Savonius I
We set about to build a Vertical Axis Wind Turbine to power our wireless access point in the fall of aught-3. The WAP is mostly a repeater so it would be appropriate for a remote, stand-alone fielding. Our current WAP is too power-hungry for this turbine so we are going to put another WAP on this in the spring for field testing. Richard bought the plans for the turbine from picoturbine.com and modified them a bit.
Why Savonius?
This design is good in low wind. The tradeoff we considered (and eventually took) was that we wanted to harvest some energy out of the numerous low-wind opportunities than harvest considerable energy out of the rare high-wind events in Boston.
Here it is. It is about 4 foot high with a 12 inch diameter rotor. Using parts mostly bought from an assortment of vendors, we fashioned a frame out of Dexion and put venetian blind slats in for the blades. This isn't the final model, but it is good enough for bench testing.
Here is the Vcc output trace (no load) for our turbine spinning at 200 rpm. Another great thing about Savonius in general, and picoturbine in particular, is that you get 3-phase AC out. Might be useful down the road...
Here are the coils twisted to the specs we had. An improvement we will try here is smaller gauge, more loops equals higher voltage.
Here is the base where you can see the coils and rotor plate. We staggered the blades to get this shot. What you can't see is the magnets on the underside of the plate.
We got north of 6 volts at 200 rpm. Don't know what the simulated wind speed was for this test but that was from a point source (compressor nozzle) so it wasn't very realistic. Our next test will hopefully be in a wind tunnel and we will have a higher degree of confidence in those numbers.
The future of this work.
We know what we need to do on Savonius II, the turbine we plan on field testing. While these initial numbers are fairly anemic, we have not done any performance optimizations yet. We have discovered quite a few places we can make adjustments, but many of those come at the expense of robustness.
the only components we ordered from pico were the magnets ... (new) Cu wire for winding the first series of coils for the first 3-phase alternator we built came from way-too-highly-priced u-do-it electronics, out on rt 128 ... (new) hardware came from local hardware stores and bath industrial sales (maine), and the (real, authentic, very expensive british dexion 1.5 x 1.5) perf'd angle iron came from our own collection of such extensive supplies on-hand here at cannery iron works (maine) ... as well as the steel from which we fabricated the disks for the magnet-rotor assemblies ...
ditto the upper (bronze bushing) bearing support and mount for the shaft top ... all the elements for the rotor vane support system and the `vanes' themselves were from additional materials on-hand - ie - nothing but the magnets came from pico ...
CAVEAT - the pico design templates AS PRINTED from the pico site were UNDERSIZED ... so surprise! due to our haste at that point in the pre-build stage we discovered the magnets WOULD NOT FIT THE ROTOR PROPERLY - and the array would not be correct - we'd already roughed the rotor disks from 1/8th inch mild steel, and they were hot off the lathe when this was realised ... even though there were TWO OF US studying the plans!
also, though we put out plenty of volts (5-6) - there was nearly NO POWER (in light winds) ...
CAVEAT - why was it so important for the magnets to fit the rotor disk at least precisely? if not in such a way that we could also turn a bit of a ridge 'round the periphery of the rotor (to aid in restricting centrifugal forces of the magnets when spinning at high rates)? well, epoxide is /ok/ ... but ... not!
------- more to come -------------