Printed in PLA, this motor can deliver up to 1.6 Nm (yes, 1.6 Newton meter or 160 Ncm) at 4 bar pressure. Forget all those other designs that can hardly lift a few grams, the real work starts here. See video how it lifts 5+ kg using a 60mm diameter spindle. Without gearbox.
The mechanical design is more or less identical to that of the Bourke engine, which was developed in the 1920s. The only difference is in the gas: here, pressurized air is used instead of fuel combustion.
The engine acts as a rotational stepper motor with step size 90 degrees, so it needs two 5/2-way valves (or equivalent) for control.
Video with assembly instructions: https://www.youtube.com/watch?v=I9DUlbN97yk
Enjoy making this thing, and feel free to ask for help if you need to!
July 8, 2019: Renamed to "R-81 Bourke Engine"
Oct 9, 2019: Renamed to "R-90 Bourke Engine"
anything, but for crankshaft 0.1 mm is recommended
To survive 4+ bar of pressure and deliver 1.6 Nm of torque, solid infill is recommended (especially for the crankshaft). Solid housing also helps making it airtight. Watch out for overextrusion at 100% infill, as this would affect dimensional accuracy.
The large spindle is the only part that needs support in printing.
Applications and assembly instructions, music by Tim de Man
Grind parts until mechanics slide smoothly and clearances between middle and top/bottom housing parts are minimal.
Put four piston plates on the crankshaft. Twist two of them and put piston heads on each side, properly aligned. If the tight fit holes are too small, file the pins and/or holes until it fits; if it is too loose, use glue.
Cut four seals using cutting tool, from about 1mm thick silicone rubber sheet or equivalent.
Drill four air inlets in bottom housing and push pneumatic tubes all way through. If loose: use glue (e.g. Loctite 770+406).
Tap M4 threads in remaining 28 holes in bottom housing.
Optional: apply impregnating agent to make housing airtight (e.g. "waterdichtmaken.nl"). Not really needed if housing is printed solid and/or some leakage is acceptable.
Apply blue silicone on middle housing and (optional) vaseline grease on pistons and crankshaft.
Place crankshaft with pistons in housing. Put seals adjacent to each piston head. Important: large face of seal must face outwards, i.e. towards the pneumatic tubes!
Assemble motor and tighten with M4 screws.
Connect tubes to two 5/2-way valves, and test!
Two spindles are provided, with radius 30mm and 55mm respectively. Use screw(s) in holes to secure it to the shaft.
Crankshaft and pistons
Additional materials and tools
Silicone rubber sheet, around 1mm thick
4mm (outer diameter) pneumatic tubes, e.g. polyurethane tubes or Lego pneumatic tubes
M4 x 30mm bolts (28 pieces)
M4 tap (or use M4 nuts, washers and longer bolts)
Sealant such as blue silicone
(Optional) Lubricant such as vaseline (petroleum jelly)
(Optional) Something to make 3D printed parts airtight, impregnating agent "waterdichtmaken.nl" works fine
Stanley knife with 9mm blade size
Drill with right size for tightly fitting tubes (3.8mm works well for 4mm polyurethane tubes)
Valves (two 5/2-way valves or equivalent)
(Optional) M3 tap and screws with sharp tip, to secure spindle to shaft
(Optional) M6 screws to mount motor on mounting bracket
Cylinder cross-sectional area is 30 mm x 20 mm = 600 mm^2. Stroke size is 16 mm. Crank throw is 8 mm (on rounded side of crank, slightly lower on flatted side).
Theoretical piston force at at 4 bar = 0.4 MPa pressure: F = P A = 0.410^6 Pa 60010^-6 m^2 = 240 N
Theoretical torque: M = F r = 240 N * 0.008 m = 1.9 Nm
Effective torque is a bit lower due to friction, but still around 1.5Nm
Friction and leakage are the main enemies of any rapid prototyped pneumatic device. It is impossible to fully eliminate both of them, but they can be kept to a minimum by good manufacturing and troubleshooting.
- Check rigid parts: operate crankshaft by hand to feel resistance. Find friction surfaces and grind them, or adjust printer settings (finer detail, lower extrusion etc). The motor is quite powerful, so a small amount of resistance is no problem.
- Check seal dimensions. Ideally, 0.2-0.3 mm larger than cavity width and height. The cutter is designed to cut 30.3 x 20.2 mm seals. If too large: reduce seal cutting block dimension (set at 99.5% etc).
- Sealant (blue silicone or equivalent) could add excessive friction if not wiped out of the cylinder. Always operate pneumatic gripper a few times at low pressure to clean cylinder walls using the seal. Leftover sealant collected by the seal is normally no problem.
- Through housing: 3D printed parts are usually porous. Solutions: a) print really 100% solid (at the edge of overextrusion), b) apply imprenating agent such as "waterdichtmaken.nl" for PLA or a dilute ABS+acetone mix for ABS parts, c) any other stuff that fills up the pores (paint? nail varnish? ...)
- Through/along top and bottom cover: if first layer not printed 100% solid, channels will be present along which air can escape. Solution: overextrude first layer, either by extrusion factor or by re-calibrating z offset (lower nozzle). Side-effect: elephant feet. Other solution: fill up channels with stuff.
- Along walls: too low resolution (e.g. 0.3mm) gives ribbed walls which cannot be sealed off effectively with the seal.
- Along seals: check seal dimensions, should be at least 0.2mm larger than cavity. If too small: scale seal cut block up by 0.5% or so.
- Between housing parts: seal off with sealant such as blue silicone; tighten screws.
- Between housing and air tube: irregular hole or too large hole, or too stiff tube. Solutions: a) use sealant, b) glue tube to housing with e.g. Loctite 770+406.
- If one chamber causes excessive leakage, then its seal might have been inserted the wrong way round.
- Seals made from a bicycle tyre's inner tube might be curled, causing problems in sealing
- Too thin or too flexible seals are also ineffective, experiment with different rubber materials