Continuing my search for a gearset that can be 3D printed and handle a 20Nm (14.7 lb-ft) load (~40:1 ratio, ~140x140x80mm volume), I toy-modeled this Carrier-Driven Split-Ring Planetary gearset (Sunless Wolfrom drive) to familiarise myself with the 1-tooth-off compound principle.
It's inspired from David Hartkop's Carrier-Driven Split-Sun Epicyclic Compound gearset, and is very similar to a Wolfrom drive
However, the sun gear of most planetary designs is often the weakest link when handling a high torque load. Because of the force concentration on this small single gear, its axle usually shears off or its teeth break.
So, instead, I replaced David's split suns by large split rings and positioned the planets inside, like in a Wolfrom drive, but drove the planets via the disc that carries them instead of via a sun gear (ie hypocyclic).
See the "How it works" section below for more on operation principles.
For now it's just a concept toy but, seeing the potential, I am now prototyping sturdier versions with bearings and Nylon gears, and putting them to the test. Results below as they come.
The Fusion360 CAD (very messy), STEP and 3MF files are on Github and Makerworld
Feel free to copy, remix, whatever (non-commercially)
Happy 3D printing to all !
______________________________________________
Settings
- 0.4mm nozzle, 0.24 layer height
- 2 walls, 3 top, 2 bottom
- 25% Adaptive Cubic infill
- no support
Align the dimples before gluing the parts together:
Done. Et c'est parti manège !
Manual rotation in action:
Now working on a full version with bearings:
update The following version that includes bearings painstakingly lifted 175lbs via a pulley, i.e. 87lbs (40Kg) straight, thus delivering 16Nm (11.8lbf.ft) of torque. But this test started destroying the tooth surfaces as temperature from friction built up, and the whole thing was as groaning and creaking as my knees after hiking Longs Peak.
It was printed in PLA though, which has a very low 50-70°C heat deflection temp, depending on pressure. PA Nylon filament would fare better, will try it next.
Drive quickspecs: 120mm (4.7") outer ring diameter, 39:1 reduction ratio, 6 planets, 11 planet teeth, 10mm (0.4") wide tooth contact area on each ring, 80mm (~3") drum diameter.
update 2 The Sunlun easyPA Nylon print passed with flying colors. It lifted 87lbs straight without any straining. And, although I tested it 10 times longer than the PLA version, upon disassembly I couldn't find any sign of wear or thermal stress. Nylon really is an affordable, easy to print and robust filament for gears.
EasyPA's Safety Data Sheet shows a composition of 99% PA + 1% additives. No description for the additives. The PA's CAS # identifies it as PA-66. Surprising, seeing the unbelievably forgiving print requirements (50°C bed, no chamber), I was expecting a blend with other plastics like PC. But who knows, since the SDS lists an HDT of 50°C @ 0.45Mpa, which is obviously wrong, it casts doubt on everything else. More on PolyAmides: types of Nylon used in 3D printing (Polymaker)
But Nylon can still be a pain to use. Requires hours of drying before printing, every single time, even in Colorado's low 20-30% relative humidity. So, will try ABS next.
update 3 The test with ABS was quite successful too. It passed the same number of cycles as Nylon, with no straining, using the same 400rpm slow drill speed. However, after switching to 1500rpm the smell of ABS appeared in a matter of seconds.
After a handful more high speed cycles, bits of ABS started pooling in the casing. And a few teeth showed signs of melted bits, although their overall surface seemed fine. Wish I had done that high speed test with Nylon too, but I suspect it'd have fared better.
Diagram
 |
| Sunless Wolfrom gear train |
How It Works
The key principle lies in the 2 rings that are off by 1 tooth, while the planets keep the same number of teeth on both their stator-ring-riding and output-ring-riding sections.
As the planets revolve around the fixed stator ring, they also make the output ring rotate because of the 1 tooth difference. After a full rotation by the planets around the stator, the ouput ring has now rotated by just 1 tooth. So, my toy version having a 60 / 59 ring setup, its reduction ratio is 59:1. Mind-boggling...
Watch David's video at 3:43 to see the 1-tooth-over-360-rotation principle in action, and his supplemental video on how to calculate the number of teeth and the module values.
Digression: note that several variations are possible: more than 1 tooth off like in David's 50/40 supplement above, different planet pitch diameters between their stator and output sections, different numbers for stator & output planet teeth, etc. For now, the model herein sticks with rings off by 1 tooth and planets sporting an equal number of teeth.
For giggles I also added a first stage by adding teeth to the planets carrier, making it a regular spur gear stage with a reduction ratio of 76 / 12 = 6.3. So this drive's total ratio is 59 x 72 / 12 = 373.7
Of course, picking 12 teeth for the planets when one of the rings has 60 teeth is a bad idea as these numbers are a multiple of each other. That will lead to repetitive uneven wear patterns on the teeth.
Instead, 9 or 11 teeth would have been better as 59/9= 6.5555... and 60/9= 6.6666..., or 59/11= 5.3636... and 60/11= 5.4545... No big deal for a toy, but for an actual application 60/12 teeth should be avoided.
Note that in this 1-ring-tooth-off scenario, the number of teeth on the planets themselves doesn't matter, as long as it's the same on both their stator and output sections. What's crucial though, is that each planet but one gets its output teeth rotated compared to its stator teeth, depending on its polar location on the carrier.
Here, since there are only 2 planets, the upper one in the pic above is straight, no rotation between the teeth on its bottom section (stator) and those on its upper section (output), while the lower planet shows a rotation of half a tooth:
That's because at the 180° point opposite the upper planet, the output ring itself is now half a tooth off relative to the stator ring. So the rotation angle for the lower planet's ouput teeth section must be 360 / 2 / 12 = 15° so as to rotate in sync with both rings.
If 2 more planets were added between the existing ones, ie at 90°, their output section would be rotated by 7.5° and 22.5° respectively. Etc.
In conclusion: the massive advantage of this design is that the load torque (output) is applied to the large-radius rings and is then spread across many planets, while each planet transfers the torque via its own axle to the carrier disc. So, nowhere is the resulting torque transfered to a unique and small sun gear like in a regular planetary gearset.
Anyway, enough of this as my head now hurts from all these tooth, teeth and rotate-all-but-one non-sense. Glad that CAD allows me to skip all that and try things monkey-poking-in-the-dark style 😅
By the way, shoutout to the extremely useful ME Virtuoso website where one can parametrically simulate their own planetary / wolfrom / cycloidal / harmonic / wave drives, and
then download the STEP / DXF files. Some of which he also CNC'd or 3D
printed, and tested on the ProMakina Youtube channel
Comments
Post a Comment