The Miller Rotor was invented by a retired Ballarat farmer, Mr. Miller, in, I think, the early nineteen sixties. He had noted the amount of force that could be generated by a sheet of roofing iron in even a moderate breeze and wondered why it should be the case. His interest was piqued further when he read a book about the famous windjammer, the ‘Cutty Sark’, and the amount of wind power its sails produced …. a theoretical maximum of some sixty thousand horsepower in a thirty knot wind. (If the rigging could hold together!) He worked out that to be useful, the wind would have to impact the vane at an angle for best torque translation in a rotating system. His result was the Miller Rotor.
He experimented with vane dimensions and shapes, and, being a well educated man he decided that the Greeks had probably got it right when he settled on the ‘Golden Mean’ or ‘Golden Ratio of 1 : 1.618 for the vane dimensions. Further experiments showed that this was not a hard and fast rule and that wide variations in the ratio made little difference until you got wildly outside the Golden Ratio norm. A ratio of 2.5 : 1 still worked, but poorly across all conditions of wind speed.
A practical design uses corrugated iron sheeting, with the corrugations running vertically. Each vane is made up of two sheets, each 2.4m long by the standard 756mm per sheet ‘coverage’, to give a vane that is 1500 x 2400 mm, roughly. This size of machine should only be built by a competent mechanic. It produces lethal amounts of power, and if it flies apart at full revs it will severely damage the surrounds and potentially kill any person within range of the flying debris. Each vane has to be properly braced to the cross bar between the vanes and solidly attached at the vane centre. You should use, for each vane, three horizontal ‘top hat’-type bracing pieces across the top, bottom and centre. The corrugations of the roofing iron will brace the vane in the vertical.
The length of the cross bar joining the two vanes determines start up wind speed. The closer the cross bar length is the lesser vane dimension, the lower the wind speed needed to start, but the lower the torque and the lower the power. Conversely, as the cross bar length approaches the major vane dimension, the faster the required ‘start up’ wind speed, and the better the torque and power. Take your pick! A good compromise is 71% of the major dimension. Miller rotors are relatively lively and will start at wind speeds that will not even move a normal windmill, so the longer cross bar is preferred.
DO NOT SET UP THE COMPLETED ROTOR ON A SHAFT SO THAT IT CAN ROTATE WITHOUT HAVING IT CHAINED DOWN!!
I had someone try this and started to run with it held in his hands to see if it would rotate. It did, of course, and the natural gyroscopic forces asserted themselves and tipped him into the dirt. He only needed thirteen stitches where the corner of the vane hit his head !
EVERY MILLER ROTOR NEEDS A ROBUST BRAKING SYSTEM AND A TIE-DOWN CHAIN TO PREVENT MOVEMENT DURING MAINTENANCE OR ASSEMBLY.
The ‘leading edge’ of each vane is bent out at a roughly forty five degree angle to allow the rotor to ‘self start’. Miller rotors will work without this, but will have to be spun up to operating speed by some other means, and, if the wind speed momentarily drops right down, then picks up again, the unbent rotor will simply coast to a stop. Easier just to bend the edges out.
You will need some form of gearing in order to be able to use an alternator. You have to get from 20 RPM up to around 2500 RPM. This is fraught with many difficulties. For instance … If you use vee-belts and multiple shafts and multiple pulleys on each shaft, you will lose 0.6 of a horsepower everywhere a vee-belt touches a pulley. So if you have a triple pulley and two shafts, you lose 6 x 0.6 = 3.6HP. You will need much, much more than this, however. There are rules about how much bigger the big pulley can be compared to the small pulley. You need to get a ‘gearing ratio’ of 125 : 1 (2500/20= 125) but you are limited, in practise, to steps of around five or seven to one, so 125/5 = 25, 125/7 = 18. This is far too many shafts and pulleys to be sensible. A better option is to use an old Toyota Coaster bus differential that has a ratio of 9.6 : 1. That will bring you up 20 x 9.6 = 192 RPM…much more practical! Now you only need a step up of 13 times. One ‘lay’ shaft with a large pulley on the diff’s tail shaft driving it and a second large pulley on the lay shaft will drive a car or truck alternator to around twenty five hundred RPM. The full output of a good truck alternator is about four Horsepower and you have another thirty to forty Horsepower spare, so maybe ten or twelve alternators? This is getting silly again. There’s a much better way!
Assuming an output of about forty five horsepower and a rotational speed of twenty RPM, the best, most mechanically efficient way, is to drive several reciprocating pumps from an eccentric on the vertical shaft directly connected to the rotor. Each pump consists of a radial ply tire, with four circles of marine, 16mm plywood clamped in pairs, either side of each tire bead and made air and water tight with silicone sealant. Two simple, plastic foot-valves (One-way Valves) are solidly glued into holes in the ‘back’ plywood face. One allows water into the interior of the ‘tire-pump’, the other allows water to get out. The ‘front’ plywood face has a push rod coupled to its centre point. Each ‘tire pump’ is solidly bolted, in an upright position, to an upright post, or posts, set vertically into the ground in a wide circle around the central rotor drive shaft, with each pump’s push rod terminated on an eccentric bearing on the drive shaft.
You then have a circle of five, seven, nine or eleven ‘tire pumps’ (always an odd number or a ‘prime’ number) around the drive shaft whose eccentric drive alternately pushes and pulls on each of the ‘tire pump’ push rods as the drive shaft rotates, thereby sucking in water from a smallish holding tank (~ 1000 Litres) and then pumping it out to a ‘plenum’ or manifold connected to a water pipe that conducts the water to a water turbine somewhere up hill, driving a large alternator for power generation. The waste water is returned by a second pipe to the Holding Tank at the windmill by gravity feed. The waste water might also deviate via a garden water feature before continuing on its journey back to the windmill holding tank.
Why an odd or prime number of pumps? Because even numbers of pumps can allow the set up of sympathetic, additive, harmonic vibrations that reinforce each other and can destroy machinery and pumps. Odd harmonics rarely reinforce and quietly fade away before they become a problem. In fact, odd harmonics more often than not subtract from each other, whereas even harmonics add up. This last sentence is a gross over-simplification. Do a formal Fourier Analysis to get an accurate picture but, generally speaking, ‘Prime’ numbers of pumps will work best (ie. 5, 7, 11, 13, 17, etc). This is why, quite counter-intuitively, 5 cylinder engines are smoother than four cylinder ones and also give 25% more power for only a 20% increase in complexity.
A few words about windmill towers. Don’t be tempted to get one of the myriad lattice-style windmill towers variously available in country areas unless your miller rotor is a very small one. A Miller Rotor the size of the one previously described, crumpled a 10 metre Southern Cross windmill tower the first night it got hit by a ‘Gully Wind’. These winds are common in hilly areas and are the result of local, adiabatic air movement, reaching 36+ metres per second (130KPH) or more. For the above Miller Rotor, you will need a tower constructed of three or four large (300 x 150) ‘I-beams’ in a teepee shape or three or four 150 – 200mm steel pipe of 8 – 9mm wall thickness buried one third of their length in the ground, in concrete. Sounds over done, doesn’t? But, that Miller rotor will generate side forces of up to 80 tonne in 130 KPH wind gusts, so it needs to be substantial.
Miller Rotors are amazing, and they can get expensive very quickly. However, they are very simple to build and relatively cheap, except for their towers. They move slowly and seem to be quite visible to birds on the wing, unlike a lot of other windmills, but they are subject to stupendous forces, particularly in a hilly environment. Proceed with caution!! I take no responsibility for what your efforts do to you and yours. Be it on your head!
I have left out a critical piece of the mechanism, but if the thing is signed off by an Engineer and constructed by a competent mechanic or steel fabricator, then the missing element will become obvious. Miller Rotors are thought provoking and interesting but very dangerous in inexperienced hands. The power of the wind, coupled with the unforgiving nature of mechanisms, makes for a deadly brew if treated lightly. How hard can it be? Forget that question. The one you should ask is …”How deadly is it?” If you cannot see that for yourself, you have no business trying to build one. Small models of balsa wood and cardboard are safe to experiment with, but as soon as you start using sheet metal, beware!
