In the summer and fall of 2004 I built a parabolic dish solar collector. I was motivated partly by a desire to experiment with solar power and partly because the couple who owned the house I was staying in had gone for the summer on a bicycle trip across the United States. I wanted to have something bizarre waiting for them on their return. (This was something of a tradition in that house. Or maybe it was just me.)
The collector had a diameter of 2 meters and a focal length of one meter. The short focal length allowed me to get away with some imperfections in the mirror, but it also resulted in sunlight converging on the focal point from a wide range of angles, which would have made receiver design more difficult. It also made it difficult to use the collector as a solar cooker -- you want the food being cooked near the focal point, but if you miss and put it right on the focal point, it just catches fire.
The collector turned out to be good at setting things on fire but I never was able to do anything practical with it. I dismantled the collector before moving to Minnesota in 2006. It was a fun project and a decent first try at a collector. If I ever get back into solar power, I have some ideas for making the next prototype with less work and better results.
The collector is constructed from sixteen ribs anchored to a central base of plywood. Each rib is a piece of two-by-four with a parabolic arc cut out of its top side and an angled cut on the bottom so that it projects up from the base at the proper angle. The result is that the top surface of the ribs form the paraboloid shape that the mirrored surface must have. Making those parabolic cuts in the two-by-fours was the most time-consuming task of the whole project.
The skin of the collector is made of pieces of 1/8” thick tempered hardboard, arranged in three concentric circles. There is a single central octagonal piece, surrounded by a circle made of eight octants. The outermost circle is made of 16 separate pieces. The reason for making the skin out of many small pieces is that, the surface is curved, less deformation is required for smaller pieces.
Unfortunately the skin, as shown at right, isn't exactly a paraboloid shape. The ribs weren't all exactly the right shape, and the sections of hardboard didn't bend into the correct shapes. Even for the short focal length of this mirror, these errors in the surface are too large.
There is a way to create a very good parabolic surface, and that is to spin a body of liquid at constant speed. Consider a pan of water sitting on a turn-table. When the turn-table is first turned on, the water will swirl and slosh around for a little while. Eventually that will die down, and then the water simply undergoes solid-body rotation. That is, all the water has the same angular velocity, there aren't any currents within the water or relative to the pan it's sitting in. In this situation, the centripetal acceleration experienced by each bit of water increases linearly as we move out from the center of the pan. That acceleration means the slope formed by the top surface of the water also increases linearly. And that's the definition of a parabola. The shape of the pan doesn't matter (as long as it's deep enough that the water doesn't slop over the sides!), and the type of liquid doesn't matter. This is actually used in astronomy; one can make a very large and high-quality mirror by spinning liquid mercury for much less money than a glass mirror would cost, as mentioned here. The drawback of using mercury is that you can't tilt your mirror - the whole thing has to remain level, so you can only view things that are overhead, or nearly overhead.
There was also an Amateur Scientist article in Scientific American several years back about how to make your own telescope mirror by spinning a dish of epoxy on a turn-table, and that's exactly the approach I used. The collector structure makes a very good base for a spun mirror. Since the structure was already fairly close to a paraboloid I didn't need to pour a terribly thick layer of epoxy over it. The trick was spinning the collector -- obviously I couldn't just set it on a record player!
The pictures at right tell the tale of how I created the spun mirror. I constructed a frame to hold shaft upright and then mounted the collector on the shaft. The shaft was supported by ball bearings so that it, and thus the collector, could rotate freely. A few guywires from the top of the shaft were needed to help hold the collector level. The shaft was rotated by a gearmotor, and I built a PI-type speed controller to keep the rotation steady at just the right speed.
I sealed the gaps between the sections of hardboard with tape and also created a little dam around the outside of the collector with packing tape. With the collector prepared, I mounted it to the frame, got it spinning at the right speed, and poured in two gallons of self-levelling epoxy from U.S. Composites. I then let it spin for about 18 hours.
By the time I got to this point, Jeff and Andrea had returned from their bicycle trip across the country. They shot some video with their digital camera (I didn't have one at the time), some clips from which you can find at right. I think they illustrate well the ridiculousness of the whole operation. Jeff put it well: “Very strange. Could be awesome.”
The collector surface still wasn't perfect -- you can see many small-scale distortions in the reflections from the epoxy surface in the photo at right of the spun epoxy surface. These were partly due to distortion of the collector frame (it simply wasn't stiff enough) and partly due to areas of insufficient depth of epoxy. There were also clearly many high areas around the outside that the epoxy didn't get to at all. I could have tried pouring on more epoxy to correct some of these problems, but that would have gotten expensive and the collector was already as heavy as it could be without being unmanageable. I did not make any further improvements in the shape of the surface -- even with these imperfections there was enough surface area with the correct shape to work with.
The next task was to make the surface mirrored. I did this by gluing strips of aluminized mylar to the surface. The glue added its own lumpiness to the surface, but again enough level surface was left to have something to work with. With this done, I built a wheeled stand that let me control the tilt of the collector, and took it outside to try it out.
Though I never ended up doing anything useful with the collector, we did find it to be very effective at burning things. With it properly pointed at the sun, anything flammable placed at the focal point would begin burning within seconds. I used a piece of hardboard with a hole burned through it to help aim the collector: I could find the point at which the sun's rays were converging by moving the hardboard around until the edges of the hole were smoking more-or-less equally. At that point, the hardboard's shadow should fall on the collector with the hole exactly in the center of the collector, and I'd move the collector around until this was achieved. I once tried unsuccessfully to light a candle with the collector, but though the wick was scorched it wouldn't stay lit. It did melt lots of wax from the candle though.
The burning ability of the collector was most spectacularly demonstrated with leftover pieces of two-by-fours. The lumber started smoking immediately upon being placed near the focal point, and by the time the precise location was found (a few seconds later), the board would be in flames. We also melted some aluminum cans. We weren't able to heat steel enough to show any visible glow, so the maximum temperature achieved was probably around 650 °C. A well-designed receiver could probably have produced a reasonable amount of power (in the form of heat) at this temperature, which would have been plenty high to run a heat engine, though I never got anywhere in trying to design one.
The collector didn't work effectively as a solar cooker. The problem was that because the focal length was so short, the region of space in which sunlight was significantly concentrated was very small. I tried cooking a hotdog with the collector only once. With the hotdog only a few inches away from the focal point nothing happened. But when it was shifted over to the focal point momentarily, the surface charred almost instantaneously. At that point, we just held the the hotdog at the focal point to see what would happen. It was eventually reduced to nothing but ash and grease. Lots of grease. Yuck. We also tried popping popcorn once, with a little improvised dish into which we put some cooking oil and a couple dozen popcorn kernels. This was a little more successful, in that we did get several kernels to pop and so we were able to have a few bites of solar popcorn (oooh!).
The collector made a bit of a stir with the neighbors. The kids on one side seemed to think it was pretty impressive. Our neighbor on the other side warned us not to point it at his house. Apparently we scared the neighbor across the street a bit, he kind of hid the first time we wheeled the collector out to try it out. (I didn't see him there at the time, otherwise I would have tried to reassure him. The goal wasn't to terrify the neighbors!)
In the end, I dismantled the collector in the summer of 2006 before moving to Minnesota. Oh well. If I ever want to try this again, I know several ways to make the job easier the next time, as well as to produce a better result.