An actuator is a device that takes energy in some form and converts it to perform some type of mechanical work. One of the most challenging and exciting frontiers in self-replicating machine research is the development of printable actuators - actuators that can be built using the constrained set of materials and processes available to a RepRap-style 3D printer. This page collects links to various resources that are relevant to this effort. Printable actuators are a big step towards a more self-replicating RepRap.
Where possible a link is provided to CAD models and/or a description that is clear enough such that a motivated experimenter could replicate the work. Currently only a few of these projects are truly "printable" in the pure sense. But even non-printable actuators that can be made by hand are relevant. Intrepid experimenters can figure out how to build by machine what they can build by hand. Some of the links go to academic papers or other sources of information that stop short of detailed design information, but still have useful information.
Feel free to add your own information, but please try to keep this page focused on projects that people can actually build.
Even better, do some experiments yourself and link to the results from this page!
- 1 Forum Threads
- 2 Electromagnetic Actuators
- 3 Electrostatic Actuators
- 4 Pneumatic
- 5 Hydraulic
- 6 Thermal
- 7 Electroactive Polymers (EAP)
- 8 Piezoelectric
- 9 Shape Memory Alloy (SMA)
- 10 Interfacing and Control
- 10.1 Interfacing with fluidics
- 10.2 Scratch-built solenoid valves
- 10.3 Scratch-built solenoid clutches
- 10.4 Magnetorheological (ferrofluid) valves and clutches
- 10.5 Electrorheological valves and clutches
- 10.6 SMA valves and clutches
- 10.7 Electroactive Polymer valves and clutches
- 10.8 Thermally actuated valves and clutches
- 10.9 Piezoelectric valves and clutches
- 10.10 Electrostatic valves and clutches
- 11 Fluidic (fluid logic) and Pure-Mechanical Systems
Printable actuators are a popular topic in the forums. For reference here are some of the threads:
The fabrication of traditional electromagnetic motors uses materials that:
- conduct electricity very well (materials with a low resistivity). See the List of conductive materials.
- that are magnetic (materials with a high remanent magnetisation). See the List of magnetic material.
- that can be easily magnetised and demagnetised (materials with a high relative permeability). See the List of magnetic material.
A possibly-printable Electric Motor on Thingiverse
Brushless DC Motor Parts v0.1 on Thingiverse
A pancake motor using printed circuit board coils: Flat brushless P.C.B motor experiment
Another Printed Circuit Motor
There are hundreds of variants of the famous Beakman motor: The Beakman Motor.
This page would not be complete without an entry on Bill McLennan's hand-made motor that won the first Feynman Prize: Small world's big achievement.
An example of a reed-switch motor: home made brushless motor.
Some of Forrest Higgs's Work on a Printable Linear Stepper
Solenoids and Electromagnets
Electromagnet with castable coils on Thingiverse
Electromagnet Demispool on Thingiverse
Ratchet drives are mechanical transmissions that can be combined with short-travel actuators to allow long-distance precise motion.
Stepping ratchet gear thing on Thingiverse
2 Way Ratchet Mechanism on Thingiverse
Plastic Experiment on Thingiverse
Solenoid-driven rotary motors
These things are just waiting to be turned into stepper motors:
A printable electric motor similar to the four stroke solenoid motor above.
Electrostatic actuators in theory can deliver a force while barely using any power. However large electric fields are generally needed to obtain useful forces. This requires either large voltages (which are dangerous) or very small features. In traditional fabrication methods complexity is expensive and an actuator consisting of many small features is not an option. In 3d printing however complexity is almost free, which opens up the way for large scale electrostatic actuators.
Dielectric elastomer actuators
Dielectric Elastomer actuators consists of a flexible elastomer in between two electrodes. The elastomer makes it possible to apply larger electric fields than would be possible in air. Dielectric elastomer's have been shown to be capable of high energy densities. Often the elastomer's are pre-stretched to reduce the amount of air inside the material and improve the dielectric strength of the material.
In 2008 Jeremy Risner got his PhD at UC Berkeley on the investigation of Dielectric elastomer actuation for printable mechatronics. His PhD mainly focused on the development of structures that could prestretch conventional dielectric elastomer materials. Investigation of dielectric elastomer actuation for printable mechatronics
Duduta et al. from Harvard university managed to make an dielectric elastomer that had a strain of up to 7 percent. They did this by spin coating 37.7 micrometer thick layers of urethane acrylate oligomer and layers of conductive ink. Multilayer Dielectric Elastomers for Fast, Programmable Actuation without Prestretch
Air based electrostatic actuators
Detailed plans for a wooden engine: Air Engine 2
This is a plastic pneumatic engine that looks almost printable: 2 Cylinder Air Engine
A clever way to seal a piston with ferrofluid: Ferrofluid Piston
An engine made with drillpress, file, and solder: 4-cylinder swash plate air engine
A laser-cuttable Simple Air Vane Motor on Thingiverse
RotaVac - A Rotary Vacuum Pump on Thingiverse
A LEGO turbine my first lego turbine
Tesla Turbine V2 on Thingiverse
McKibben-style air muscles
Peristaltic Pump on Thingiverse
These might be printable or laser-cuttable: Gear Pump
Electroactive Polymers (EAP)
Basic info on Electroactive Polymers
A paper about 3D printing electroactive polymers:
A paper about fabricating piezo actuators: Development of PZT and PZN-PT Based Unimorph Actuators for Micromechanical Flapping Mechanisms
Shape Memory Alloy (SMA)
Basic info on Shape memory alloy
Composable Flexible Small Actuators Built from Thin Shape Memory Alloy Sheets by E. Torres-Jara, K. Gilpin, J. Karges, R.J. Wood and D. Rus.
Interfacing and Control
One of the drawbacks of pneumatic, hydraulic and other non-electrical actuators is that they need to be controlled with valves (or clutches in the case of pure mechanical systems). Since we are typically using a computer for control, the valves/clutches must ultimately have an electrical interface. This is often accomplished with some type of solenoid. But if the motivation for using non-electrical actuators is to avoid the need for making coils, we haven't really solved the problem if we still need solenoid-driven valves and clutches.
In principle it is possible to build printable, electrically operated valves and clutches based on thermal, piezoelectric, SMA, electrostatic, electrorheological and other methods. Unfortunately it seems that there is a lot less activity in this area than there is for actuators in general. Some different categories are listed below as place-holders for future results. Maybe with enough material this section could be moved to a new article on transducers.
Interfacing with fluidics
One can interface with fluidic circuit and use the circuit to control actuators.
Fluidic tone and sound sensor can be used for acoustic interfacing.
Putting a heating element in a fluidic amplifier can also be used for electrical interfacing.
Scratch-built solenoid valves
Scratch-built solenoid clutches
Magnetorheological (ferrofluid) valves and clutches
Electrorheological valves and clutches
SMA valves and clutches
Electroactive Polymer valves and clutches
Thermally actuated valves and clutches
Piezoelectric valves and clutches
Electrostatic valves and clutches
Fluidic (fluid logic) and Pure-Mechanical Systems
- Main article: Mechanical Computer
For the truly ambitious among you, consider doing away with the computer entireley. There are several examples of functional fluidic and mechanical computing machines, both classical and modern. Fluidics and mechanical logic do have the disadvantage of requiring more power than electronics and having slow speeds. Power requirements for mechanical and fluidic control systems are expected to be more than the 60 watts prescribed in the Gada prize.
Required reading in this area is Chris Phoenix's design proposal for a fluidic controlled macro-scale machining self-replicator.
One should also be familiar with the Jacquard loom.
Detailed build information for fluidics is hard to find(on the internet at least), but here are a few sources:
"Fluidics Quarterly" provides useful information on fluidic element design
"Microfluidics: History, Theory and Applications" by William B. J. Zimmerman provides a good overview of fluidic circuit design.
"3D Microfluidic Devices Fabricated in Layered Paper and Tape" by Andres W. Martinez, Scott T. Phillips and George M. Whitesides 
There are several printable designs for machines that might be called "printable proto-computers", meaning that they have many of the buildings blocks needed to make a mechanical computer. Some examples of these are:
waterjet clock on Thingiverse
Wind-up Toy on Thingiverse
And if that isn't hard enough for you, you can always make things more interesting by hopping down a few notches in scale: