Kites and Darts are a type of non-periodic tiling invented by Roger Penrose. The tiles cover the plane in an infinite variety of beautiful, non-repeating patterns. The Golden Ratio (1/2)*(1+sqrt(5)) features prominently in the tiles, in their construction and in their patterns. In the limit, the ratio of kites to darts is the golden ratio. So if you print 1000 darts you should print 1618 kites. There are seven ways the tiles can be arranged around a vertex. These seven patterns were given names by John Conway and appear in the stl files: sun, star, ace, deuce, jack, queen, king.

The circular markings show how the edges of the tiles line up. The circular patterns themselves are interesting, and if you want you can print and use only the markings.

See these links for more information:
intendo.net/penrose/info.html
mathworld.wolfram.com/PenroseTiles.html
en.wikipedia.org/wiki/Penrose_tiling

This experimental motor is another attempt at building a printable actuator. The advantage this design might have over a motor like this one thingiverse.com/thing:11164 is that it does not use magnets. This is not really a practical motor, but others interested in this topic might be able to build on this design.

It was designed for use with this electromagnet thingiverse.com/thing:9608 and actually operated for a very brief moment before the electromagnet burned out. The electromagnet is driven by a pulsed DC current (a square wave at 5 to 10 Hz). With each pulse, the arm is attracted to the electromagnet, and it advances the ratchet wheel by one tooth. Gravity returns the arm to its rest position. The motor sits in the funny-looking wheel stand so that different inclinations can be tried. This is a simple way to adjust the return force on the arm.

A video of the motor running with a hand-wound electromagnet can be seen here youtube.com/watch?v=bqnEIj00C6k

The electromagnet uses magnet wire and a half-inch steel bolt as a core. I did not count the turns, but it is just a simple old electromagnet. It draws a few amps at 6V. There is a steel washer glued to the moving arm.

The arrangement of ratchet wheel and pawls was inspired by
thingiverse.com/thing:329

There are some other ratchet motor designs like this one
thingiverse.com/thing:946

and this one vimeo.com/3169018

There are lots of other actuator ideas listed on the
Reprap Actuator Fabrication Page
reprap.org/wiki/Actuator_Fabrication

This is an electric motor made from a printed circuit board, some 3D printed components, 44 magnets, 44 steel washers, and a handful of electronic components for drive circuitry.

It is similar in design to thingiverse.com/thing:802 but it is much easier to make and the performance is much better. Maximum mechanical power output is about 600 mW. Top no-load speed is about 2000 rpm. This is probably enough power to do something useful.

Also provided is an open source script that runs on Matlab or GNU Octave to generate custom motor coil patterns. The script will export CAM files in KiCad or EAGLE format, so you can fabricate coil patterns of your own liking.

While this motor used a commercially produced PCB for the coilplate, the idea is that users can produce functional motors using nothing more than their own 3D printers.

See reprap.org/wiki/Automated_Circuitry_Making for an overview of using Reprap-style printers to fabricate circuit boards.

Videos of the motor in operation are at
youtube.com/watch?v=aIj4dKaEXnU
youtube.com/watch?v=Gj_GX-TvAQU

This work was presented at ASME IDETC 2011. The paper citation is DETC2011-48602, Design of an Electromagnetic Actuator Suitable for Production by Rapid Prototyping, by Matthew Moses and Gregory S. Chirikjian

This is a follow-on to thingiverse.com/thing:904

It is a cylinder with a helical channel. The channel is designed so that the part can be printed without any internal support material. This was printed on a Dimension uPrint with zero internal support material.

The hollow channel can then be filled with low-melting-temperature metal alloy to form an electromagnet.

Due to the low number of turns, the performance of the electromagnet is poor. This is really just a proof-of-principle of the fabrication method, not much of a practical electromagnet.

Malone and Lipson report better results using similar materials with a layer-by-layer approach in the paper "Multi-material Freeform Fabrication of Active Systems" which can be found here: creativemachines.cornell.edu/publications

The SpoolHead is another promising approach for making embedded conductors. Info here: reprap.org/wiki/SpoolHead

UPDATE June 27 2011 - Also be sure to check out Rhys Jones' new results on the RepRap page here: blog.reprap.org/2011/06/new-approach-to-printing-metals.html

This is a derivative of
Chainmail by Zomboe
thingiverse.com/thing:8724

While Zomboe's thing uses square links on a square lattice, this one uses triangle links on a hexagonal lattice.

UPDATE June 6 2011 - At Zomboe's suggestion I chaged things so the vertical connections are trapezoids instead of small triangles. Hopefully this will make it easier to print. New Stls and OpenSCAD have been uploaded; the new files have "v2" in the filename.

With the recent flurry of screw-related things I decided I should upload this before it was too late. As Syvwlch said, publish or perish! This is a screw-based connector system intended for use with a modular robotics project. The parts were originally designed in Alibre. The raw Alibre files are pretty hacked-up, so uploading stl-only for now. Of course, an OpenSCAD version of these parts would be fun also, but that will have to wait.

A robot made of components based on this connection system can be seen here:
youtube.com/watch?v=wUivg15i2-Y

There is more information in this short paper:
Simple Components for a Reconfigurable Modular Robotic System
which can be found here
custer.lcsr.jhu.edu/Publications#Metamorphic_Robots

By request of Syvwlch ( thingiverse.com/syvwlch ) and WilliamAAdams ( thingiverse.com/WilliamAAdams ), here is a stand-alone public-domain OpenSCAD cycloidal speed reducer. As with the Wankel Engine and Roots Blower I recently posted, this is intended more as an example of an interesting mechanism than as a practical device. If you want a practical printable speed reducer, you might consider one of the other alternatives like

the worm drives on this Tank thingiverse.com/thing:8080 or
differential planetary gears thingiverse.com/thing:7390 or
cascaded spur gears thingiverse.com/thing:7379 or
this planetary gear reducer thingiverse.com/thing:8460

There are several cycloidal-type mechanisms already on Thingiverse, such as
thingiverse.com/thing:3617 and thingiverse.com/thing:3736

There are also several interesting external sites like:
zincland.com/hypocycloid/
fabricationsofthemind.com/2010/07/09/extruder-design-1-printable-1001-hypocycloidal-gearbox/
github.com/triffid/Differential_Hypocycloid
reprap.org/wiki/Differential_Hypocycloid
en.wikipedia.org/wiki/Gerotor
en.wikipedia.org/wiki/Gear_pump

and many many interesting youtube videos such as
youtube.com/watch?v=bRn1K2XeWVE
youtube.com/watch?v=3WvPF6uGCq4
youtube.com/watch?v=CG2sPuqEXBg
youtube.com/watch?v=AMtyFwMDL7w
youtube.com/watch?v=h236SP86nnQ

This present script is based on a design by M.F. Hill described in his 1928 patent "Internal Rotor", number 1,682,563:

google.com/patents/about?id=mdF5AAAAEBAJ&dq=1682563

Note that this design is based on an offset hypocycloid, similar to Figure I in Hill's patent. Most of the contemporary designs appear to be based on an offset epicycloid, more closely resembling Figure V in the patent.

The motivated student can modify the code so it generates epicycloidal-based profiles. Hint: start by making a module ``epitrochoidBandFast(n, r, thickness, r_off)". The motivated student could also probably clean up my train-wreck of code and/or figure out how to do arrays in OpenSCAD.

Note also that these rotors can be used for pumps - see the gifs in the comments for an example.

This is a stand-alone Public Domain model of a Wankel Engine in OpenSCAD. This is intended more as an example of an interesting mechanism than as a useful engine.

See en.wikipedia.org/wiki/Wankel_engine

The file uses code appropriated from
Leemon Baird's PublicDomainGearV1.1.scad
which can be found here:
thingiverse.com/thing:5505

A Roots Blower is an air pump with strangely shaped rotors. They are used in automotive superchargers and high-vacuum pumps, among other places.

See en.wikipedia.org/wiki/Roots_type_supercharger

This is a stand-alone Public Domain OpenSCAD file you can use to generate these mechanisms. This is intended more as an example of an interesting mechanism than as a useful pump.

The file uses code appropriated from
Leemon Baird's PublicDomainGearV1.1.scad
which can be found here:
thingiverse.com/thing:5505

This is a Public Domain OpenSCAD file with epitrochoids and hypotrochoids.

These shapes are useful for making mechanisms like
Roots Blowers thingiverse.com/thing:8068
Wankel Engines thingiverse.com/thing:8069
and Moineau Pumps thingiverse.com/thing:7958

An EPITROCHOID is a curve traced by a point
fixed at a distance "d"
to the center of a circle of radius "r"
as the circle rolls
outside another circle of radius "R".

An HYPOTROCHOID is a curve traced by a point
fixed at a distance "d"
to the center of a circle of radius "r"
as the circle rolls
inside another circle of radius "R".

An EPICYCLOID is an epitrochoid with d = r.

An HYPOCYCLOID is an hypotrochoid with d = r.

See en.wikipedia.org/wiki/Epitrochoid
and en.wikipedia.org/wiki/Hypotrochoid

This is a parametric toy model of a piston engine, inspired by Neophyte's Air Powered Motor: thingiverse.com/thing:7792

It is written in OpenSCAD, so you can change all the parameters and even animate it. It is kind of fun to animate it and change parameters on the fly.

All the parts should be easy to print. Clearance between moving parts can be changed parametrically.

Does anyone want to add some valves and make a real one that runs off compressed air?

This project is in the spirit of Fdavies's ( thingiverse.com/fdavies ) work on printable Sarrus linkages such as
thingiverse.com/thing:684
thingiverse.com/thing:1112
thingiverse.com/thing:1425
thingiverse.com/thing:1969

The idea is to build a mechanism that creates precise large-displacement linear motion, without using precision mechanical components like ground metal rods, ball bearings, and so on.

You can see a video of this mechanism at:
youtube.com/watch?v=y6Y6iKvTpIc

The main article is here:
reprap.org/wiki/Compliant_Linear_Motion_Mechanism_1

The waxuum is sort of a reverse-extruder. The idea is that a heated hollow needle is used to remove material from a block of wax in a controlled manner, analogous to conventional machining using an endmill. A vacuum pump* pulls the molten wax through the needle, after which it is deposited in a reservoir where it can be recycled. The wax part can then be used directly, or more likely it can be used as a pattern for a mold. Below is a basic overview of how the process might work in a well-developed system. See video of a simple test at youtube.com/watch?v=rnKD9oIzcnM .

1. A heated basin is filled with wax, which is then allowed to cool. This big block of wax forms the working material for our little setup.

2. A waxuum mounted where the extruder usually goes on your 3D printer carves a master pattern (including containment walls) out of the wax block.

3. Elastomer resin is poured directly into negative pattern in the wax block.

4. When the elastomer cures, it is peeled off the wax master and used as a mold to crank out large numbers of parts.

5. The wax master is remelted within its heated basin. The removed wax in the reservoir is added back. The basin is cooled, and the wax is ready to use again for a new master pattern.

--------------------
ADVANTAGES

1. The wax is melted instead of cut, so high forces are not required at the tool-tip. The low force requirement means that the XYZ positioner can be much more flexible (and cheaper and easier-to-make) than is normally required for conventional machining. Unlike conventional machining, there is no need for chip removal.

2. In many cases the surface finish on parts is better than that on filament-extruded parts, due to the smoothing effect of the hot needle moving over the wax. There are also fewer warping, delamination, and anisotropy problems.

3. You get a mold instead of a part: casting from a pattern can produce parts at a much faster rate than printing. You also have a greater choice of materials (plastics, metals, ceramics, and wax can all be cast in silicone molds).

---------------------
DISADVANTAGES

1. You get a mold instead of a part: there are several additional steps and materials, including the manual work of pouring resins. Sometimes you really want to just print your parts and go.

2. Limited geometry: unless you make multi-part molds with cores and such, you are limited to relatively simple geometry. Hollow and concave parts are difficult.

3. The waxuum is relatively unproven compared to the many filament extruder designs currently in use.

------------------------
RESULTS

The test showed it was feasible to use a tool of this design to make wax patterns. The most important thing to consider seems to be optimizing heat transfer from the needle to the wax, while minimizing heat loss due to movement of cool air through the needle. The thin-walled brass tubing is just barely acceptable for this purpose. The next version will use a material with higher thermal conductivity, such as thick-walled copper tubing. The heater should probably be located outside of the case. Possibly two heaters should be used - one for the needle and one to keep the wax molten inside the case.

------------------------
*Update March 6 2010: See this cool article on converting an aquarium pump to a vacuum pump garage-shoppe.com/wordpress/?p=109

This is a set of plastic snap-together blocks that was designed for a prototype universal constructor. You can see a simulation of a robot made from the blocks here youtube.com/watch?v=LumLzSQE5Vw

You can get more info on the strengths and weaknesses of the actual physical plastic blocks here molecularassembler.com/KSRM/3.18.htm

This is the final episode of a three-part mini-series on making things with low-melting-temperature-alloy. The collection of things here form a hot material transfer system that can be used to extrude (sort of) molten metal or wax. This is inspired by earlier work done by Dr. Sells and Prof. Bowyer reprap.org/bin/view/Main/AutomaticDepositionOfMoltenAlloyInt oACastingChannelToCreateAVerySimple Electro-mechanicalComponent .

As usual, remember that low-melt alloys are toxic and hot enough to cause burns and fires, so only use this material if you are experienced in a lab environment. On the plus side, this system can also be used to make things out of wax, which is much safer to use, although you still have to watch out for burns and fires! Possible uses for a wax-handling system include: 1) automated wax casting; 2) use of wax as a support material.

If you omit the heaters, you can probably use the nozzle and syringe pump to deposit slurries or liquids.

The idea of operation is that a standard RepRap prints a layer of plastic (HDPE, ABS, etc). Then the toolheads are switched, and the metal/wax toolhead deposits molten material in the channels or cavities formed by the plastic. If necessary, the process can be repeated for constructing thicker and more complex parts.

We have had good success using the nozzle as a hand-held tool for transferring metal. For example, we used it to make this thing: thingiverse.com/thing:802 .

We have had somewhat less success using it when mounted to a RepRap, as seen in the pictures. The main problem is poor control of flow-rate (see photo of pattern on hot plate). With some modification (such as a smaller nozzle hole diameter) and tuning of parameters (feed rate, height of nozzle above plate, etc) it might work much better. We RepRapped a simple test piece, manually filled it with metal while it was still on the hotplate, then let it cool (see photo of widget with 25-cent piece). This yielded fairly good results, indicating that an automated system may be able to produce usable parts.

The main components of the system are

1) Heated cup - this is a heated reservoir that contains the molten work material.

2) Heated copper nozzle - the nozzle slurps up material from the cup, then moves to a desired location and deposits the material.

3) Syringe pump - a motor-driven syringe controls air flow in and out of the nozzle, which in turn causes the nozzle to slurp or deposit material.

4) Heated plate - this keeps the work hot to improve the flow of the heated material. It also melts the base of the deposited part, creating a good seal between part and plate, so that molten material does not leak under the part. In general, the heated plate seems to help avoid warping, but the downside is that the bottom of the part starts to ooze outwards after a while.

This is an attempt at a printable/castable electromagnet. The coil is made from plastic castings and low-melt alloy. There is one base-disk and four coil-disks in the assembly. Each coil-disk has 1.5 turns, so there is a total of 6 turns in the coil. The magnetic core is made from iron powder mixed with epoxy.

This electromagnet did not work very well. With a current of tens of amps it can barely move a paperclip. Poor performance is probably mostly due to the low permeability of the magnetic core. It is possible that the same device with a solid iron/steel core would work much better.

Maybe you can improve on the design, or at least avoid making the same mistakes that we did. The construction techniques may also be of some interest. We used plastic castings, but the disks can probably be printed or laser-cut. The use of multiple disks avoids the overhang problem common in 3D printers, but it does not capitalize on the printer's strengths, such as the ability to print a single part with complex internal geometry.

An experimental solid is also uploaded with coils formed from internal cavities. The coil block is about 3cm tall and contains 11.5 turns. The internals were designed to minimize overhangs. I have no idea if this object will print well or not. As a visual aid, an inverse of the coil block is also provided (coilChannel) so you can see what is going on inside the block.

If anyone makes a working electromagnet you could put it on this thing thingiverse.com/thing:329 to make a motor!

This is a possibly-printable electric motor. The motor can be operated as a DC motor or a stepper motor, depending on how you set it up. We built the motor by casting plastic and metal parts, but most of the parts can probably be built with a laser cutter or a Reprap/Cupcake/Fab@home type machine. It runs at about 400rpm at a voltage of 6V and a current draw of 7A (yes, seven amps).

You can see a video of the motor in operation at
youtube.com/watch?v=XSAof007cS4

A video of the first prototype, which is easier to make, is at
youtube.com/watch?v=cHML3gVQ-uU

For more info, also check out our paper
Towards cyclic fabrication systems for modular robotics and rapid manufacturing, by M.S. Moses, H. Yamaguchi, and G.S. Chirikjian. Proceedings of Robotics: Science and Systems, June 2009.
custer.lcsr.jhu.edu/Publications#Robotic_Self-Replication

Before you try to make the motor, you should understand what it is and is not.

*It IS* An experimental design that you can build, try out, and hopefully improve so it does something useful for you.

*It IS NOT* An inexpensive alternative to an off-the-shelf motor. If you need a motor you can put in your project, go buy a motor. This motor is very inefficient, produces low output power, and takes a lot of work to build.