Shafts in Drivetrains

[AndreaV]

Cantilevered shafts!
You should do a search for it, I am sure many people go over it in great detail.
Last year we had .5" hardened steel cantilevered shafts for our drivetrain,they stuck out just under 3". After 2 regional, champs, and IRI the shafts were exactly the same as they were during assembly. Many teams use aluminum with no issue, it was just our first year trying it so we played it safe. We used bolts and circlips to hold the extremly sticky colsons on, no problems. the weight isn't too much less, aluminum shafts would have saved some weight. Yhe biggest advantage is space, the ease of wheel/drive train access.

Cantilevered wheels are a must have for West Coast Drive, I suggest you look into that too.

As long as the shaft is sufficiently supported, there's absolutely no problem with cantilevering the shafts. It's commonly found on a West Coast Drive, where the axles are live and well supported by a bearing block.

[Akash Rastogi]

Originally Posted by Andrew Lawrence

As long as the shaft is sufficiently supported, there's absolutely no problem with cantilevering the shafts. It's commonly found on a West Coast Drive, where the axles are live and well supported by a bearing block.

This is an overly simplified statement. Quite a few factors go into shafts for a WCD, including material used (I don't use 6061 to cantilever since friends reported bending of shafts on their bots), stress risers along the shaft, length of the shaft, load at the end of the shaft, etc... These factors are most important for the output shaft of the gearbox.

I haven't seen teams have wheels "fall off," but I have heard of the occasional shaft snapping near a retaining ring groove, and causing issues. You can retain wheels using snap rings, I prefer to tap the end of the shaft in order to drop in a large washer and bolt to hold the wheel on. The retaining rings at the ends of a shaft don't/shouldn't cause problems though.

I only have anecdotal evidence from other teams and from 3929's 2013 drive, but as far as I can tell, cantilevered shafts are A-okay with the proper setup. Check out any 254/968/973/1538/1323/987/60/696 bot as examples.

There's a good amount of discussion on this photo as well - http://www.chiefdelphi.com/media/photos/35912 but deals with a shaft used in a shifting PTO gearbox. I don't think single speed gearboxes have nearly as many issues to worry about.

Originally Posted by Akash Rastogi

This is an overly simplified statement. Quite a few factors go into shafts for a WCD, including material used (I don't use 6061 to cantilever since friends reported bending of shafts on their bots), stress risers along the shaft, length of the shaft, load at the end of the shaft, etc...

I should have gone into more detail - my apologies - Yes, as Akash said, you shouldn't have a problem with cantilevered shafts if you use the right grade material in the correct manner. For a live axle setup, 7075 aluminum in 1/2" hex should work perfectly. I've yet to have problems with it. If you're doing anything smaller or dead axle, I'd go with a steel alloy (I hear 4140 is good), though I find it better to stick with live axle and use 7075 al.

Also, as Akash mentioned, I also prefer to retain parts on the shaft by drilling and tapping the two ends of the shaft 10-32 by about 3/4" and putting a 5/8" 10-32 bolt with some washers on for size to keep everything on the shaft.

[Tungrus]

In our experience it's is the robot weight supported by the shaft that has more impact. Also the distance between the wheel and the bear shaft is supported on. Our team has successfully bent 1/2" keyed Aluminum shaft. For cantilever shaft steel is the probably the best choice.

[sanddrag]

We've bent 6061 and 2024 1/2" round shafts. Never a problem with 7075 1/2" round for wheel axles. Just don't cantilever it out too far. In 2012, we had 6" pneumatic tires and 7075 axles, and we jumped the robot over the center field bump (and speed bumps at school) to where the whole thing was a clear 8" in the air, and it was never a problem. Also, I would sometimes ride the robot, and I weigh twice as much as it does.

[Tottanka]

im planning on a 3/8'' dead shaft, for a 6x6 or 8x8 DT.
is that a problem in your opinion?

[Qbot2640]

Originally Posted by Tottanka

im planning on a 3/8'' dead shaft, for a 6x6 or 8x8 DT.
is that a problem in your opinion?

There are many variables to consider: total load, length of shaft from support to load, Modulus of Elasticity of the chosen material, etc. But the comparison of section modulus in 3/8" round to 1/2" round (and to 1/2" hex) is rather surprising.

3/8" Round has a section modulus of 85mm^3
1/2" Round has a section modulus of 202mm^3
and 1/2" Hex has a section modulus of 213mm^3

We have bent 1/2" steel round in a non-drivetrain robotics application, and we have bent 1/2" bolts used as dead axles supported on both sides of the wheel (estimate 5" spacing)...this comparison would make me very hesitant to use 3/8"

Sources:

http://67.237.145.72/PDFs/SECTIONPROPERTIES.pdf
http://www.engineersedge.com/calcula...re_case_15.htm

[Chris is me]

Originally Posted by Andrew Lawrence

I should have gone into more detail - my apologies - Yes, as Akash said, you shouldn't have a problem with cantilevered shafts if you use the right grade material in the correct manner. For a live axle setup, 7075 aluminum in 1/2" hex should work perfectly. I've yet to have problems with it.

You're still missing way too much detail to speak in general here. The point of interest is more than just the shaft sizing; how the load is applied and supported also matters. Essentially you want the cantilevered load as close to the support as possible and each of the support bearings a reasonable distance apart. In a drivetrain it's fairly easy to get those supports far enough apart, so unless you're doing something really strange like half-inch tubing, the main area of attention should be getting the load as close to the support as possible. Don't space out wheels or use excessive spacing between sprockets.

If you're doing anything smaller or dead axle, I'd go with a steel alloy (I hear 4140 is good), though I find it better to stick with live axle and use 7075 al.

You're saying that for an equal size shaft and load, you need steel for a dead axle but can use aluminum for a live axle? I don't follow; the bending loads on the shaft should be identical. If anything the lack of torsional loads on the axle would slightly reduce the stresses involved on a dead axle.

OP: If you're deviating from the tried and true 1/2" shafting west coast type setup, you should probably do some math to be sure. Very simple statics, just draw a free body diagram and look at the forces and moments at play. If you need help with this I can walk you through the math; I'm a bit busy at the moment though.