Wide gears

Plastic gears can't transfer much torque; at some point, the plastic somewhere will start to deform.

I was thinking about heavy earth moving equipment, and how they use very wide wheels in order to distribute the weight over a large surface area so that they don't sink into the ground. I assume that this same principle can be applied to plastic gears.

I bet the proper technique, from an engineering point of view, would be to use a stronger material. Making a large, cylindrical-like gear will increase angular momentum and increase the size of whatever you are building, over just using a stronger material. It would also increase friction.

However, we only have plastic, and engineering is also about working within constraints. it makes sense in my mind that the wider a gear is (the height of a cylinder, not the radius) the more surface area it has to grab an axle, and the more surface area it has for tooth contact, and therefore, more torque can be transferred.

This is probably old news to any of you engineers out there, but I'm not an engineer. Are there any formulas to determine the max torque for a certain gear of a certain material?

Edited 2 time(s). Last edit at 07/07/2011 05:06PM by Buback.

Wider gears only help to a point. Unless perfectly aligned you never get contact over the whole face anyway, until the gears have already worn down somewhat.

Quote

This is probably old news to any of you engineers out there

Quote

Are there any formulas to determine the max torque for a certain gear of a certain material?

Yes, there are, but that's a really complex matter. The max tourque is typically limited by when teeth start to break off.

What you're after is (low) wear. A wheel can hold some tourque without breaking, but what helps that if it's worn down within hours?

Increasing the diameter of the bigger wheel lowers wear. Because you need less force at the tooth to get the same tourque at the wheel's center.

Making the wheel thicker also lowers wear. Andrew Diehl is right when talking about misalignment, but I think with printed parts you always have to deal with some initial wear. Even grinded gears wear a bit, if orders of magnitude less.

And then you might want to experiment with lubrication. Lubrication substantially reduces wear, especially if both partners of the movement are of the same material. Vaseline / peroleum jelly should work well.

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

I hadn't thought about the alignment, but that's a good point. herringbone would be a must.

ok yeah i guess i was asking a bit much for a formula. :-)

I can imagine that there are just too many variables when designing a gearbox, so a simple formula is out of the quesiton.

I still feel it's a valid printable solution for high-torque design problems. The big question is where that crossover point is; where it is just easier, cheaper, and/or more efficient to abandon the 'printable' goal and just pay to get the parts made out of metal.

Edited 1 time(s). Last edit at 07/14/2011 04:55PM by Buback.

Herringbone suffers the same alignment issues as any other gears.

There is usually not going to be any good answer for when to switch from printable to bought. Gear design is a very complicated issue, with no clear cut "correct" solution in of itself. Add trying to design for cost into the equation and it becomes a full time job.

I can tell you a few places where it is a no-brainer, but other than that it gets tricky.

Any place the gear will experience any of the following, I would highly recommend buying the gear over printing it.

-Temperatures over 60C
-over a few thousand RPM
-too big to print.
-designed for 10^5 or more cycles.
-requiring high efficiency
-used near organic solvents

Also, for high torque, one of the biggest issues will be attaching the gear to the shaft.

I agree with all of the above, and I'd also add, any application that is safety-critical.

I would not be overly concerned about alignment because after many cycles, the gear teeth will wear down at the 'high points' and the contact area will gradually increase to its maximum. In general the torque capacity of the gear will increase in proportion to the height of the gear.

There are engineering formulas to determine the torque capacity of gears but they always assume a uniform material, not the kind of layered cross-hatched materials that we print gears out of. They should work as an approximation but it's best to perform the test to gather data - though not many people have testing equipment or the patience to run a battery of tests. I don't know if anyone has found the stress limits of reprapped gears yet, or if there've been any documented gear failures. That might indicate that all the printed gear designs have been sufficiently conservative so far.

Printed helical and herringbone gears raise an interesting question. The stress on a helical or herringbone gear is usually higher than on a similar spur gear transferring the same torque, because the contact angle creates an additional thrust force. In a herringbone gear there's two opposite thrust forces that cancel out overall, but still increase the stress. But printed helical gears don't have a true incline to them, instead they are more like a stacked series of thin spur gears. It might be that the contact pressure is somewhere between that of a spur and helical gear.

I've been trying to think of use-cases where you need high-torque/low speed. best i can come up with is some type of wind, water, or human powered grindstone for milling grain. it's probably best for low-tech applications.

High-torque, low speed application would be some kind of rotary actuation, lifting a bridge or maybe hanging a clock's weight.

We once had a customer that produced servo actuators for closing valves measured in meters. Sewage, water and hydro power kind of stuff. The total gear reduction on the big ones was something like 20,000:1. When discussing emergency closures, the engineer said "It won't close quickly, but it will close". They also had hand wheels for closing manually. I wouldn't want that job.

Edited 1 time(s). Last edit at 07/26/2011 11:20AM by Dale Dunn.