If you've ever been on an airplane before, you've probably encountered some turbulent flow, and the wings on the airplane start to shake and seem to be very flexible. This back and forth shaking motion is called aeroelastic flutter. Normally you don't want this in an airplane, as if it develops and worsens, it turns into aeroelastic divergence and the wings are in danger of snapping because of the high frequency it is fluttering at.
In the case of these devices, this aeroelastic flutter force is taken advantage of, to produce electricity. These are actually more efficient than a wind turbine, as they work in any wind environment - slow or fast moving, turbulent or calm, and in any direction these will work. With wind turbines, if the air is moving too slow, it just won't move, whereas these flutter energy harvesters will, and if it is too fast, they will have to shut the wind turbine down to prevent damage. Turbines also need one direction of wind, where these devices work with any. With scale taken into account, these actually produce more electricity, but they do produce less current. They are also a lot smaller than wind turbines, meaning they can now be placed in places wind turbines can't, such as in the city on skyscrapers, in your own backyard, or even in the side scoop of a car.
Step 1: Technical Explanation and Materials
At the end of the cantilevered, flexible metal beam is a 3D printed shape (sphere, prism, flat plate, or bullet) which produces drag. Aeroelastic flutter works because of a constructive feedback loop - one waveform is added to another waveform exactly like the first, to produce a waveform that is two times larger than the original. As mentioned earlier, aeroelastic divergence is a worsened stage of aeroelastic flutter, where it is now a destructive feedback loop, where the two waveforms are different, and produce a totally different outcome, which is in turn, destructive. Also, drag is directly proportional to the fluttering motion. There needs to be the right amount of drag to trigger this constructive feedback loop. This is the purpose of the shape on the end. I have included the CAD files, so they can be edited to suit your needs, as it will need to be fine tuned to your energy harvester. When it is in the wind, the beam will start to flutter from side to side, and generate electricity.
To generate electricity, piezoelectric discs (the little white discs or strips on the beam) are glued or fastened to the beam. These are special crystals that, when motion is applied to them (bending, tapping, etc.), the internal molecular structure shifts around, and as it shifts around, it generates electricity. Conversely, if electricity is applied to the crystals they will move. This can be heard audibly in smoke alarms, for example. You probably have used piezoelectric materials before, like in the ignition source for your grill or on a handheld lighter.
1/2" PVC Pipe
Provided CAD and STL Files
ABS 3D Printer Filament (or another strong material, such as nylon - feel free to experiment)
Printed STL Files
Full Wave Bridge Rectifier (individual component, or a homemade circuit)
Ferroelectric Materials (man-made piezoelectric crystals - these are the flexible, film-style ones)
Acetone (for smoothing 3D prints - optional)
Aluminum, Brass, or Copper 3/4" Beams
Step 2: 3D Printing and Post-Processing
All models were designed using Autodesk Fusion 360. Start by loading ABS filament into your 3D printer. I chose ABS because it is tough and resilient, whereas PLA is brittle, and acetone can also be used to smooth out the ABS prints, for a smoother finish that won't effect its ability to flutter. Feel free to experiment with other filaments - I'd like to hear how it goes. I used a Printrbot Simple Metal, printing at 230C with a heated bed temperature of 60C. If oriented the right way, no supports are needed, and make sure to print all models at a layer height of 0.1mm. After printing, treat the models with acetone vapor, for a smooth surface finish.
Step 3: Assemble the Beam
Cut the beam to 18" long, using a hacksaw with a metal-cutting blade. I used aluminium, as it is cheap and still flexible enough. Brass is also an option, but as soon as it gets thicker, it because much less flexible. If it is too thin, it could bend towards one side when fluttering in the wind. Copper is almost too flexible, but if it is thick enough it should be fine. It is also significantly more expensive, and for that reason alone, I have not done much experimenting with it. Next, insert one of the "PVC Pipe Attachment" fittings into 1/2" PVC pipe at a length of your choice. The "Dual PVC Pipe Attachment" doesn't have one end blocked off, allowing for multiple generators to be placed on one length of PVC piping. On the beam, use super glue to adhere the piezoelectric crystal discs to the beam. I found that it is most efficient to place them near the ends and in the middle of the beam. This is where the most flutter will occur. If you opt to use ferroelectric materials (man-made piezoelectric crystals), then can gently be taped on the the beam. Keep in mind, these are significantly more expensive, but they do produce more electricity. My advice - purchase a lot of the small piezoelectric crystals and adhere them all over the beam, front and back, to produce a lot of electricity at a low cost. Let this dry, and then slide on one of the many shapes onto one end. The other end is placed into the "PVC Pipe Attachment" on the PVC pipe.
Step 4: Wire the Electronics
Using thin-gauge ribbon wire, solder one wire to the silver pad of the crystal (positive) and one to the gold pad (negative). If you are using ferroelectric materials, the connections are usually labeled. Leave enough length to each wire, to bring it down to the electronics control box. Ideally, each crystal should pass through a diode and then be wired together after passing through the diode, that way the voltages don't interfere with each other. However, that is a lot of diodes, so it is also acceptable to just wire them all together and skip the individual diodes. Note that some of the voltages will cancel out, and the final output voltage will be lower than if individual diodes are used. Then, once all the crystals are connected together, connect it to the full wave bridge rectifier, to prevent any voltage from going back any trying to power any of the other crystals (remember the converse piezoelectric effect?). It also converts the AC crystal power into DC electricity, making it compatible with rechargeable batteries. Add an optional buck boost converter to step the voltage down, but gain current which these crystals do not produce much of.
Step 5: Bring It All Together and Test It
One force that acts on aeroelastic flutter is mechanical vibrations, meaning a transducer, or a bass shaker, can be attached to the beam to generate the back and forth flutter motion. If the output is connected to an oscilloscope (not a multimeter - it won't be able to accurately detect the quick spikes of electricity the crystals produce), you will find that the piezoelectric crystals produce around 20 volts each. The piezoelectric crystal prototype produces 180 volts in total, when in the wind. If this was stepped down to 120 volts, which is what comes out of an outlet, there would also be a gain in current while stepping it down to a more usable voltage. The second prototype with three ferroelectric strips, produces 80 volts, and the last prototype with one larger strip produces around 5 volts. This is so low, as this type of ferroelectric crystal wants to be curled to produce electricity, not bent back and forth from side to side. When using this project outside to actually produce electricity, it can be placed anywhere, as it works in any wind environment. I find using an anemometer helps find the best spots for placement. In the future, it can even be placed in the side scoop of a car, to produce electricity for the car. This project is a totally new technology, and there is still experimenting to do, so please share any results or modifications in the comments below!