Building The Big Bamboo

So in a hurried attempt to get something submitted to Robot Wars, I came up with the idea for a heavyweight pyramid/wedge type robot. Having never built a robot before, I came on to this lovely website and quickly realised the world of robotics has changed since I last designed or researched how to build a robot. So now have scaled back the robot somewhat, and on advice, going to build a featherweight.

Starting with the chassis, the frame of the robot will be built from bamboo. It's a light material, cheaper than steel and around the same sort of strength, and means I can build something straight away without having to find welding equipment. The skin of the robot is still to be decided. I had originally thought 6mm steel on the skirts and other areas where the robot will likely take a lot of bashes, with the rest of the skin coming from polycarbonate sheets.

Weaponry comes from a flipper, which would have a sensor of some sort (pressure pad or optical) which would trigger a Raspberry Pi Zero to trigger the flipper at the optimal time. This should reduce the waste of precious gas!

For defence I have two ideas. The first is to make this robot a "ground effect" machine, with the robot being sucked to the ground. inside the armoured skirt would be a rubber skirt which would seal the machine to the ground. This would, hopefully, improve grip and mean I could run with no ground clearance. Along with this, in the inevitable event where someone gets underneath, I would have a gyroscope attached to the same Raspberry Pi that would monitor the X-axis of the machine. So if it was to be tipped over to the right, the Pi would instruct a movable counterweight inside the robot to move towards the part of the machine that is titled up. Hopefully I think this would make it harder to flip over or correct the issue straight away. But it's a limited weight category, and whether or not I could get sufficient weight or not to be the counter weight remains to be seen.

That's all I have so far in terms of design. I will be going out to get some bamboo canes during the weekend and start to work out what components to get. Currently I bought two 12v windscreen wiper motors that used to be in a Transit Van on the basis of motors connected to long wiper arms would be quite torquey.

I think you are falling into the classic trap of overcomplicating your first robot. My advice: Drop the armoured skirt that seals to the floor, the floor isn't smooth enough at live shows to seal onto and you'll probably find that if you do seal onto a metal floor (like the champs arena) you won't be able to move from friction. And the lowered ground clearance from it is a SERIOUS hindrance that will mostly likely cause you mobility issues.

The flipper mechanism will be fire more effective manually firing it, a computer can't really know what's actually happening in the fight and won't be able to make a better decision more often than you would manually firing it. That said pneumatics are dangerous to work with, for your first robot focus on an all electric robot that is some variant of rammer/lifter maybe an axe if you're feeling adventurous.

The counterweight is a nice idea but feathers aren't borderline lifting or flipping things, they won't notice any extra weight on the arm at all and your bot will still be sailing many feet through the air. In addition there isn't as much weight to play with as you think.

On the bamboo front I'm not really sure but I'm hesitant to say it would up for full combat. Live shows maybe though. HDPE will server you far better, it is a plastic that can be worked exactly like wood but is very tough and has flex to absorb impacts, you may find polycarbonate to be a bit brittle.

Simple, reliable, maintainable. That's the key to getting a good robot to compete with. Making sure there isn't much to break and when it does the fixes are easy and components can be removed without stripping down the entire chassis.

On the motors I can't comment on wiper motors having never used them but the cheapest drill argos sells makes an great first feather drive motor and is much easier to work with as it already comes with a gearbox.

My final important piece of advice: Don't fret too much about your first feather. You WILL make a ton of mistakes that you'll realise the moment you enter the arena with it and you'll want to immediately build version 2. Your first feather should just be something to get you in the arena and learning about fighting robots.

Good luck in your building!

Thanks Shakey. The counterweight idea came when I saw there was a 100kg limit, so I thought 10/20kg on just the internal weight would be enough to counter movement. If I scale it down, it'd be no more than 500g so it's pointless. I may however build it without the weight, just to see if the idea would work.

I'm not sure of the bamboo either. I'm basing that idea on when I used to watch Robot Wars and the robots would be built from scaffold poles. Then I remembered bamboo is used for buildings, scaffolding etc. It is strong, however it could split under impact. But I think that's a trial and error process, depending on the diameter and the thickness of the bamboo. Besides, if I use HDPE I'd have to change the name to "The Big HDPE" :P

The wiper motors I bought come with a bit of a gearbox I think - I haven't stripped them down. I'll try out the wiper motors as I've got them, but I can switch to the cheap drills if I struggle with them.

And thanks for the advice. In the space of a week I've gone from cramming everything to get an application in to RW through to reducing the size of the robot with a view to doing local events that I didn't know exist. It's a bloody lovely learning curve this is and just keeps getting more interesting.

Look up the featherweight Beauty 2. If you could build something like that out of Bamboo it would look great :L

The best way to drive a flipper, is to not actually flip. When you flip a robot you effectively loose control of it, whilst also using gas.

If you get under another robot, you gain such an advantage over it - you might aswell use that advantage to keep control over them and place them somewhere advantageous before you flip. This is why automated flippers dont really work.

The most obvious choice is to push the robots towards the arena wall to try and flip them out - but this could also work for getting them in the pit, so on.

Wiper motors have been completely superseded by drill motors; they do have plenty of torque but their RPM is far too low unless you are planning to use huge wheels.

The output shafts of wiper motors are a bit awkward for when it comes to mounting wheels too. Quite stubby, with no way to lock the wheel on. Drill motors are a bit more accommodating in this respect, and it's relatively easy to make up wheels for them.

But if you've got the wiper motors and can work with them, may as well use them. They're a bit old-school but those of us who have been doing this for a while have a tendency to go with what is easiest/most effective/what we're used to rather than going a little off track, so it's refreshing to see a different approach to things. The bamboo idea sounds particularly intriguing; not sure how well it'd hold up in the arena but would love to see how it goes :)

I must say that is the one thing I've found this week from doing research and talking to you guys: the whole scene has gone from robots rolling around with stone wheels to amazing high tech creations. 15 years ago when I was first designing and researching I found you would either use wheelchair motors or wiper motors, now it's all drill motors and brand new dedicated motors. I remember once seeing a robot (can't remember the name) and it used two electric screwdrivers as motors. I don't think it did very well, but if drills similar to it are being used then their quality must have improved as well.

I have a photo of the two motors which I'll upload tonight. They do could with some sort of "gearbox", although I'm not sure how good it is. I'm going to strip it down to find out the specs of the motor anyway so I'll find out then. But I did go for for motors from a vehicle that would have to drive long blades, so that it would be a bit more powerful.

And you make a good point about the automated flipper. It may work if, somehow, I work out a way of it knowing the dimensions of the ring so it would trigger to flip it out of the ring if it was near the edge of the ring.

And yeah, I'm just as intrigued about the bamboo idea. If it works how I think it should then it might be a good environmentally friendly way to build a chassis with. Plus it's relatively cheap. However, the real test will be when I ram the robot in to the arena wall and see if the construction snaps in half.

"It may work if, somehow, I work out a way of it knowing the dimensions of the ring so it would trigger to flip it out of the ring if it was near the edge of the ring."

Surely it's easier to just press the button yourself lol...

Put it this way, I am that type of player on FIFA that struggles to do anything more than walk with the ball and kick it. I thought it'd take a bit of pressure off my shoulders delegating the flipper to the a computer. You're right though it's probably easier to do it myself, but it would add a bit of coolness to the robot wouldn't it? :)

Put the weapon switch on a cord and have a co-pilot like the Beauty 2 & 8 team does. Have you thought about how to join the bamboo frame parts together?

*raises hand* ill be the co pilot lol, ive got a mean thumb and nowt better to do!

Quote:

---

Originally Posted by overkill

Put the weapon switch on a cord and have a co-pilot like the Beauty 2 & 8 team does. Have you thought about how to join the bamboo frame parts together?

---

Yes I have, there are a few options available. I will either go down the route of using a piece of metal that sits inside the bamboo, or a piece of metal that sits outside of the bamboo and the bamboo connects inside the metal, or have a bracket of some description. I've got some bamboo ordered which I will be picking up this week, so once I have the stuff I will start to see which is best to use.

That was pretty much what I was going to suggest. The bamboo is stronger in compression, so metal around the outside is probably going to be stronger. Google 'bamboo bike frame' for plenty of ideas.

I thought about it a lot over the weekend while doing other things, and I think the fact that if I can get a metal pipe to act as a link I can then thread a bolt through the metal then the bamboo. I don't think it'd be a good idea to glue it, as going down this path means if a length of bamboo breaks I would be able to unscrew the section and replace it fairly easily. However I've read that bamboo can split, so working out the optimal location for a bolt needs to be worked out too.

Hopefully either tonight or tomorrow I'll have the bamboo I ordered, then I can start to lay the robot out.

Wood, mother natures own carbonfiber.

If it was good enough to build the fenomenal Mosquito it should be possible to build a competitive machine with it.

There is a de Havilland museum down south and they're rebuilding a Mosquito. It's in the hanger and you can look around it while the volunteers rebuild it. It's some machine I can tell you. There's also a WW2 glider there that you can sit in, and an example of a training plane that's built from wood and canvas. You can touch the skin and it bends. When you think that the material was used on a machine that could fly hundreds of feet in the air and deliver some damage, it's just quite incredible to think the primary material was wood.

Another way to hold the bamboo together is with a bolt down the length, like this cross section:

http://www.nswrfc.org/Nick/bamboo-1.jpg

Its basically a huge barrel bolt :). As long as the bamboo has metal sleeves on the ends, it shouldn't split under tension. one problem you are going to have is that the bamboo doesn't come in fixed diameters, so finding steel tube that fits over the ends tightly is going to be difficult.

My example, the Mosquito, wasn't bolted, but glued. Wood is a very glue friendly material.

Glue would make changing parts quite difficult I think, plus if I rammed your robot would the glue on its own hold together?

Just the same for welded metal monsters, isn't it? dented and ripped pannels have to smashed back together, grinded out, replaced and rewelded.

If a Mosquito could hold together with glue @ 600kph...

Several sports cars using aluminium and carbonfiber for chassis are glued as well. Not to mention the wings of several large modern airplains. And that's with materials that are less glue friendly than wood.

Just use the right glue and work it according the specs.

I have to admit, I didn't do that yet, as I'm a metalworker by training, and a welder by experience.

Mario's favourite plane got it's strength from curved plywood rather than a strong internal frame. It wouldn't be possible to get enough strength from gluing bamboo, there just wouldn't be enough surface area in the joints. If you wanted to slice all the bamboo up and make laminated panels, things might be different :).

Learn from the past. I myself lean to the battlecruiser approach.

BIG weapon, and just enough armor.

Not the One hit wonder approach as this wannabe historical example gives. Battlecruiser HMS Hood as real life example.

And Nick, if there is a favorite plane in my heart, it's the Vulcan, closely followed by the ugly duckling the XB70 Valkiri and the awe inspiring SR71

I could just imagine me with lengths of bamboo and the thing butchers use to slice ham, when my fiancée comes in and asks what the hell I'm doing. Although I'd have thought if you were using HMS Hood as an example you may have considered the Bismark? I mean what are the chances of a squadron of Sopwith Camels diving in to the arena and shooting the robot? :P

My philosophy follows from Boxing, which I used to do. "Hit and don't be hit". Really Boxing is the same as fencing. It's not about pure power, it's about landing a strike on your opponent then getting the hell out of there so they can't hit you back. So something that's agile, hard to hit, that's what I'd like to build. The benefit of this bamboo structure is that I should be able to pop parts out and replace them relatively quickly. I don't do welding, never have. But I have drilled holes in wood before.

That said, plywood is fairly strong. I might just make a steam engined bamboo robot with a plywood skin. Proper environmentally friendly!

Quote:

---

Originally Posted by mickburkesnr

I could just imagine me with lengths of bamboo and the thing butchers use to slice ham, when my fiancée comes in and asks what the hell I'm doing.

---

She's going to say "Got wood?"

She probably wood knowing her...

Bamboo is basically a hardwood (wayhay) so you need to treat it as such. It will be hard to screw together and resist glue, but will be really good for drilling and bolting together.

Yeah that's what I thought daveimi, plus it should withstand some impact. Whether or not it'll be as strong as steel in this situation is anyones guess. But I'll find out!

The way it's meant to work is through an accelerometer, the same as what is in your iPhone to detect the position of the machine. If it's gone 180, then it'll flip after a small delay. You should be able to move it up to 90 degrees and it wouldn't trigger. So if the automatic self right isn't turned on until the machine is at a specific tilt angle.

Or you could bend some bamboo as shown below and automatically self-right with no power or coding:

http://www.nswrfc.org/Nick/P9230188.JPG

If you grow your own bamboo, it is possible to train it into shapes like this with a frame - if you are aiming for an environmentally friendly bot, its the only way to go! :)

Looks like a parachute :lol:.

How did those bars fair against opposition? Did they not fly off? And I didn't consider growing the stuff, I don't know how I'd manage to get that past the girlfriend to be honest.

The roll-over hoops are made of UHMW and have been really reliable; because they are quite springy, they also provide some shock protection when the bot lands upside down. The front one was cut in half by a vertical saw bot but that is the only real damage they have ever taken.

You would probably need a greenhouse to grow bamboo in Birmingham but as some species grows at up to a metre per day, you could have a frame is about a week.

I'd need a greenhouse with central heating to grow it in Birmingham!

I see your point though. I bought some bamboo and it was delivered last week, turns out the bamboo I bought is too thin to be used how I intended it. BUT, there are 50 canes at about 1.5/2m lengths and incredibly light. I tried to bend one and it was difficult to get any sort of bend in the cane.

So I think I'm on the right track with the frame at least, I just need to order thicker bamboo. I ordered 14mm thinking it was in reference to the thickness of the cane wall. What I'll be going for next is 25/30/40mm thickness. However the thinner bamboo would probably be ideal in a triangle formation around the base of the frame. I'll give it a try at the weekend.

You can bend regular wood after steam treatment: https://en.wikipedia.org/wiki/Steam_bending that may or may not work on bamboo. Wood can also be bent with an ammonia solution - pretty sure your GF is going to veto that one!

Quote:

---

Originally Posted by overkill

Or you could bend some bamboo as shown below and automatically self-right with no power or coding:

http://www.nswrfc.org/Nick/P9230188.JPG

If you grow your own bamboo, it is possible to train it into shapes like this with a frame - if you are aiming for an environmentally friendly bot, its the only way to go! :)

---

Only problem with that is other robots hook you don't they Nick ;)

Not just robots, but inanimate objects too. None in particular, definitely not a small tyre used to activate the pit in the RC arena. Absolutely not.

I HATE that tyre!:evil: The 2016 version of Mr Mangle has a fix for that.