This is a customisable GT2 belt / pulley / idler with the following additional features:
- models the GT2 belt using arcs, based on the publicly available belt profile
- draws the teeth using involute curves, which models the belt release which is dependent on the gear's radius (see pictures)*
- M3 grub screws / offset screws / no screws (for pulleys)
*I'm not a mechanical engineer.
- % detail to render the teeth (an arbitrary setting that affects the number of segments in a circle + number of steps to model the involute curves. Attached stls are rendered at 100%)
- shaft diameter. Set this to your bearing OD if you're making idlers
- a lip to encourage the belt to stay centered
- nut type (square, hexagonal, none for idlers)
- offset the mounting screws. Set this to false, if you intend to use grub screws that are in-line with the teeth. The customiser will override this if there is insufficient space for placement of nuts
Suggested print settings:
- 0.25mm nozzle
- 0.1mm layer height
- no supports
- 5 perimeters
- octogram spiral top/bottom infill
- 30% gyroid infill
- PETG [use a strong plastic that doesn't warp]
I used the GT2 belt profile from here: https://www.adafruit.com/product/1184
This is the same GT2 belt profile published in multiple places.
There are two modifications.
The tip of the teeth is filleted (<0.15mm) to follow the belt profile. As gear sizes get smaller, I decrease this radius so that it remains tangential to the gear's outer radius (since this part will also curve as the belt wraps around). This probably isn't the best approach, but should be more correct than 'wrapping gear teeth around the circumference'. I couldn't do the trigonometry, so I linear regressed an accurate solution instead. An FDM printer nozzle (and molten plastic surface tension) limits the printable accuracy here anyway.
The belt teeth needs clearance as it engages/disengages the gear. Thus the gear 'valleys' need to be wider (the smaller the gear, the wider the valleys need to be). I modelled this with involute curves. I can't do the calculus to model the curves, eg. https://www.youtube.com/results?search_query=NIGw_dlEzQ4 , so I modelled the teeth at various stages as it disengages instead (see picture).
GT2 (half) belt tooth travel path as it disengages from a pulley
Closeup of (half) tooth following an involute curve path