Building

The drive is made up of 4 identical assemblies like this, one for each wheel.

​

I didn’t really want to mess around trying to cast my own treads onto a custom hub is it looked difficult, and time consuming, and expensive. I decided to use off the shelf 63 shore A castor wheels from Wicke.

They are the same type used in Griffin MKI and I liked them. They were pretty light weight and grippy.

​

The downside of using off the shelf castor is that I am restricted to the sizes they supply, so I designed around that.

The pulleys and support bushing mount were watercut in aluminium blanks and finished on a borrowed lathe.

They are keyed onto the hex drive shaft.

The live shaft spins on cage bearings press fit into the plastic chassis and side armour.

​

A Igus plastic bushing sits in the drive motor mount plate to control the engagement between the two spur gears. I also hoped this would support the drive shaft if the bearing on the far end were to get smashed off.

​

The larger driven spur gear is waterjet from 6082 grade aluminium.  This is quite a soft grade, as I knew the waterjet cut quality isn’t as good as a conventional machined finish, and wanted the teeth of the larger gear to wear into the harder teeth of the small gear attached to the motor.

The aluminium teeth should work harden against the harder material of the drive gear eventually.

The 150 series motors are a snug fit through the UHMW frame. I decided not the put the rear tie plates supports on the motors, as they were already pretty well supported in the frame.

Left side of the robot assembled.

The interlocking waterjet profiles needed a bit of force to fit together on the first assembly. I designed them to have a snug fit as I didn’t want them to work loose under vibration.

The robot could support its own weight on its wheels at the point, without the side guards which is reassuring, as if they wind up being removed in a fight, I might still stay mobile.

The side armour pods are CNC milled from an 80mm thick block of UHMW plastic.

To save time and cost when machining, very deep cuts were taken resulting in this corduroy effect finish, which I really liked.

Vee belt guides were housed on the inside of the armour pods in a track to protect them.

The side pods are supported by the centre frame using two arms, and retained by bolting through a steel plate.

I think this is my favourite part of the whole machine.

I made use of the 5 axis waterjet at CWC and added chamfers to the UHMW frame.

​

At the back this allowed me to keep the rear wheels in contact with the floor if the front end is lifted of the ground.

It also breaks up the boxy shape of stacks of 2D profile cuts.

The hydraulic power pack is a pretty much standard off the shelf unit from Related Fluid power.

​

I fit my own motor onto it, so had to machine my own coupler to the pump, and design a way to mount this.

I also sealed the breather cap for the reservoir with epoxy, as otherwise it would leak oil when inverted.

I couldn’t afford to remove the material all the way through for such a shallow clash, so I planned to use a router to carve out voids for features of the hydraulic power pack so it would sit flush in the frame.

The power pack lives in this cut out in the centre of the frame

The power pack is held in place by a recessed plate on the underside, that was also counter-bored with the plunge router.

Another waterjet sheet bolts onto this to protect the tank, and to keep all the electronics from falling out all my weight saving holes.

The hydraulic actuator is held between these two machined blocks, which are clamped between two thick waterjet plates that mount back down to the frame on a bar of titanium.

Here is the assembled actuator and pivot mount shaft

The tie plate runs through the plastic front plate and resolves the actuator force so that it doesn’t pull the plastic frame apart.

The holes on all the shafts for the weapon are oversized on the steel parts, and tight on the plastic.

This way when a force is applied the parts all move around until they hit the steel, which acts as a limit stop.

​

My thinking is that when impacted, the more parts that are able to move the less concentrated the force is in a single place.

​

The scoop was lasercut with interlocking locations tabs, for both structural strength and to make welding fast and simple for me.

I designed a 3d printed block to house my removable links and seated them in the frame.

​

A thumb screw retains the link so that it cannot fall out.

I’m not going out like that!

I designed these terminal blocks for multiple ring crimp terminals.  

I like to zip tie all the wires together to reduce the chance that a single wire will get snagged and pull out of the crimp

The speed controllers were mounted at the side of the hydraulic unit. The batteries lived below the speed controllers between the drive motor rear caps.

These were covered by the self-righting arms

And it was time to crate up and ship out.

​

The night before I had disassembled for a quick paint so I didn’t ever get a chance to fully test or assemble!

This is the weapon system after a quick test on the bench.

The difference in size between the actuators might not look like much, but the weight saving was considerable as the jaw strength could be reduced because of the drop off in forces being applied.

Here some of the components for the drive assembly are able to be seen a bit better.

The drive shaft was turned from stock Hex bar over at Wycombe Engineering, and case hardened so that it could run directly onto the needle roller bearings seated in the frame.
the wheel is retained by circlips cut straight into the hex stock.