Possible to extend output shaft on WCP Gearbox?

Hello,

I am designing a West Coast Drive in which of course the wheels, but also all pulleys are mounted on the outside frame of the robot. This means I need a good bit of shaft space on my gearboxes to mount everything. (Subtract an inch as the shaft passes through a clearance hole in the frame. However, I’ve ran into a problem and am not sure where to go. On the output shaft of my WCP SS Gearbox (http://www.vexrobotics.com/vexpro/motion/gearboxes/wcp-ss.html) I have mounted 2 15mm pulleys, and need to fit a wheel on the end, but the axle is not long enough. Is it possible to order an extra long output shaft/swap it out for a custom-cut piece of hex shaft from our stock, or should I ditch the gearbox I’m using all together and find something with a longer shaft?

On the page I linked, it says the “extra long shaft” can be cut to length, but it seems that is still too short. Any ideas?

The WCD version of the gearbox should be able to fit your pulleys and a wheel on the end.
It’s a rather simple part to make on the lathe, so it’s definitely possible to make your own.

Thank you for the response, Asid. Our team does not actually have a functional lathe at the moment, so I’d rather try to find a pre-made solution. I am using 1.5" thick Colson wheels, and 30mm of pulley thickness total, and it does not fit on the part I downloaded. Does the actual shaft come longer than represented in the CAD?

I would assume not, but I haven’t checked.
Would it be possible to use 9mm pulleys instead, or 7/8" wide colson wheels? Have you though about using chain instead of belts?
The end is tapped for 1/4-20s, so it might be possible to stick a 1/2" Vex tube axle on the end and throw a 1/4-20 on it.

You should be able to make an output shaft for this gearbox with just a piece of hex or Thunderhex stock. However, you will need to replace the 1/2" round bearing on the motor plate with a 1/2" hex or Thunderhex bearing, and add some method of stopping the shaft from sliding out (shaft collar or a bolt+washer in the end of the shaft).

Asid: I’m a little weary of using 9mm belts to transmit power to the outer wheels, but it would give me plenty of space. Is there any risk of the belts breaking if I go down a size? I could try smaller wheels, that seems like a pretty elegant solution.

Just figured out how to quote. That’s neat. And this seems like a pretty good solution, I will try it if I can’t find a fix within the stock gearbox.

Assuming 5M HTD Belt; the Tensile Strength of the 9mm width belt is around 780 lbs. Provided you use reasonably sized pulleys and don’t over tension it, 9 mm belt should be fine in this application.

You can definitely screw this up, and several teams have had failures using 9mm belts and 24T or smaller pulleys in the drivetrain. If you’re not sure I really would go 15mm.

That this gearbox doesn’t fit two 15mm pulleys is one of the reasons I haven’t used it. It’s a lot of space to ask for, true.

Use of small pulleys really decreases the torque rating of these belt drives. Compared to a 36T pulley a 24T is about a 53% reduction in capability, while a 22T is down to ~47%.

A 15 mm wide belt is 40% stronger than a 9 mm belt. The extra strength may help some, but it is still poor compensation for a bad design.

For smaller pulley sizes, it would be better to go down to a 3M belt since you would have more teeth in contact and there would be less of torque rating decrement.

I would encourage everyone designing belt drives to take some time and understand the product, and engineering equations involved, before slapping it in to their design and hoping it will work.

How is going to a 40% stronger belt “compensation for a bad design”, but going to a 47% stronger pulley size the proper decision? They seem to both fix the problem about the same, it’s just horizontal space versus vertical space which is design dependent.

I’ve built several belt drives - 15mm wide belts and 24T pulleys on a WCD with 4-6" wheels have been completely issue free for us. 9mm/24T has not treated us as well. I don’t see this as a bad design decision to use wider belts when we don’t have the clearance for bigger pulleys.

Definitely look at the formulas and understand the rating for each belt - and then also remember that the resources specified by the manufacturer are generally for machines with 20x the run time of the average FRC robot, so you can go a bit closer on the safety margin than you might expect.

You have a Design choice of v belts, synchronous belt, poly cord, and chain.

Using 5M HTD belt on small pulleys with only 12 teeth engaged and the small 1.5" turning diameter is not the intended or recommend use. The fact that you broke 9mm wide belts, suggests you probably were just getting by with 15mm belt with very little margin.

These HTD belts have some elasticity which reduces the effective number of teeth in mesh on the pulley. This effect is magnified when the number of total teeth is under 30.

Going to a 3M belt would provide additional margin at a given pulley diameter, as there would be more teeth engaged.

SDP-SI recommends a minimum of 6 teeth engaged and has a correction factor for less engagement: http://www.sdp-si.com/D265/PDF/D265T146.pdf

Going to 3mm belt has a significant reduction in strength. Belts are recommended to be used at 1/15 of their breaking strength (Timing Belts and Pulleys - D265 Design And Installation Suggestions - Technical Section T-42)

3mm HTD breaking strength http://www.sdp-si.com/D265/PDF/D265P1023.pdf
6mm = 213 lbs.
9mm = 320 lbs.

5mm HTD breaking strength http://www.sdp-si.com/D265/PDF/D265P1027.pdf
9mm = 640 lbs.
15mm = 1068 lbs.

3mm HTD working strength (1/15 of breaking strength)
6mm = 14.2 lbs.
9mm = 21.3 lbs.

5mm HTD working strength
9mm = 42.6 lbs.
15mm = 71.2 lbs.

You can probably get away with more than their recommended 1/15th reserve, but that would require some testing and development.

The Problem with FRC usage (except 2015) is you will eventually get into some form of low speed, high torque pushing contest.

Assuming a 120 robot with bumpers and a battery weight, you could assume around 35 pounds of normal force per wheel (w/ four in contact). Assuming a coefficient of traction equal to 1, that would provide a traction limit force of 35pounds. With eight inch tires, that’s a traction torque limit of 140 in*lbs

A two CIM drive shifted to a low speed ratio of 24:1 has a stall torque much higher than this per wheel. Thus the wheel traction becomes limiting factor rather than the motor torque in most cases.

A 24T Pulley (PD of 1.5") has a resulting belt tension of

 Ftension = T/r = (140 in*lbs)/0.75in = 187 lbs.

The 140 inlb torque value is way over the recommended limit for a 24T Pulley for either 9MM (36.9 inlbs) or 15 MM (70.7in*lbs). In a very defense game you could potentially break either belt width.

Larger pulleys have higher torque limits, and lower resulting belt tension.

If you run those calculations with the kit drive train pulley sizes and wheel diameters, does a 9mm or 15mm belt meet the strength requirements? Is pulley size the only way to get a drive system from failing?

What happens if all the weight of the robot rocks on two wheels? What happens if all the weight of the robot rocks on one wheel?

9mm belts are great, but as Chris said can easily get messed up if you’re not sure what you’re doing. We ran a very successful 6wd where the middle wheel used a 15mm pully that held two 9mm belts going to the outer wheels. Its just another option. If it doesn’t feel safe to you, stick with the 15!

The ratings are for a maximum recommended load over a long life, not ultimate tensile strength, so the belts aren’t going to snap the instant your robot tips on two wheels.

Dozens of teams have run 15mm wide belts on 24T and larger pulleys and 9mm wide belts on even larger pulleys. These aren’t about to snap. 9mm belts on 24T pulleys are pushing it. Smaller pulleys can lead to failure in a drivetrain. Load ratings are conservative and for use cases that aren’t like FRC - using them as gospel here isn’t necessary, and it isn’t a poor decision to go with what’s worked for many teams for several seasons, really. We push the rated limits of 25 chain all the time too.

Agreed, the ratings are for 3,000 hours of use, where an frc bot would see 10-20 conservatively.

What about the 10th, 100th, or 1,000th time it rocks? The point is there is a lot of advice in this thread, it could be very helpful or not so helpful. Most of it is not fully quantified. Yes, increasing wheel diameter increases load on the belts. Yes, increasing pulley diameter decreases load on the belt. Yes, increasing belt width increases srength. Yes, increasing tooth engagement helps with power transmission. It’s basic physics, but what is the design point? How much load is the difference between a belt breaking or lasting? What tooth engagement is necessary? Taking a look at the numbers can be very helpful in understand the failure modes.

Define pushing it, what does that correlate to? When does it become an acceptable FRC design and the requirements it is held to?

How do you quantify a 24t pulley 15mm setup: What size wheels? What power input from the gearbox? That can be very helpful information to say we loaded our belts to X pounds and didn’t see a a failure.

9mm wide belts with 24t pulleys are the setup I’ve had the least experience with, but in a 6WD WCD with 4" wheels, there were issues with ratcheting, and I believe belt failure was inevitable at some point. I have heard of several other teams failing belts in this configuration.

How do you quantify a 24t pulley 15mm setup: What size wheels? What power input from the gearbox? That can be very helpful information to say we loaded our belts to X pounds and didn’t see a a failure.

In 2012, my old team ran a 4WD with 6" wheels, neither wheel direct driven, using a pair of 15mm belts and 27T pulleys. All good.

In 2013, we ran 6" wheels on a 4WD with a 4 CIM drivetrain, single speed, probably a 10:1 gear ratio or so. 15mm wide / 24T pulleys. Majority of the weight was on the direct driven wheel, so maybe not a perfect case study.

In 2014, we ran 4" wheels on a 6WD with a 6 CIM drivetrain, maybe a 7.5:1 gear ratio? (around 12 FPS). This robot had the most run time of any of our robots, with at least 100 hours put in to practicing with both a competition and a practice robot, and obviously that game had plenty of pushing defense, so in a way this was the hardest test of this belt setup yet. Never ratcheted, never failed, no issues at all.