FC Gears: Feedback thread

[freman]

Yes, it has to be a rotation. A linear offset is specified on the reference circle, so is readily convertible to an angle. I would be looking to express this as a property of a pair of gears since you cannot have a backlash on one gear. There may be a question of whether the backlash is split between two gears or just one profile is modified to gain the full backlash for the coupling.

You could envisage three gears where each pair was specified with a different backlash, for whatever reason.

I seem to recall earlier discussion of positioning gears relative to one another, which you see as future way forward. This property similarly seems to need to be treated as a property of two gears.

For helical gears the distance probably needs to taken perpendicular to the face of the teeth, as it would be measured with a feeler gauge for example. The same could apply to bevel gears but may be a little more complicated to calculate. Though calculating the normal vector should be accessible.

[looo]

There may be a question of whether the backlash is split between two gears or just one profile is modified to gain the full backlash for the coupling.

It's known that the bigger gear has normally stronger teeth. So it might be useful to add the backlash only to the stronger gear. So in my eyes it makes sense to set it individually.
But maybe we should call it different for one gear and use the term backlash only for pair of gears.

For helical gears the distance probably needs to taken perpendicular to the face of the teeth, as it would be measured with a feeler gauge for example.

Sounds difficult, and I am not yet sure if I would do it this way.

[freman]

Sounds difficult, and I am not yet sure if I would do it this way.

If you have a 30 degree twist on a helical gear, the divide the required backlash clearance by cos(30) to get the equivalent distance along the reference circle. ( ie it's a bit greater )

[softmoth]

There's a great discussion of backlash on pp. 43–44 of the October–November 1984 issue of Gear Technology, "Design of Involute Gear Teeth" by Fellows Corporation, retrieved from https://www.geartechnology.com/issues/1 ... basics.pdf

A sample: “In checking helical gears, the backlash is measured in the normal plane, instead of in the plane of rotation, as is the case with spur gears. The method just described can be applied satisfactorily to both spur and helical gears. In providing for backlash, it is customary, when a small pinion is to operate with a larger gear, to reduce the thickness of the teeth on the larger gear to provide the necessary backlash, leaving the pinion teeth of standard tooth thickness.”

P.S. The previous article in that series (“Functions of Gearing and Application of the Involute to Gear Teeth”, August–September 1984) is also very worthwhile reading: https://www.geartechnology.com/issues/0 ... basics.pdf

[freman]

excellent reference, thank you.

[looo]

Thanks for this document! This is very interesting.

In providing for backlash, it is customary, when a small pinion is to operate with a larger gear, to reduce the thickness of the teeth on the larger gear to provide the necessary backlash, leaving the pinion teeth of standard tooth thickness.”

This means setting the backlash for one gear must be possible. But I still believe it's better to define backlash for one gear as an angle (because this is the same for all gear types) but maybe rename the property.
For a pair of gear, we can define backlash as the measured distance at the pitch-circle normal to helical angle and compute the angular value for each gear by another proportion property.

BTW : I never wanted to go in such detail. This workbench was my introduction to freecad [1] to understand the python basics.
But feel free to add some motivation [2] or directly create a PR. ; )

[1] https://forum.freecadweb.org/viewtopic. ... =10#p39897
[2] https://liberapay.com/looooo/donate

[Zolko]

Hello,

I'm having a case that I didn't find with the standard PartDesign Involute Gear: I'm being proposed a gear with 186 and module 4, with "addendum modification xm2=2.0". I've found references to this addendum modification in standard gear calculations (for example here) but it's a multiplicative coefficient and I don't think that xm2=2.0 is multiplicative.

Anyway, the PD Involute Gear doesn't seem to have this addendum modification parameter: did I miss a magical parameter or is this addendum modification indeed missing ?

[looo]

Hello,

I'm having a case that I didn't find with the standard PartDesign Involute Gear: I'm being proposed a gear with 186 and module 4, with "addendum modification xm2=2.0". I've found references to this addendum modification in standard gear calculations (for example here) but it's a multiplicative coefficient and I don't think that xm2=2.0 is multiplicative.

Anyway, the PD Involute Gear doesn't seem to have this addendum modification parameter: did I miss a magical parameter or is this addendum modification indeed missing ?

I guess this is what is called "clearence" and "head" in the freecad.gears (previously FCGears) workbench and isn't yet included in the part-design-gear. This is the issue of multiple solutions for the same problem...

[obelisk79]

I really appreciate the features in this WB, they have proven to come in quite handy. Most recently for me with the cycloidal gears.

If it's possible to request a feature, I'd ask that you look into adding advanced forms of worm gear generation or matched worm reducer sets for example, things like double enveloping worm reducers offer mechanical advantages over the more simple standard worm gears, however they also tend to prove to be much more difficult to model with the correct geometry. Particularly properly meshed teeth.

[Zolko]

I'm having a case that I didn't find with the standard PartDesign Involute Gear: I'm being proposed a gear with 186 and module 4, with "addendum modification xm2=2.0".

I guess this is what is called "clearence" and "head" in the freecad.gears (previously FCGears) workbench and isn't yet included in the part-design-gear.

it's called "shift". Thank-you, that works.

gear_shift.png (255.97 KiB) Viewed 1761 times