Twisted Extrusion

[Martinm]

I'd like to make a herringbone gear like this: http://www.welter-zahnrad.de/wp-content/uploads/WELTER-Pfeilverzahnung.jpg
Application is a 3D printed microscope focusing rack (like this one https://www.amazon.ca/KOPPACE-Microscope-Focusing-Accuracy-Aperture/dp/B07MDGSF76). So it need not be geometrically perfect, just good enough. I already have the linear part and am now struggling with the gear.
The idea was: Use the gear generator python script to generate a 20 tooth gear sketch (20 mm pitch diameter) and make a twisted extrusion, rotating the profile by one tooth (360°/20). How do I do this?
Using a loft, I'd probably have to split the 18° twist in several steps, making several sketches (on offset planes) rotated by like 6° each, but I'd say this does not give a geometrically correct result (not that it would matter) and seems quite unelegant.
Somewhere in this forum I found an example of a helical boss with profile direction (advanced) set to normal. However I can't get this to work and documentation is zero here.
So what would you recommend? (Of course I could use a "straight" gear. But what would be the fun in that?)

Regards, Martin

 

[Lew_Merrick]

Hi Martin -- I can see two ways to accomplish your "gear," but I would need more information to begin -- What is the "Herringbone Angle" needed and what is the Diametral Pitch (or Module) to which it is being "fut?" [Irrational minds want to know.] -- Lew

 

[bigseb]

Helical boss tooth profile in one direction. Then repeat in the other direction with reverse twist. This makes one tooth. Circular pattern and fill the centre with a revolve/extruded boss.

 

[Martinm]

@seb I have not tried a helical boss with one tooth - is there a reason why the whole gear profile would not work?
@Lew_Merrick : I'd design the herringbone so it overlaps one tooth over the full with. Tooth module is 1 (pi mm pitch, 20 mm dia, 20 teeth) and width of the gear 2*10 mm. I tried it with a loft, which works:

which works, but I think inner and outer "diameters" of this gear are not perfectly cylindrical, so this is just an approximation. Good enough for 3D printing, but I'd like to know how one could do this properly. I guess the helical boss would be the correct option?

 

[bigseb]

A very basic version but you get the idea...

 

[idslk]

Made one with the scripts: Gear Generator and CopySketch
Made a sketch with Gear Generator with zero depth (it made an extrude error...ignored it because i only used the sketch...)
The copied the sketch withCopySketch with rotation
Then a Loft and finaly a mirror of the loft.
The scripts are included in the part.
Regards
Stefan

 

[Martinm]

Hmmmm... is there any reason why the full profile helical boss does not work?

 

[Martinm]

@Stefan: Basically I wanted the twist to be one full tooth, so first and last profile are identical. If you add just one intermediate sketch (rotated one half tooth), then the loft function can't possibly know whether to rotate left or right. I guess this could require a guide curve (?) or just as I did two intermediate sketches.

 

[bigseb]

Hmmmm... is there any reason why the full profile helical boss does not work?
View attachment 29054

As with many things in AD there are limitations as to what can be done. Do one tooth and pattern.

 

[MilesH]

Hmmmm... is there any reason why the full profile helical boss does not work?

The Helical Boss won't allow your Profile Sketch to bridge the helical axis. If you split the Profile leaving a one micron gap either side of the axis, it should work. If you want the tooth profile normal to the helix, you'll need to do a single tooth and pattern as above.

 

[Martinm]

Thanks everyone for the explanations!
- Martin

 

[idslk]

I've appended a part with loft and helix in one part only to compare...

just as I did two intermediate sketches.

yes, for one tooth shift you need more than one step. Loft takes the shortest way without guide, so with the known accuracy the result with a half shift won't be stable...
If you use loft and copysketch with rotate and scale, you can create "gears" like this
Regards
Stefan

 

[Lew_Merrick]

@Lew_Merrick : I'd design the herringbone so it overlaps one tooth over the full with. Tooth module is 1 (pi mm pitch, 20 mm dia, 20 teeth) and width of the gear 2*10 mm. I tried it with a loft, which works:

But what is the "contact angle" and the relative angle between the "herringbone rack" feature to which you are mating?

 

[Martinm]

I used a pressure angle of 20° (did you mean that with "contact angle") - seems to be common.
The herringbone rack has an offset of 1 tooth (pi mm) over the half width of 10 mm. So angle would be atan(pi/10)= 17.44°.

- Martin

 

[morgandc]

Reading all of this with lots of interest, I am interested in 3d printing a working Buffalo 200 Silent shaft/gear. I have read multiple approaches, but come back to Lew's statements of approximation. Looking at the pictures it appears that this is not an involute gear, so it will be interesting to make. Assuming I can produce the part in Alibre, for a "low" speed application is the accuracy close enough? There is a lot of pressure on this gear, so I am guessing the material will have to be metal. I have no idea what the cost is to have a custom gear made the traditional way in a one off manner as needed.

Lew what are the rules of thumb for 3D printing gears and having them work vs traditional production?

 

[Martinm]

These are worm gears, where the teeth slide over another. I doubt these can be 3D printed with reasonable performance as metal laser sintering leaves a very rough surface, in contrast to the ground finish you are looking for here.

 

[morgandc]

As I research more on it....accuracy may not be good for the metals. I wonder how long an industrial strength plastic would last...

 

[Martinm]

Just looked up what a "Buffalo 200 silent" is. It is a 1910 hand-crank blower for small forges (am I correct)? Looking at the photos, there seems to be no place for a worm gear (90° angle between two axes), so I might be wrong and this is simply a helical involute gear (although one with large helical angle), which might even work with 3D printed plastic. Maybe you should post some more pictures including what the remaining gears look like. Is the shaft you are showing the one the final blower wheel sits on? In that case, I'd just print the gear (not the shaft) in PETG. Chances are that this works.

 

[morgandc]

Correct on the item and the helical part. I hope it is involute, but looking at the picture (all I have) I am wondering if it is prior to common use of the involute curve in gear cutting. I don't have the parts in hand right now, I will in a couple months so I am doing some research. It sounds like this shaft/gear is the common issue that takes these out of service. I pulled the pictures from this youtube until I can get my hands on one for measurements.

 

[Lew_Merrick]

Lew what are the rules of thumb for 3D printing gears and having them work vs traditional production?

Martin -- It is not so much "rules" as "practice & experience." Gear designed using Cubic Spline entities will never truly be "involutes" in any trues sense. Therefore you will need to increate "Assembly separation" to provide sufficient Allowance. -- Lew