New user

[KirkB]

I looked at Gearotic years ago for intellectual interest, but decided to pull the switch and purchase only yesterday.  I am a hobby machinist who builds models of steam engines, and many of these have governors.  The motion of the crankshaft is often transmitted by either helical or bevel gears to the governor.  On my two recent builds, I needed to purchase the appropriate gears for seemingly exorbitant prices given the small sizes of the gears.  Being able to make my own seems both economic and true to the hobby ethos of not buying something one can make.

As soon as I can get my 4th axis set up, I think the first project will be to see if I can duplicate the helical gear I purchase previously.

In the interim, after watching several of the tutorials, I "designed" a random round gear to get a feel for how long machining might take using an endmill as opposed to a form cutter.  I chose to find a gear close to 1" in diameter, and ended up with 14 teeth at 17DP.  This fits a 1/16" endmill.  In brass, I used F&S from GWizard, and found that in 2.5D machine time would be 21 minutes.  On the 4th axis with rooting and shave, it's 127 minutes.

If change the tooth count to 12 and DP to 14, the useable tool size grows to 3/32" allowing faster feeds.  Are there any design rules for number of teeth assuming the gear ratio works and the shaft separation isn't fixed?

Finally, is the ability to generate 4th axis g-code now available in the current version for bevel gears?

[ArtF]

Hi Kirk:

>>As soon as I can get my 4th axis set up, I think the first project will be to see if I can duplicate the helical gear >>I purchase previously.

Shaving a 4th axis helical works pretty good. It was only recently fixed as it had some errors on smaller gears,
but a forum member here worked with me to find the errors and the helicals turned out proved to be good.

>> In brass, I used F&S from GWizard, and found that in 2.5D machine time would be 21 minutes.  On the 4th >>axis with rooting and shave, it's 127 minutes.
>>If change the tooth count to 12 and DP to 14, the useable tool size grows to 3/32" allowing faster feeds.  Are >>there any design rules for number of teeth assuming the gear ratio works and the shaft separation isn't fixed?

  No hard rules, BUT it is generally recommended by gear specs not to use too few. Trochoidals weaken a tooth
and occur only in the roots of gears with low tooth counts. Not a rule, more a design suggestion but for maximum
strength the number of teeth should rarely be less than 12.  How many is optimal depends a huge amount on
all the other consideration of the train, power, speed, etc. In general the simpler the root the stronger the tooth
and higher tooth counts give simpler roots.

>>Finally, is the ability to generate 4th axis g-code now available in the current version for bevel gears?

  No. Bevels are incredibly hard things to machine in 4th axis generally. Ive never attempted to make
a code generator for it. It is possible to machine them as 3d machining but require fixing up for the
tool marks inevitable on the faces. Its been discussed here a few times as to the difficulty and suggestions
for ways to do it perhaps, but Ive never verified such schemes would work generally.

  Im currently making some with round teeth as theory shows if one had balls as the tooth ends they
would in theory work at any angle from 0 - 90 degrees with the same friction and single point contact
on the teeth. I hope to show a video of these crown type bevels in a video soon, but more as a
suggestion of how to do decorative gearing on angles. I dont think they would carry high loads though
as they have only a single point of contact between teeth, a sphere pushing on another sphere.

Note for those interested in crown gearing (quite a few have asked)..

  To make spherical teeth gears easily as I can, I am machining slots into a blank and installing
brass pins with a rounded end. This makes spherical teeth. The pitch diameter (the point at which
such teeth mesh) can be found using Gearotic using a gear Module setting equal to the formula

  Module = (Diameter of Rod * 2) / PI. So a 1/4" brass rod for example is a 
module of  (6.35mm*2) / PI = 4.0425, though I use 2.2 for backlash and a loose tooth fit
as Im making a clock type mechanism. Ill post a photo of such a gear set below with
20% designed in backlash.These two gears will have a hinge so they run at any angle
their tilted to up to 90 degrees. For hinged gears the hinge point must be inline with the
tooth contact point at pitch radius.  Ill post a video when and if it runs. :)


  So to date, bevels are still a hard thing to make unless 3d printed. Helicals can be 4th axis shaved
as they are really just a twisted involute.

Art

[KirkB]

Thanks for the info.  I'm going to create STLs of the bevel gear and pinion I bought and see how the cost of printing in brass at Shapeways compares.

I have a high-speed spindle that I can attach to my mill, which at 30K rpm reduces the machining time of my sample round gear to 18 minutes.

One thing that I am wondering about is whether shaving both sides of gap is the best method.  Since my 4th axis uses a worm gear, might not its backlash cause problems on small teeth?

[ArtF]

Kirk:

  Yes, backlash can be an issue, doesnt matter how many axis.I havent had the issue
bad enough to comment on how much difference it makes based on how much
backlash. Id be interested in any experience you may have base don it and how
much you experience. It has never really been mentioned much around here as
an issue, though I do recall it in a post or two awhile back.

Art