Team Astro Build Log (Cheese bots)

All things antweight

Moderators: BeligerAnt, petec, administrator

User avatar
Kyro
Posts: 412
Joined: Fri Jan 18, 2019 12:36 am

Re: Team Astro Build Log (Cheese bots)

Post by Kyro »

with nothing attached, the be 1806 is pretty quiet if its well balanced... add a bar/disk and you will get that hum as it cuts through the air
Team Rocket
Trappist 1(4wd grab 'n' lift)
Ton 618 (4wd expanding bot)
Io(4wd flipper)
JoeHLloyd
Posts: 25
Joined: Fri Jan 10, 2020 8:18 pm

Re: Team Astro Build Log (Cheese bots)

Post by JoeHLloyd »

I got a test chassis printed in PLA and boy am I glad I did - it has thrown up numerous issues. In general I think I've pushed to keep it as compact as possible and unfortunately I'm now paying the price.

https://imgur.com/gallery/ackSYQE

Basically:
  • The weapon does not appear to sit centrally. I'll bolt everything down to confirm but I think some dimension has not beem modelled correctly. I'll be getting out the verniers to confirm where I've gone wrong.
  • The drive motor holes are too tight. I can't really file them out and I can't exert much force on the motors because I'll snap the ESCs off. Mounting the motors securely whilst being space-efficient is a big problem. I think I'll just expand the holes by 0.25mm and hope it's still reasonably snug. Alternatively I could have detachable wheelguards, which would make installing the wheels/motors and filing the holes to fit a lot easier. If I was to do that I'd want them out of polycarb, and I'm not sure I actually can spare an extra 1.5mm either side because the design was bang on the cube limit.
  • I forgot to model the ESCs like a muppet, and they protrude into the cover. I'll have to raise the cover height. Not too big an issue, unless it goes above the heigh of the wheels in which case inverted manouvering will be impacted, and I'll have to get use bigger wheels. Not necessarily a bad idea as I was concerned about the amount of tyre contact the small wheels would have whilst inverted already.
  • The RX won't fit the bay. The given dimensions online were not quite accurate. This should be an easy fix - I've got 3-4mm I can increase the bay heighy by so I'll do that.
  • The weapon ESC wires/solder prevent it from fitting the bay, and may also clash with the right drive motor. I can re-solder all the big wires to angle into the ESC rather than away from it which may solve this issue, but it may make more sense to move the ESC to sit on top of the battery. There's plenty of space there. But it does mean running the motor wires through the congested wiring passage at the back.
  • The left drive motor clashes against the weapon support mount, preventing install. I can probably cut out the necessary section of this and still have it work OK.
  • There is a big clash of wiring at the back of the RX bay against the right drive motor ESC. This might be solved with a higher cover which is already necessary. If it isn't then I'm not sure what to do. Potentially swap the rx into the empty ESC bay if I move that to the battery area.
A lot of these problems would go away if I made the chassis deeper. Without wheels it's only 60mm deep. I really don't want to make it deeper because that means it will be heavier (and I don't know if I have the spare weight yet) and it will move the wheels further away from the weapon axis which will make it harder to control.
JoeHLloyd
Posts: 25
Joined: Fri Jan 10, 2020 8:18 pm

Re: Team Astro Build Log (Cheese bots)

Post by JoeHLloyd »

I'm having to make minor changes to pretty much every part of the bot. It'll be another week at least before I'm ready to print the next test chassis.

What's happening:
I'm going with 30mm wheels over the 25mm I originally intended. This unlocks quite a few benefits.

1. I get 2.5mm more space at back of the robot to expand without changing the wheelbase. This 2.5mm, couple with a curved tunnel, has added about 50% more space to the congested cable tunnel between left and right, alleviating that problem.

2. I get better tyre contact if the bot is inverted

3. I can still get tyre contact when inverted above where the drive motor ESCs protrude.

4. The weapon support will now rotate out without hitting the back of the bot, which means I can remove the slanted support at the base, which means the drive motor will install more easily.

I'm also changing the drive motor securing method. I'll use 3 sides of a rectangle to secure the motor and drop it in vertically through the open side, then cinch it down with tiny cable ties. I wanted to avoid them if possible but unfortunately they are the most space efficient solution that will allow me to install the motor without accidentally breaking it.

I also discovered the DYS 1806 model was too short by 1.5mm. I have no idea why. Be warned the rest of you.

Also, this is the last time I do a bot in sketchup. Having to manually mirror every change from one side to the other is a PITA.
JoeHLloyd
Posts: 25
Joined: Fri Jan 10, 2020 8:18 pm

Re: Team Astro Build Log (Cheese bots)

Post by JoeHLloyd »

Image

Here's the stuff you want to see so I've put it first, out of chronological order haha:

Weapon test1 (rips a chunk out of a cutting mat)
https://imgur.com/a/FS9xexU

Weapon test2
https://imgur.com/a/wX4T7iQ

Aftermath pictures and pictire of the weapon disc
https://imgur.com/a/cciXR1a


I’ve been awfully quiet on this but time for an update: the bot is nearly finished!

I’ve mostly been delayed by numerous print problems.

Got another test chassis printed in nylon. Like the toughness of nylon, hate the printed result. It’s a bit warped, and very scrappy. Heavyish too at 40g, which with all the other components only left me 17g for a weapon, which is not enough. The G5 titanium sheet I got off ebay is 4.7mm thick (a totally normal dimension) and getting a disc down to 17g just doesn’t work - too small a diameter or too fragile.

Back to refine the chassis design, this time stripping out any spare weight I could find. Reduced the number of screws by 2, decreased the size of some other areas, chopped some holes out of various places that looked like they could afford to lose the material.

A friend suggested I try TPU, as it has many of the good qualities of nylon and also prints nicely. I found the hardest TPU filament they do (85D as I recall, about the same as standard nylon) called Armadillo, got it in hot yellow, and my friend tried that. Works like a charm, really beautiful prints, zero warp. Chassis came in at 30g this time, which makes life a whole lot easier - now have 27g for a weapon and any extras.

TPU chassis
https://imgur.com/a/gTdRcR4

Now, the weapon. My good friend Richard Manning designed this evil looking disc, 23g, and after some trouble finding a waterjet company willing to take on such a small job (Jack Green engineering, Thame, recommended), I got it cut, along with the polycarbonate wedge I wanted (give the bot a good tough face that should stand up better to damage than any printed front panel/wedge would.

Flawed weapon design and other intermediate progress pics
https://imgur.com/a/G8cdf8D

Tested the weapon, and unfortunately discovered a design flaw, which really we should have both seen. The outermost point of the weapon isn’t actually the tooth, and it showed when I tested it on the old chassis in the form of poor bite. If I bolted the target down and rammed into it at full speed then I got some spectacular and nasty results, but nonetheless this was clearly going to be a problem. Shame to waste the titanium (got 4 copies), but lesson learned. When I do the plastic weapon prototyping, do check for bite. Mr Manning and I came up with a new design, which we’ve dubbed “Rogue Squadron”. There was then a very long delay whilst I waited for 1 of my friends with a printer to fix them - I wanted to print the new weapon design before committing it to metal.

Anyway, eventually it dawned on me I should just get a random company to print the weapon, which I did, in carbon-infused PLA (I think). Really nice prints, really heavy duty plastic. Might investigate it as a chassis material in future, although it feels more dense than TPU.

Tested this weapon, bite looked really good. Got it committed to metal this week, installed it yesterday. Also I wanted some "HARMESAN" stickers printed. Unfortunately the minimum order was for around 500 :lol: so if anyone wants some please ask, hahaha

A couple of test videos attached for your enjoyment, with Harmesan doing some of its tricks.

Really happy with the weapon potency - I was a little concerned it might not hit very hard because it’s not the biggest diameter weapon at 53mm, but it seems to be OK, should weigh around 23g. Also has the advantage of being 4.7mm wide over the usual 2mm wide, which means I have more than twice the contact area in each hit, and therefore more energy transfer. Ripping off a 6 inch chunk of cutting mat is insane. Heal that, self-healing mat! Am really happy that it self-rights quickly without ricocheting all over the place (so far), which you see with a lot of shallow, top-heavy verts. It also doesn’t zoom backwards across the ground (that I’ve seen) with the disc as a big wheel as I’ve seen some do on occasion. And it spins up to full speed in about 1 second. Nobody is going to be able to box rush me.

What next? Well, I need to figure out why the weapon isn’t spinning as freely as it was. I think it’s probably the weapon support that isn’t aligned correctly. Will need to spend some time fine-tuning that.

Need to put some mini-deans connectors on, fit the battery-side cover (maybe using zip ties to ensure it is secure), make some acetate blades to fit on the front for that wedging goodness, spray the disc black, reduce the low-speed sensitivity as it is still too twitchy for my liking. Should have it all finished this weekend! Hopefully there will be a competition to test it out in soon!
JoeHLloyd
Posts: 25
Joined: Fri Jan 10, 2020 8:18 pm

Re: Team Astro Build Log (Cheese bots)

Post by JoeHLloyd »

It's finished!

Key facts:
  • Yellow
  • Chassis: 3D printed TPU
  • Front armour: 1.5mm polycarbonate
  • Weapon: 23g (I think) 4.7mm grade 5 titanium 53mm diameter single tooth
  • Motor: DYS BE1806 2300kV
  • ESC: Turnigy Multistar BLheli_32 21A (safely over-specced)
  • Drive: N20 gearmotors with built-in ESC
  • RX: 4 Channel Flysky FS2A AFHDS
  • Battery: 180mAh 2S Turnigy Nanotech
95% sure it's <150g, will take it to the scales next week to be sure. I've definitely prioritised combat effectiveness over practicality. I hope I never have to change the chassis because the electronics are hard-wired in there. In future versions I think I will design it to make it possible to swap the electronics from one chassis to another without soldering. I also think in future versions I may move the disc forward a little so there's less ramp to climb, but having it where it is does mean it self-rights really easily from the back side. Depending on how destructive it is I may also design a larger disc, although I'm not sure where I'll get the weight allocation from. Maybe a gram or 2 from using N10 gearmotors? I'm not very happy with the screws holding the acetate wedge on - I've covered them with sellotape to keep the ramp smooth but I don't really like that solution. Maybe double-sided tape for the wedge would be better. Will give it some more thought.

Image
Image
Image
Image
Image
User avatar
Kyro
Posts: 412
Joined: Fri Jan 18, 2019 12:36 am

Re: Team Astro Build Log (Cheese bots)

Post by Kyro »

looks great... cant wait to see it in action...
Team Rocket
Trappist 1(4wd grab 'n' lift)
Ton 618 (4wd expanding bot)
Io(4wd flipper)
Dazmund
Posts: 40
Joined: Wed Dec 27, 2017 10:38 pm
Location: Deptford, London
Contact:

Re: Team Astro Build Log (Cheese bots)

Post by Dazmund »

Ah, the hard cheese that grates back! That vertical disc looks terrifically unpleasant. Looking forward to it take an opponent to pieces!
DAZ DUZ BOTZ

Fili-PA (wedge flipper)
Watta Doze (tank tracked pusher)
Wolfhead (front wheel drive flipper)
2-2 Liftman (fork lifter)
The Ting Goes Grrrab (grabber/control)
JoeHLloyd
Posts: 25
Joined: Fri Jan 10, 2020 8:18 pm

Re: Team Astro Build Log (Cheese bots)

Post by JoeHLloyd »

At less than 24h notice I caught wind of an Ant competition going on without me and luckily was able to fill a last minute drop out! So Harmesan had its first combat outing @ the Bristol Bot Builders Bodge Off

2 wins; 2 losses.

Same as Briestroyer actually. Weirdly similar fights, fight 1 was vs a wedge, fight 2 was vs a flipper cluster, fight 3 and 4 both against vertical spinners, EXACTLY the same as Briestroyer's fights. How strange. And just as with Brie, the 2 verts did me in.

Fight 1: Win: Lay in wait, opponent rode up the wedge into the disc, got flung in the air, drove into the pit in the confusion after.
https://imgur.com/yg7pWt0

Fight 2: Win: Lay in wait, slid under each cluster and flung them around a bit. One self-pitted and I popped the other into the pit with a half-power hit.
https://imgur.com/K53HP9q

Fight 3: Loss: Against Mr Mangles, just went straight for each other. I shouldn't have let him point me at a pit, but I was confident I would win the wedge war. I did not, and got blasted into a pit. Weird, I should have got under him.

Fight 4: Loss: Against a bot with no spikes or wedges at all, just a vertical disc on a box. Decided to go straight at it as I'd definitely win the wedge battle. Somehow didn't, and took a couple of big hits. Took bigger hits from myself ricocheting off the weapon (there's plenty of energy in it) and on the last of these sheared the weapon mount right through, landed in a compromising position, and that was game over. My opponent was also upside down and immobile after the impact, but on seeing the damage and knowing I had no way of fixing it, I forfeited and he continued. Honestly I'm a bit disappointed with this one. I have no idea why Harm was not getting under the robot better. I didn't lie in wait for this one, I was going forward quickly, so maybe the wedge flicked up on a bump during the approach? The acetate was looking a bit worse for wear, I might not have got it on the remaining piece properly. And the impacts were definitely to the wedge - I have the chunks to show for it.
https://imgur.com/leGzcC3

What went well
The acetate wedge worked brilliantly for the first 2 fights and scooped the opponents into the disc.
The disc has plenty of energy in it - you can see from fight 4 that there's enough energy to throw itself into the battlebox roof.
Soldering held together
Polycarb provided good armour

What went poorly
Made a poor decision to get head-on with another vert near a pit. Got a chunk out of the motor to show for it. Really risky move because my disc will have a shorter reach than most others, so the motor can get hit.
I designed Harmesan to out-manouver other verts and going head-on was silly
Wedge game failed me against the 2 verts, one of which had absolutely not wedging ability at all.
Critically, the weapon support sheared. Which is particularly galling because making that bulletproof was one of my top priorities. This is quite astounding actually as the weapon axle was even supported on the other side and STILL enough shear force was exerted to break the 10mm support.

What to do in future
The chassis needs a re-design before it's printed again.
  • That weapon mount must be printed with full infill and maybe some other tricks like chamfering.
  • Can it ever be strong enough 3D printed? Is TPU the best choice? If I got with a stronger structure, will I have the weight for it?
  • I must take the opportunity to re-design the wiring passages so that the electronics can be easily lifted in and out without soldering. This is going to be hard
  • Consider replacing printed weapon support with CNC perspex bolted on
  • Need to find a way to protect the weapon motor at least a little
  • Consider moving weapon forward slightly
  • Time to learn Fusion 360 properly I think
Driving choices need to be improved
  • Must remember to try to get behind other verts. That's what Harmesan should be best at
  • Need to figure out why I didn't get under that last robot
  • Do NOT get inbetween a spinner and a pit, it's too easy to end up in the pit like that
Disaster
Image
Image
User avatar
MarkR
Posts: 375
Joined: Mon Dec 18, 2017 12:46 pm
Location: Reading Hackspace
Contact:

Re: Team Astro Build Log (Cheese bots)

Post by MarkR »

Excellent job for a first outing.

I'm glad I helped- I was a last-minute drop-out :)

I see the spinner mounting piece sheared off despite being made of TPU - which is impressive. I suggest maybe either try a softer TPU, or perhaps remake that piece out of a different material or as a bolt-in piece of TPU which can be printed lying down so the layer lines give it more strength (and better for replacements)
Robots: Betsie - RaspberryPi controlled flipper bot with gyro stablisation - too clever for her own good?
Stacie - tidy flipper; 4wd driven by hair bands
JoeHLloyd
Posts: 25
Joined: Fri Jan 10, 2020 8:18 pm

Re: Team Astro Build Log (Cheese bots)

Post by JoeHLloyd »

Yeah I really like that idea of a replaceable piece. I don't really want it breaking at all though, but if it does at least it won't mean a whole new chassis. I watched a video on stressing 3d printed parts and yeah having the layers run vertically should really help resist shear forces. Another upside of the replaceable piece is that I can keep my current (inelegant) cable routing as is and bolt the motor mount over the top. Couple of 25mm M3/M4 aluminium screws should hold it in nicely, will countersink the bottom. And will include a small motor guard as part of this piece. If this keeps breaking (I will test it thoroughly before future fights) then will consider the CNC polycarb option.

Any suggestions on screws vs bolts vs nuts & bolts?

And thoughts on printing the new part in TPU or maybe something stronger (but less flexible) like carbon-PLA? I got some carbon-PLA test prints of the weapon and was really impressed by the strength and quality (3D print direct).
Post Reply