fight:


    • Zero Gravitas decided to wear some red fur on the top of its jaw for this fight, a knowing nod to fellow Irish robot Díotóir, which I thought was cute.
    • I was still having issues with my wires sometimes getting caught in the wheels, but overall I'd say we were even on control. For once the torque-ier motors actually worked to my advantage: ZG could never push me even with a clamp, but every time I got him from the sides or back I could push no problem.
    • A minute in and I get a nice push that nearly pits ZG. A while later and pit shenanigans ensue, with each looking for the right twist that would get the other in the pit. Unfortunately I am unsuccessful, and after a while later I lose drive on one side again.
    • Once I lose drive, I know I've lost my only advantage, so by backup strategy was to shuffle near the pit in the hopes that I could get a cheeky move in that would pit ZG. Fortunately I get a chance, and some iffy driving from ZG leaves them teetering on the edge.
    • Fortunately for me, all of my time driving on one wheel had gotten me pretty well practiced in driving like that, and with some skillful shuffling I was able to pit my opponent, scoring my first beetle win and a clean sweep for Team International!


    POST-FIGHT:

    So I was pretty happy with that last fight; I'd say the roughly 9 minutes of roasting Léim got throughout the competition was at least offset by some decent driving at the end there. Unfortunately the damage to the drive was a little worse than previously, as one of my motors had actually seized up, as well as one of the ESC's no longer working. It was worth it though, and the fight seems to have gone down well with people, can't complain when you end an event on your highest note!

    POST-EVENT THOUGHTS


    • Due to the damage sustained, Léim is now retired. I am considering rebuilding the bot at some time down the line (replace the motors, ESC's), and putting in a beefier servo to at least have a spring system that can at least lift. If I did that I'd even be willing to sell it outright, though I do have other priorities at the moment, so that's on the back burner.
    • I severely underestimated the energy needed to make a viable spring mechanism. The next iteration of the machine will be using a low RPM high torque motor that would put the previous servo to shame.
    • The Banebots hubs consistently failed me over and over. I know of some teams who've modified them to be more reliable, but my next design will use wheels that bypass the hubs entirely.
    • The thing I'm most proud of with Léim was how well built it was structurally. Any damage it sustained was shrugged off completely, and I reckon the machine could take hits from some of the nastiest spinners in the country over and over and keep going the full three minutes if I mounted the electronics and wheels right.
    • The success of Maximum Ogredrive has made me wary of beetle saws in the future, and the next machine I build will aim to bypass them in a creative manner.


    Overall a very fun event, and everyone there deserves praise. Special thanks to Mark for being pit buddies (congrats on winning btw!), Alex Mordue for being so friendly and donating 4 brushless ESC's to me, Jeroen and Cosmin for sharing a house during filming and the lifts towards the event, Robert for sharing some of his designs with me, all of the other roboteers, and all of the event organizers for putting on such an incredible event. If there is a second season of Bugglebots, I would be absolutely delighted to return, hopefully this time with a machine that can actually live up to the hype!

  4. #24
    Quote Originally Posted by Shooty View Post
    Not really a lot to add here, just a sucker for consistent formatting.
    Damn right!

    Great write up, and congrats on the last fight. I was properly on the edge of my seat while Léim was single-wheel-jiving over to ZG. Looking forward to seeing what you take to the next Bugglebots!

  5. #25
    Shooty's Avatar
    Roboteer

    So with Bugglebots coming back and after the disappointing performance of Léim last time round, I've began a new spring flipper build with the intention of fixing all of the mistakes the old model had. I'm calling the new build Léim Thart, and the goal of the build is to make a viable spring BW flipper, same as last time but hopefully without the same failings (pronounced "lame heart", and means "jump around" in Irish. Also now you can listen to this while reading and suffer as I did as a kid:




    An overview of the CAD. As you can see the design has gone through a complete overhaul.

    the previous design unfortunately didn't live up to my expectations, and for this machine I had two goals driving the entire project: to have a stronger, more reliable, four-wheeled drive, and to optimize the flipping mechanism to get as much energy as possible from the springs into the flipper. As such, I've forgone the old Díotóir-inspired design for a design inspired by US Beetleweight Butterfly Effect's winding mechanism, and American flippers such as Bronco or Sub-Zero, which aim to reduce the distance between the rams and the flipping blade. I've scrapped the choo-choo mechanism for a more traditional snail cam mechanism, and have swapped over to using two compression springs mounted on shock absorbers to allow them to hinge. The trick for me here was to design the flipper in a way where a) the springs were as far forward as possible to the start of the flipper for maximum energy output and b) the cam was as far forward as possible so that I could get as much useable torque out of the motor as possible after all other losses (size of the cam, increased forces by being behind the springs etc.).





    Shots of the flipper mechanism in the wound and fired position.

    The motor I'm currently using is a Servocity 34RPM Econ Gear Motor (https://www.servocity.com/34-rpm-econ-gear-motor), which has an output of 110.20 kgf/cm, though I may upgrade to the 19RPM model in the future for an absurd +180 kgf/cm. For the mechanism I'm currently using two Flexo Springs 133414 (http://www.flexosprings.com/stock-sp...springs/133414), with a rate of 10.51 N/mm, compressed down to 15mm. I've taken most of my parts for the motor mech from Brandon Zalinsky's US Beetle Butterfly Effect, with the exception of using a snail cam winding method instead of his choo-choo inspired method.

    For the drive, I was impressed by how much power WeeWoo had at Bugglebots, and after getting some brushless ESC's from Alex Mordue there, I decided to go brushless for Léim Thart. The drive will use 3D printed wheels with hydraulic hosing for rubber grip, and I will be replacing the Banebots hubs with press-fit wheels instead protected by 5mm HDPE. I'm using DYS BE1806 2300KV motors (https://www.banggood.com/DYS-BE1806-...r_warehouse=CN), combined with the Rotalin Planetary Gear Motor's gearbox (https://www.ebay.com/itm/Rotalin-Pla...1-87ca813ef861), with belt drive to drive the front wheels. As well as all of this, the new design will use interchangeable titanium flipper blades, as well as a D2 inspired hinged wedge for the rear for horizontals, though the thickness is entirely down to what I can source.

    Given how irritated I was at the performance of the bot last year, I decided to segment the build to get all of the new parts of the build working before combining. As such, I set myself the challenge to not apply for Bugglebots S2 until I had a) a driveable base for the 4WD brushless drive, and b) a working spring mech that could send the old Léim chassis airborne. The next few posts here will be my build logs for how I got both working, to be written in the coming days. Know now though that I have accomplished both, and have now submitted in my application for Bugglebots 2019. I look forward to publishing all I've done so far.


    The old Léim chassis along with my flipper and drive base, to be explained in the upcoming posts.

  6. #26
    Looking great! I like the symbolism of New Léim having to defeat Old Léim before taking its place...

    Oh and I guess thanks for the video

  7. #27
    That's a really neat design - looking forward to future posts.
    Also you need to play this song when your robot enters the arena!

  8. #28
    Shooty's Avatar
    Roboteer

    Well, this post may be a smidgen late (ahem), but I’m finally at a point where Season 2 has aired, I have enough time to do a large write-up, and most importantly, I can be arsed to do this. So before going into the event, let’s take a quick look at how Léim Thart came together.


    The Léim Thart chassis nearing completion.


    For the drive setup, I built a mock platform which was meant to replicate the drive of the final machine on a one to one scale. Getting the DYS1906 motors to work with the 22mm planetaries wasn’t easy, but a handy guide done by Cosmin Gorgovan on the Beetleweight Robot Combat page got me through, and the end result was very compact. Flashing the ESC’s proved a bit trickier, requiring an Arduino to program as well as some custom firmware to get everything working correctly, but after a bit of tweaking I had a working drive system going!


    The new wheels and brushless drive. Note the mould for cutting tires in the background.


    This project also marked the first time I could use a 3D printer (in my case a Wanhao i3 Duplicator Plus), and this would prove very useful in the creation of my new wheels. The wheels on one end press-fitted onto the gearbox shaft, and the other wheel ran on a bearing at the front of the machine. I could even use 3D printing to make a guide that allowed me to get perfect cuts out of a silicone hose I was using for tires (thanks for the inspiration RobertK!). I used rubber bands to transmit drive in the short term in place of the proper PU belts, and when run on 4S the speed and control of this bot was a substantial improvement compared to the original Léim!


    The drive platform I used for testing.




    Having good drive was just one piece of the puzzle however, as if I wanted to have a good bot, I needed a weapon that would actually function this time. This involved 3D printed shock absorbers that required multiple revisions to take the forces demanded of the springs, and though frustrating to redesign and reprint over and over again, the fifth revision would ultimately prove successful.


    The testing platform for the spring flipper.




    Some more views of the mechanism.

    Cutting plastics comprised a lot of the challenge of this mechanism, most notably the snail cam that drove the mechanism (thanks to Mark Joerger for the advice for improvements, and Shakey for the milled cam later on). After a few revisions though I finally had everything assembled and could test it on the old Léim chassis. While not a massive flip, it *technically* went airborne, and as those were the less powerful springs I had in the mechanism, I was confident I could improve it more in the future.


    Mid-way through cutting the weapon cam. This was all painstakingly done by hand, and while the finish was excellent, I outsourced the final product just to be even more safe.


    The machined finished cam next to my first attempt (I gave it two tries before outsourcing).



    So at this point I had improved both my drive setup, and my weapon, and all that would be required now would be to combine the two setups. Seems like smooth sailing doesn’t it? Well, while everything looked good from this angle, we are merely in the eye of a storm that brought up many problems that nearly sunk the whole project.
    Last edited by Shooty; 18th August 2020 at 11:26.

  9. #29
    Shooty's Avatar
    Roboteer

    The Weapon:



    The problems are two-fold. First of all, I again messed up with my understanding of springs. While I had far more power at my disposal than previously, Springs differ from the pneumatics used in similarly shaped machines in one major way: the further the travel, the less energy they have. This meant that if I didn’t have a bot touching my flipper blade, I wasn’t going to flip anything at all, and it instead became the world’s most disappointing lifter. In hindsight, I shouldn’t have used springs with such a short travel, and instead have used a flipping mechanism that brought springs from 100% tension to maybe 75% or 67% tension instead of going from 100% to 0%.

    Secondly, I needed torque, and lots of it. The Servocity gearboxes promised a lot of torque, and while they (mostly) delivered it, there was one small snag. These gearboxes were a bunch of spur gears stacked on top of each other compared to the standard planetaries, and the second the motor couldn’t compress the loads demanded of it, instead of the motor stalling, it just sheared the gears at the end of the gearbox. This lead to a *lot* of testing where I had to dismantle and rebuild my entire gearbox. Fortunately, the old drive gearboxes on the original Léim had the exact replacements I needed, but this setup was still incredibly fragile, and I’d have to be extraordinarily careful not to mess the entire mechanism up.


    Was this image taken in March, April, May, or during the event itself? Who knows, because this image is nearly identical to other images taken during this time.


    Titanium:


    Along with the digital drawing files, I also provided this guide to get the most out of the piece I had.

    I was very fortunate to be able to buy a nice piece of 2mm grade 5 titanium from Mark Leigh following the event, and the size of the piece allowed me to get multiple flipper blade configurations from the piece on top of a very nice wedge to counter any deadly horizontal spinners that could show up. After the good job done on Léim’s scoop, I contacted the same local machine shop I used last time to get the pieces cut and bent, providing them plenty of detailed drawings and with months to spare for myself. What followed was one of the worst experiences I have ever had with outsourcing some work. Some highlights include:


    • A delay of weeks from the original promised date, putting me under more pressure to get the build finished on time.
    • Massive tapers left on each part.
    • Several holes drilled in the wrong location from the drawing.
    • One piece cut several centimeters shorter than what appeared on the drawing.
    • A crack in the bent titanium wedge that was mentioned *nowhere* and that I only noticed in person seconds before paying for the work.



    My finished titanium wedge, with misplaced holes, and a rather noticeable crack on one side.


    Weight:


    Isn't it fun to go on a diet?

    While every issue up until now would be detrimental to my bot competing, one that kept creeping up concerned the weight of the final product. I had given myself some allowances for items such as bolts and nuts, but the end product was still substantially heavier than I expected. Some major work was done to lighten the machine, including hollowing out as much of each bulkhead as was possible, switching to a lighter battery, and even scrapping the notion of running the titanium wedge entirely, and even after all of this I left for the final show uncertain that the end product would fit in underneath the weight limit of 1530g.


    Taken shortly before leaving, I was hoping to lose the last 35g with the new battery I'd get at the event itself.


    While wrestling with these issues I finished up the final parts of the robot. A link holder was 3D printed which mounted to the side bulkheads, and would be re-used in both Barróg’s. The side armour was bent and fitted, which served both as armour and to give the bot a bit more character. I could also work on some full-function tests, from driving to self-righting (a little spoiler, the latter didn’t work).


    The finished link holder, which also held the battery fuse and LED power light.


    The beauty shot I took of the spring flipper the night before leaving Galway.

    It was only with a week to go that I had a realisation with the weapon. I knew that the springs were a little unreliable at best, but that being said, I did have a motor with a lot of torque sitting in the bot. I theorised that I could convert the spring flipper into a four-bar lifter if pressed for time, and so in a few hours, CADded up a design which I could cut out and print in a day as soon as possible. The concept not only worked, but this would also save a lot of weight, and allow me to use the titanium wedge when running it! I kept it a secret to all but the producers, packed up the rest, and left hoping that I could get everything competition-legal once I arrived.


    A top down view of the bot with the four-bar in place.


  10. #30
    Shooty's Avatar
    Roboteer

    Bugglebots 2019:

    After a quick stop-off in Cambridge, I arrived in Bristol a day before the competition, and got to see the new and improved pit-space for the first time. I got myself unpacked, got friendly with my pit buddies (Felix Townsend of Rev 4 in my case), and spent the next few hours frantically trying to remove weight off my bot. I had an awful lot of help from other builders, especially Felix who grinded my Titanium blades and melted together my PU belts, bought a spare battery from Rory Charlesworth, and did a ton of cutting and countersinking of both plastic and titanium to get a finished bot together. I was staying in what I’m just gonna call the ‘Drizzle’ house (Tom, Sarah, Tim, Catherine, Charlie, and Aaron), but I actually didn’t end up leaving with them that night just so I could get the bot checked and ready to go. After a final desperate session of swiss-cheesing, the robot had *finally* made it into the legal weight limit with not a gram to spare. Phew!


    Talk about cutting it close!


    We all arrived the following morning, got ourselves ready to compete, and heard all of the draws for each heat. Ours was an interesting one with a variety of weapon designs, and while I didn’t ever believe that I could win the heat, I remained hopeful that I could at least show some improvement from my last competition.

    Fight 1 VS. Unconscious 514:

    My first fight was against what may have been the most unconventional design in the competition. Unconscious 514 (Shirumi Eku) was a crusher bot coming all the way from China, two things that have never been seen in the European circuit. I was fortunate that this was my first opponent, as it allowed me to show off my new spring flipper design. I knew that Unconscious had a weapon that could go through my 2mm HDPE top armour like it was nothing, but my one advantage was that it was slow to self-right, so if I could get around him, I could have enough power to flip him over, then start to control the fight while he was upside down.


    Fight starts at 11:02


    The fight started with me getting accustomed to how Léim Thart drove in this new arena, as well as noting how Unconscious was struggling to get past the lips on the floor. I knew I couldn’t take Unconscious head-on due to him winning the ground game and the lethal potential of the weapon, so much of the fight was me just trying not to overcommit to one attack, but also to try and get behind Unconscious, which was a challenge indeed thanks to Shirumi’s excellent driving. You can actually see the flipper fire once around the 40 second mark of the fight, but the blade was just too far away.

    Things were going quite well, and I managed to control the bot better and get behind him more often as the fight went on, but unfortunately one unlucky shove left me stuck in the jaws of Unconscious, with little chance of escape. I actually tried firing the flipper here, but the gears inside the lifter gearbox had exploded like usual, and I was left with no more options. Interestingly enough, the exact same thing had happened to Unconscious, but he still managed to get the KO via pitting.

    The good:


    • The drive is fast, grippy, and controllable. Heads and shoulders an improvement over last year!
    • The flipper fired in a fight, and rewound itself without breaking.
    • My battery was barely missed, so no irreparable lipo fire.


    The bad:


    • The weapon never landed a hit, and so I never got to show off the months of hard work that went into the weapon.
    • The gearbox shredded some gears again, resulting in another disassembly and rebuild.



    The mark left in the top armour by Unconscious 514.

    Fight 2 VS. Rev 4, W I D 3 B O I:

    This match put me in the heat’s redemption melee against two people who had actually helped get Léim Thart to a finished state, and both called for very different strategies to beat. The titanium wedge to take Rev 4’s (Felix Townsend) monstrous hits was an obvious move, so I dismantled the spring setup and replaced it with the experimental four-bar lifter. WIDEBOI (Alex Shakespeare, also I’m not doing the full name again) meanwhile was an excellent controlbot, and I was uncertain I could get around him, so I needed a strategy for head-on collisions. I had brought along a longer titanium arm with a bot to act as a keepaway stick, with the intention of countering forked machines such as Thunder Child, and with a bit of measuring by eye, I positioned the bolt in such a way where I could catch WIDEBOI between its toes and lift it up, shifting all the weight to my end and giving me an edge in any resulting pushing match. Speaking of weight, even with the four-bar I was still over the weight limit, so I actually had to scrap some of my top armour to even make it into the arena!


    The re-modified robot ready to go into the arena.



    Fight starts at 27:05

    The match started cagey enough, with my priority being to keep my wedge pointed at Rev 4 for defense, though once it KO’ed itself after deflecting off the wedge I could focus on WIDEBOI head on. This was a fun matchup; the waggle-sticks on WIDEBOI could easily hook onto the top of Léim Thart, but the keepaway stick and lifter combo allowed me to push WIDEBOI around quite a bit, and when all four wheels were on the ground Léim Thart had a massive amount of speed and pushing power. After a lucky slam near the pit, I managed to go around it while WIDEBOI found itself falling in. Léim Thart had managed its goal of getting further in the competition than the previous machine, and with next to no damage I could finally take it a bit easier in preparing for the next fight!

    Fight 3 VS. The Apprentice:

    This match saw me against The Apprentice (Mark Smith), an overhead spinner that had a lot of energy and that, unfortunately for me, had a weapon that would hit over the titanium wedge at the back. I modified the keepaway stick in such a way where I hoped I could lift a still-spinning robot so that it would destabilise itself and give me a KO victory.


    Fight starts at 40:31


    Starting out the fight, I noticed that The Apprentice took quite an amount of time to get up to speed, so my main priority was to stop the weapon as often as I could and to stop it from doing any significant damage to me. The HDPE armour actually stood up very well to the blows from The Apprentice that landed, and though the keepaway stick didn’t really get a chance to work properly, I was able to get some aggressive slams into my opponent. Near the end of the fight, The Apprentice seemed to be losing either grip or drive, and I was able to get a very unlikely KO on a bot that I was quite afraid of, and somehow managed to make it all the way to the heat final!


    Fully repaired and ready to go into the heat final!

Register To Comment

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •