This small detail was a big improvement for us.
For anyone who has actually built a dead-axle drivetrain with two sprockets bolted to a wheel, and designed it with no need for chain tensioners (perfect link spacing), concentricity is a challenge. A small lack of concentricity becomes a significant issue, and it gets amplified the farther the sprocket gets from the wheel.
Screws aren't really meant to take shear force, and everything is a clearance hole, so there is some play in where the screw sits within the hole. Even if we "cranked it down" and set it up perfectly initially, we had concerns things could slide around, or even loosen/wear/slide over time. Robots smash into things, and are constantly going forwards and backwards at full power. All those forces are taken up by the clearance between the screws and holes, impacting the side of a screw that isn't ideal for transmitting shear force in the first place.
The roll pins solved all of these problems. The screws were relegated to simply holding the wheel, sprockets and spacers together - exactly what screws should be used for.
No we never had to replace a wheel, chain, sprocket, shaft-bolt, roll pin or service our drivetrain at all, other than pulling bits of carpet out of it, and inflating our tires. We played 46 official matches last year, an off-season event, and countless demos, and not so much as having to tighten a fastener.
...and that's largely thanks to little details like this that were implemented.
Yes we had spare assemblies made in case of a failure.
Yes, if needed, even I could manage to hammer a few roll pins in - probably in less time it would take for you to find a screwdriver for the crazy Canadian Robertson #10 screws we have up here.
To think that one can "learn" engineering in 6 weeks while under the strain of building a robot too is proposterous. FIRST isn't about teaching. It's about inspiring.
- Collin Fultz [more] 
18-12-2012, 02:35
Threaded Mode