Updating Official Feature Scripts and Using the Updated Version

greg_frost204

I'm working on a design for compound planetary gear boxes with the intention of 3D printing them: https://cad.onshape.com/documents/9ca554446455ea6f1adca427/v/115c4f54e64515268982634d/e/10697cc92f7ab71e3933066c
I have developed a design using the offical Spur Gear feature script, but that script lacks backlash control, which given the 3D printer manufacturing is problematic.
I know how to change the script to add the backlash feature that I want, but in order to do this I need to make a copy of the script.
Is there a way that I can update my original design so that it uses my copy of the feature script instead of the official version?

NeilCooke

No it’s not possible, but if you share your backlash code I can look to include it in the official feature (no guarantees). 

NeilCooke

No it’s not possible, but if you share your backlash code I can look to include it in the official feature (no guarantees). 

greg_frost204

The backlash code is here:
https://cad.onshape.com/documents/2e3daf86e6e6fa1b9f04b75f/w/4ffc7a791d446675e6376600/e/6342e8414c3ff2e4e06a66ba
It is pretty much a straight up copy of the official spur gear with a single new parameters:

annotation { "Name" : "Backlash" }

        isLength(definition.backlash, BACKLASH_BOUNDS);

const BACKLASH_BOUNDS =

{

            (meter) : [-500, 0.0, 500],

            (centimeter) : 0,

            (millimeter) : 0,

            (inch) : 0

        } as LengthBoundSpec;

With this change to the involute generation code:

    // involute definition math

            var angle = (t + definition.backlash/definition.pitchCircleDiameter/2)*radian;

     
It moves the face of each tooth 1/4 of the specified backlash along the line of the pitch circle.
This means for instance if you specify 0.2mm as the backlash and you specify this on both gears, you get an additional gap between the coupling teeth of approx 0.2mm.

NeilCooke

There are different types of backlash calculations which may also need to be considered: https://khkgears.net/new/gear_knowledge/gear_technical_reference/gear_backlash.html

greg_frost204

It would be awesome if you could include that in the official script. Also, if you are looking to provide a bevel gear script, you might want to check out this:

https://cad.onshape.com/documents/f13ca17c18c21fa266e7385a/v/0468cf9cfdbc892b4ffa7611/e/81a2b382873011411a859c57

greg_frost204

Another suggestion for the spur gear script would be a helix angle parameters (with a check box for double helix).

greg_frost204

The backlash implementation that I provided is what the reference you posted calls Circumferential Backlash. You can also get backlash by changing the distance between gear centres, but in that case, the gears will no longer mesh using the proper involute. Normal backlash should be equivalent, where the Normal backlash is equal to the Circumferential Backlash times the cosine of the pressure angle. For a typical pressure angle this is about 10% less than the Circumferential Backlash. The other types of backlash mentioned by that reference seem inappropriate in the modelling context.

Thanks for that reference too. Some good info there. If you are open to accepting contributions and including them in the official script, I'd be interested in working on the helix change too. At the moment my compound planetary gear uses the feature script to generate the gear, and then uses the gear face to sweep the helix and then delete the original gear. It would be far better to have helix built in. If you have a view on how the helix parameters should be organised, let me know. I was thinking:
Helical Gear (checkbox)
Helix Angle (postitive for clockwise, negative for counter clock wise)
Double Helix (checkbox).

But another way would be to select clockwise or counter clockwise as selectable options (this seems to be what Helix does).

NeilCooke

So I should code both options or just Normal?

greg_frost204

I think from a modelling perspective, you should only code one. Perhaps Normal Backlash would be more intuitive for people who know the fit tolerance of their printers. For Normal Backlash, it would be the same as my implementation, but you would want to divide the adjustment factor by cos(pressure angle).

Either would allow backlash adjustment, whereas the current script does not. A bit of trial and error is probably required to settle on the right backlash for a given printer in either case.

NeilCooke

@greg_frost204 - I've added all your suggestions and shared you into a doc for testing - if you are happy I will publish in the official feature.

greg_frost204

That is brilliant! I have had a play and it seems to work perfectly. The development of the helix feature was a nice surprise.

NeilCooke

Excellent.  I’ll post it tomorrow.