Setting gear mesh properly

[Phil Tucker]

What's the proper way to set gear tooth mesh? In other words, how far apart should the two gears be spaced? If they are too close together they start to bind up and don't spin freely. But if they are too far apart so that only the tips of the teeth touch, they make noise and wear out prematurely.

I usually set the spacing by simply "eyeballing it" until it seems about right. I've also heard people recommend putting a piece of paper between the gears to add about .003 but I'm sure there's a more scientific way to do it. Any mechanical engineers out there who want to enlighten me on the subject?

[Bill Shields]

The piece of paper is just fine....don't worry about adding 0.003

For everything we do, that is more than close enough.

this assumes that the gears are truly concentric and will not bind...

if they do, then jam the paper in at the 'tight spot'...and go have a beer...

[neanderman]

Paper is the technique I've always heard.

[Bill Shields]

on really big gearboxes, where it is necessary to ensure that the shafts are in parallel alignment (gearbox not warped), we would take scotch tape and put it on either side of the gear tooth and smear one with red the other with blue.

this would ensure that we had uniform contact across the entire face...

[Phil Tucker]

Thanks for the good advice! I'll stick with the paper method and try not to overthink this. Often, the best solutions are the simplest ones.

[Bill Shields]

you could always jam the gears tightly together, lube them up with some lapping compound and set them to work.

in a few hours they would be 'set' and ready for a wash out and proper lube...

[Pontiacguy1]

Best method I've heard of, and actually seen used, was for setting the pinion depth and lash on an automotive rear end.
The guy basically would take a piece of newspaper, cut it and then roll the strip of newspaper between the two gears.

If it cut the newspaper anywhere, it was too tight and needed a shim taken out.
If it creased the paper so that you could clearly see the tooth pattern, then it was set correctly
If it didn't leave clean creases in the paper, it was too loose and a shim needed to be added.

[JimJam]

The proper way to set gears to mesh is to add the "Pitch diameter" of the gears. This diameter is a little smaller than the outside diameter of the gear. If the pitch diameter is 4.75 on one gear, and 3.25 on the pinion gear, add the pitch diameter, in this case 8", then drill your axle holes 8" apart. Perfect.

See this link for more information from Boston Gear:

https://www.bostongear.com/pdf/gear_theory.pdf

- Jimi -
Harvard, MA

[Bill Shields]

ah yes - gear theory, I remember it well....

which is absolutely correct as long as you have a way to measure the distance and ASSUME that the shafts are exactly at the intended orientation (parallel / skew / designed angle)....and the gears are spot on 'tolerance'.

If you are building a parallel shaft gearbox, then you know where to put the shafts given the gears, given the design PD...and pray that someone does not twist the gearbox when mounting it.

I have spent many days in the 'outside world' with a gearbox the size of an old VW camper, bolted to a steel frame mounted on a concrete bed that may or may not be absolutely flat..and trying to get the gears to mesh correctly by warping the gearbox back into it's intended shape.

Sometimes you have to INTENTIONALLY WARP a gearbox out of mesh when it is COLD so that when it (along with the gas turbine driving it) gets HOT it is where you want it.

We even had procedures whereby gear mesh was only allowed to be set at night so that the sun shining on one side of the gearbox or another would not upset things...really!

With an automotive differential, everything can be 'within limits' (gears and shaft positions / angles) and you can very easily have a binding gear set.

It is as much an art as a science, even with today's modern laser measuring devices.