Team Darklight - Broadax (Featherweight Axe/Hammer)

[qbnut]

Long post follows!

Last year, I took my previous featherweight robot, Battleaxe (Build Thread) to FRA Featherweight champs at Insomnia 61. It did OK I guess for a 1st generation robot, but 722 eventually tore it a new one :/

The event taught me a lot, and highlighted that there were a number of design flaws that needed to iron out, and some of those couldn't really be sorted out without extensive changes. Hence I decided that rather than change Battleaxe (which will continue to run at live shows), I would rebuild entirely for a 2nd generation robot. Hence the new build, Broadax!

Conceptually, it's the same base design as Battleaxe is, in that it is an Axe/Hammer bot, and the general aesthetic remains the same, however I wanted to make several improvements to the weapon, armour, and drive that I couldn't otherwise do with the existing build.

Broadax-1.0 v72.png

October 2017

The build started properly around the time of Extreme Robots Gloucester last year. I had made the decision that this time, the robot would be fully brushless, for both drive and weapon. Parts were chosen and ordered, after a bit of research I started from the base of a drive system similar in specifications to 30haul, Charles Guans 30lb version of Overhaul. Whilst it's not exactly that now, this eventually ended up with me picking 4248 650kV motors coupled to 13:1 ratio planetary gearboxes. It's plenty torque-y enough!

We also had a cardboard mockup made for sizes to be worked out, as Broadaxs' footprint is quite a bit smaller than Battleaxes'

IMG_20171116_213839.jpg

The weapon was also designed to be much bigger, heavier and made of stronger stuff than before (8mm Hardox 500 to be exact!), so I started to design the axe arm in CAD. The axe mechanism itself has ended up as a 5060 270kV motor coupled to a 22:1 chain/sprocket 2 stage gearbox. We made a plywood model to check for positioning and sizing, tested, and it delivered quite a kick! Back at this stage we had only tested it on 4s with a fake axe head (as the real one hadn't come yet) but it will eventually run on 7s. The tests are not videoed for that

November / December 2017

During this time, we were working primarily on our Beetleweight, SCD, so didn't really touch Broadax much over this time. We were however looking into CAD milling in our workshop rather than sending out to get the parts done externally. This ended up in the top and bottom of SCD being made successfully at home using our custom-made CAD mill, with the idea to also then use it for making Broadaxs' bulkheads. This turned out not to work in the end, but we hope it might be a path for the future.

January 2018 / early February 2018

This was where I turned things on their head a little. Up until this point, I had been working with Cardboard Aided Design, but I thought it's about time I actually learned proper Computer Aided Design, as some of the parts that Broadax needed wouldn't be possible to make at home (the Hardox armour for example). So I set to making the model in Fusion 360, and learning this from the ground up. It took some time, but as I had used Sketchup for Battleaxes' early design, I had some familiarity. This ended up with everything we needed nicely available in a 3D render, to the point that I was even able to use the viewer on iPad whilst in the workshop to check dimensions for some of the supporting parts.

IMG_20180303_161354.jpg

Late February 2018 / early March 2018

The crap weather kicked in, and it meant that getting to the workshop was not feasible (it's ~30 miles away from where I live and I can only get there on weekends). So more 3D modelling was done on those days that I couldn't leave the house. At this point, I had pretty much everything modelled the way I wanted it, so I started to send off for the parts. It was always the plan to get the axe and armour waterjet cut from Hardox, but due to the issues with the workshop mill, I also got the Aluminium bulkheads ordered in. All of this was ordered from KCut who did an amazing job, not only being spot on what I needed but it all got here in 2 days!

IMG_20180315_183116.jpg IMG_20180316_172621.jpg IMG_20180307_183950.jpg

Late March 2018

This weekend and last were intended for the build to be completed, now that we have all of the parts available, but the weather has scuppered the ability to do this. So it has been mocked (taped together!) to make sure everything works for fit, and so that I can at least work on wiring and other odd jobs before we bolt the whole thing together as soon as I get access to the workshop again.

IMG_20180317_153442.jpg

The aim is to get this finished and ready for Extreme Robots Manchester coming up at the end of April 2018, but it's all very weather permitting now at this stage!

[Maxamuslead]

looks like really good evolution of the design.

Fancy uploading some gearbox and motor CADs to grabcad as it make finding accurate verions of those items easier

[qbnut]

Well the gearbox is literally three cylinders of the exact dimensions, I didn't go all out on modelling it any further. The motors on the other hand were taken off GrabCAD anyway - https://grabcad.com/library/turnigy-...2-4248-650kv-1

[qbnut]

This weekend I managed to do some more work on Broadax. The armour plate shell is now welded up and painted, the frame mostly complete and the wiring loom is ready to go in. The drive plates need attaching to the frame now, which will implement a slightly modified version of the suspension system used in Battleaxe. Once that's done, all the parts need to be brought together to finish it off. With any luck this shouldn't take much longer to finish and ready up for Extreme Robots Manchester

IMG_20180401_180251.jpg IMG_20180401_180301.jpg

[qbnut]

So late last night I completed assembly, and was able to test it very briefly this evening, such that I've got some time (not much!) to get any obvious kinks out. I won't know fully what will and won't work properly until it's gone into combat of course.

[Outlaw]

Sounds like that axe has one hell of a thwack behind it. Tried cutting up any fruit with it yet?

[Ocracoke]

Is the axe mechanism sticking in the cycle? The last swing didn't seem to be particularly free, looks and sounds pretty tasty though, well done

[Roboteernat]

Nice, i like the sound of that axe hit, bet the neighbours wondered what it was!
Yeah looked like maybe some rx loosing signal maybe causing it to suddenly stop mid swing? Maybe battery draw?

[cosmin]

Yeah, those swings don't look right to me. It reminds me of Mega Melvin's failed tests with a generic ESC which was losing sync (see the last video here: http://www.botsfc.com/home/building-a-brushless-axe).

[qbnut]

The testing was done that I could identify any last-minute kinks that are obvious from a quick run. Of course you never know what will actually break until it's gone through a round or two in the arena!

Overall, was pretty happy with the test. The drive torque caused spinning at first but I learned quickly how to get that under control, and adjusted the curves on the TX so now it's much cleaner to drive.

Post test I looked at why the axe swing was occasionally a bit sticky and it was totally a mechanical issue. The bolts that attach to the top shaft of the axe gearbox were overtightened causing the bolts that hold the axe to the top sprocket to bind against the side of the bulkhead. Loosening the bolts ever so slightly and it's now nice and free moving.

Arguably the tolerance is too tight and when it gets knocked about there's a slim but not zero chance it might come back, but I have no time to strip it down, tweak spacings and put it back together before Manchester then Burgess Hill immediately afterwards.

Will have to see how it goes though I'm pretty hopeful it won't be an issue for now.

[Maxamuslead]

The testing was done that I could identify any last-minute kinks that are obvious from a quick run. Of course you never know what will actually break until it's gone through a round or two in the arena!

Overall, was pretty happy with the test. The drive torque caused spinning at first but I learned quickly how to get that under control, and adjusted the curves on the TX so now it's much cleaner to drive.

Post test I looked at why the axe swing was occasionally a bit sticky and it was totally a mechanical issue. The bolts that attach to the top shaft of the axe gearbox were overtightened causing the bolts that hold the axe to the top sprocket to bind against the side of the bulkhead. Loosening the bolts ever so slightly and it's now nice and free moving.

Arguably the tolerance is too tight and when it gets knocked about there's a slim but not zero chance it might come back, but I have no time to strip it down, tweak spacings and put it back together before Manchester then Burgess Hill immediately afterwards.

Will have to see how it goes though I'm pretty hopeful it won't be an issue for now.

our bots will try and help with the disassembly =D, looks great.

[dotDominic]

This looks seriously awesome - I thought Battleaxe was cool, but this is a huge step up! Great work!

[qbnut]

Post report on Extreme Robots Manchester

TL;DR - Brushless axes work! But Broadax needs a lot of tweaking, and more importantly, a rebuild!

It's maiden voyage was beset by a number of issues, mostly mechanical, design and build flaws, only one electrical. Here's what happenned:

IMG_20180421_164951.jpg

Melee 1
As I loaded Broadax up into the pen, I wasn't sure how well it would do... and the conclusion was unfortunately not all that well! A number of things were immediately apparent - the ground clearance was supposed to be low, but it was too low for the difference in panel heights between the steel plates of the arena. So upon transitioning from one plate to another, it got beached frequently. It needs the suspension system that wasn't done in time for the event, as Battleaxe (now after more tweaking) doesn't suffer the same.

The spinning is because I had the channel curves on the TX set up wrong, so it span too easily. Also the transition from -100 to +100 on the stick happened too fast due to the curving which caused the ESCs to failsafe and reset all the time, thinking it was a signal fault.

The axe worked though, and with a hell of a swing! This was short-lived though, as one of the phase leads came off the motor.

Melee 2
Made some adjustments to the TX, and Team Hitec (Aegis) offered me some foam wrap for the wheels, which helped to give it extra clearance and grip. Unfortunately, the drive torque effectively tore most of it away early on. So, much more of the same drive issues, but the Axe ESC gave up this round as well.

After hours
The after hours was effectively one in, one out until the half hour session time ran out. I think most of us uused it for ad-hoc testing. After the ESC that died was replaced, the Axe worked again for a bit, before all the screws shook themselves loose. Decided that the ground clearance needed to be higher, the axe could be fixed in the morning.

Melee 3
Sunday came around, and I used the morning to strip Broadax's drive and lower the wheels to 'fix' the ground clearance in lieu of being able to actually mount the full suspension system.

IMG_20180422_112130.jpg

Team Blacksmith also noticed that when I got the drive back together, it was all back to front! I had cable tied it all down so rather than undo it all to redo it, I mirrored it on the TX (by the way, thoroughly recommend the new OrangeRx TX6i).

This time round, the drive worked correctly, the axe worked correctly, and it did the things it was meant to! For around 1:50...

Then Binky gave it a hit to the side, which tore the side welds and threw the battery out onto the arena. After that, Broadax was inactive. Drumroll came in later and made sure the panel that was hanging on no longer was. And there ended my fights for the weekend!

IMG_20180422_135638.jpg

Autopsy
The most glaring one, the armour welds snapped. It wasn't welded on both sides, only the underside, and the panels were edge-to-edge straight welds with no tabs. I think this was almost accidentally designed to be a weak spot. The axe needs to be stripped down and rebuilt, the bearings have separated from the bulkheads and it's all maligned. I watched the videos back, and noted that Binkys disc also dinged off the axe head, this is probably what broke it.

The issue of ground clearance is how Binky got a bite. As I had to raise the armour from the floor, it left a 10mm-ish gap at the side corner away from whatever was the front at the time (remembering, it's 2wd so effectively see-saws on the wheels). In a 1v1 I can theoretically put the hammer down hard which would plant the armour on the opposite side into the floor, but that's nigh on impossible in a melee with 20+ robots to keep track of.

There's a whole category of other issues, some carried over from the inherent flaws of a 'pyramid' design carried over from its predecessor Battleaxe, but mostly new as a result of some design choices and build being rushed towards the end.

But it's not all bad news! Post fight, even though the side armour is missing, the wheels are still on, the motors and gearboxes still drive, and the batteries, whist separated, took no visible damage.

[Ocracoke]

Well done for going in despite all the issues you were having. The axe certainly sounds impressive and I wouldn't want to be hit too much by it.

For your ground clearance issues, perhaps mounting the motor and wheel onto a plate with some captive nuts and then having some bolts run through them might allow for adjustable ground clearance before each fight so you can tailor the clearance as much as you need to?

[qbnut]

Before entry for the FW champs it will get redesigned quite substantially based on what I learned from Manchester and from Battleaxe upon which it was designed, as even in a 1v1 I think the current design doesn't really work that well - ground clearance is only one thing on a very long list. I'll keep the rough aesthetic as I do like it a lot but some of that nice symmetry will ultimately need to get traded for something stucturally better designed and hopefully competitive.