pic: Gear Drive Concept

WCD-esque gear drive hybrid concept. Uses 5" Colsons and .125" thick aluminum tube.

A few comments.

First, your Colsons are on backward. The features on one side of the hub are intended to prevent the wheel and hub from rubbing against the stationary parts of the bearing, entirely contacting the inner race. Having them on the outside defeats their purpose.

Second, it seems you have the exact same gear on the output of your ball shifter as you do in the gear drive line. In that case, why not face mount the Ball Shifter to the tube and have the second stage output drive the other gear? Your current setup is just redundant.

Third, I would be very concerned with just how much you’re pocketing the top and bottom of the tube to make room for gears. Your tube’s profile gets closer and closer to resembling a pair of thin plates, which is susceptible to bending. Bending would be the absolute worst thing for a gear drive, which relies on rigidity to work. Gears in a tube are not necessarily a bad idea, but you would want a bigger tube if you were to go that route.

Fourth, have you thought about how you’ll actually get the gears in those slots? They are smaller than the diameter of the gear.

A good start, I just wanted to throw these out there to keep you working :slight_smile:

Going the all gear route you might as well just go 10 wheels if you can squeeze them in.

Use smaller wheels. He said they are 5". There are plenty of sizes smaller than that, you could use 1-7/8" if you were so inclined :stuck_out_tongue:

That Getting-the-gear-in-the-slot problem would be a killer for the design.

I like the idea. I am always a fan of differing drive ideas. (Now on to that 2-CIM-in-an-axle swerve drive)

I’d second this. Unless we have a crazy terrain for the field next year, smaller wheels are probably going to serve you better. Less reduction, smaller gears, less weight all around to spend on other mechanisms. Plusyou could probably solve both the cheese problem (the top slots) and save some money at the same time.

We found that using 4in wheels with 50t gears on both the wheels and the idlers works pretty good for spacing.

Personally, I think belt may be a more elegant solution to this. No need to pocket any tube, just make the axle go all the way through the box tubing and put a pulley on each. Team 649 does this with our 6 wheel WCD and we love it! Great start!

What’s the benefit of using gears in the drivetrain over belts or chain?

IMO, gears are heavier, less flexible, more expensive, and easier to jam than either chains or belts. Perhaps they’re harder to break (ie, to break them such that they will stop transmitting power), but I’d bet they’re a lot easier to jam by getting something sucked into them. Also, having so many gear to gear contacts, I see a lot of places to lose efficiency.

I’d venture to guess the gears take less space (horizontally) though, and honestly we used dry lube on our gear drive when we assembled it and never had to deal with it after that. Regardless, without knowing their reasoning for using this style drive it’s hard to know one way or another what an ideal configuration is for it.

Properly designed, on a system already using 8 or more wheels, gears are a small weight increase in exchange for not needing a lot of maintenance or attention like chain can. Chain or belts can be better but it’s not an absolute; for some teams / robots a gear drive is a valid solution. Don’t really know what you mean by “flexible” or why it matters here.

This was purely a mental exercise to see if a gear drive could fit in the smallest possible envelope (much like chain in tube, except different). Originally, the thought was that 4" wheels would be the way to go. Unfortunately, that left 0.150" of clearance between the lowest gear and the floor. Depending on the terrain/use case, that may work.

The bottom face of the tube is pocketed almost twice as much as the top plate in order to open up installation options for the gears. Oddly enough, there is constant contact between the hex bearings on either side of the tube owing to the 0.700" diameter boss on the vex gears which contacts the inner races of both opposing hex bearings. No idea how that would work under sudden side loading, but I’d hazard a guess that it’s better on the tubing than a completely unsupported sidewall.

How much would a riveted/brazed bellypan help those side tubes?

If the bottom is even further pocketed, you’re going to have structure problems with or without a bellypan – with dramatic pocketing, there’s no easy way to mount the bellypan to the box except at the ends short of welding it, and the bellypan itself would need to be heavily pocketed under the gears, and would provide only marginal reinforcement.

I would suggest using a larger tube, and choosing your gears such that the smaller gears are completely contained within the tube with little/no pocketing required above them.

Perhaps it may be possible to find 3x1 tubing?

That is possible … If I recall correctly, I think 610 is using something like 4x1 this year.

EDIT - This is the drive I was thinking of, not sure if they are using something similar on their competition bot.

So, some interesting things happen with 3" tube and a 50t-72t spacing.

Lots more structural webbing is available even if you open up slots large enough for installation of the 72t gears through the top. A 50t gear on a 1/8in drop center pattern has a minimum of ~0.015 clearance to the inside wall of the tube. Unfortunately, since McMaster Carr lists a wall thickness tolerance of +/- 0.019", odds are at some point it’ll interfere. I may try a 44t gear next (though this will shorten the wheelbase even further). The other option would possibly be to eliminate the drop center and add omnis to the front and back and sizing the wheels down or up to either 4" or 6".

Just raise the outer wheels 1/16th and drop the middle wheels 1/16th to make it fit. A design like that one based on 3x1 is much better.

Another iteration:

60t gears running down the whole length. A stock vexpro 2-stage 2 cim ball shifter can be face mounted to any of the hex bearings depending on space and or orientation needs. 10wd/6wd 1/8in drop center - flip the chassis upside down for 8wd 1/8in drop center. All of the gears are installed via reciprocal slots in the tube.

Stock ratios: 3.61:1 High, 8.33:1 Low. JVN Calc says that on 3.25 wheels, its a 16.6fps / 7.3fps split.

I’m curious, what size wheels are you using now, and what are the dimensions of the frame? This is starting to look more and more like our robot this year…


3.5" Colsons, though I’m also toying with 3.25 Versa DT’s.

What were your impressions on gear drive this year?

It worked great for us, our drive system was zero maintenance all season. Aside from the bit of gunk we had to clean off after our last event, the gears still looked brand new, no chipping or anything.

Our system used 50t gears on the wheels and idlers which allowed us to mount the gearbox either direct drive or to an idler without having to modify any ratios (we did end up using direct drive though). The spacing between the gears also worked nicely this way and it guaranteed virtually every part was interchangeable so it meant we didn’t need to bring as many spares to events (though it turned out we didn’t end up needing any spares at all).

We used 4in VexPro Traction wheels in a 1.5in wide configuration with traction tread. We mounted the wheels on a 32 foot arc pattern, which gave a good offset for basically zero-scrub turning, but also helped in pushing matches by keeping partial contact on the other wheels.