Dr. BAJ said: >> I (Peter) said: >> >> The problem with the parabolic shape proposed is, you cannot make an >> array from them (the focal point is way too close). So it turns out >> that just taking plain glass mirrors (at $6 / sq. meter if what I >> read was true) and warping them slightly in a frame (they will warp >> anyway) will yield something closer to the 10 meter focal distance >> needed for an array. > >Why 10m?! I actually wanted a much closer focal point. That simplifies >tracking everything. Coordinating a mirror array and then a target >that's 30 ft away is a daunting task. An ideal focal box would be 2-3m >from the center of the array. If you use the fixed mirror on a pole and target below it scheme then you need that kind of focal distance. Because of the distance from the pole required, the moving mirrors will need to have a large focal distance (10 meters or more). This is easily done by taking a plain glass flat mirror and simply attaching a suction cup on the back and pulling it out with one or a few kgf. The equivalent glued (epoxied) assembly would be your individual moving mirror. The mirror will not be parabolic but it will likely be much better than mylar experiments (my limited experience with mylar is not goot. It has to be fixed on something very rigid which tends to be heavy). >> Then the secondary mirror can be on a pole, and IT can be hyperbolic >> (using a parabolic dish as casting form and mirroring the 'wrong' >> side), and the target on the ground and fixed, where you can get at >> it. > >Again I was planning on tracking the target too, moving the whole >assembly from east to west during the day. I agree with James that >tracking individual mirror elements is way more complex than having a >fixed mirror array that tracks as a single unit. In your opinion, what is easier to build, 10 or 100 or 1000 individual identical mirrors weighing less than 50kg each freestanding or attached to a simple pole driven into the ground, which can be individually made, one at a time (one per weekend ?), or a third of a ton of parabolic mirror with attached assorted hardware such as a generator, a condenser, wiring and whatnot, which needs to be pointed to 2 degrees precision in 2 dof in despite of wind and structure bending, and requires a small construction crew and good weather to put up and get down. >The good news is that arrays scale. > >But the complexity scales too because the mirrors must track >individually right? Or can each mirror be fixed in such a way that you >can track an entire linear row with a single motor? Or am I missing >something? You can track the entire row with 1 motor but you probably do not want to. A small ~2W servo per mirror would do the job fine imho. Maybe even a usual model servo with the pot removed and attached to a sector gear made @home. A model servo produces about 42 oz*in. If geared down to 1:100 with a simple worm gear (threaded rod + rough cut gear in mild steel sheet) it would make 3 kg*m which is likely enough to point an aerodynamically balanced 1 meter square mirror in calm weather (assuming a nice bearing such as hollow-axle-and-steel-ball). The sector gear would be about 200 mm radius half circle, easily made even by hand. Of course it would have to be able to take 300kgf at the rim (and so would the worm gear and its bearings) but this is not impossible. >> You can start with 2 mirrors and increase to >> 50 later. > >That's why I wanted to go with the mylar. You build the whole assembly >once as a single unit. And limit its power for good. >>The bad news is that a moving linear array is subject to cos >> phi/2 losses and that the mirrors shade each other beyond a >>relatively >> small sun angle (say 45 degrees). But the time of useful sun is only >> about 6-8 hours per day and at that angle cos phi/2 loss is <15%. > >Complexity is a killer here. A backyard setup has to be simple to be >effective, even if that simplicity causes some losses. >Tracking a single big mirror along with a target that tracks with the >mirror is much less complex than tracking a wide array of individual >mirrors. That's why Curnutt style furnances are a fixed array of >mirrors with the whole assembly tracking the sun. The current Sunflower >technology uses some type of gimbal arrangement where all 25 mirrors of >the array are individually tracked using only a couple of motors. But >it's complex. And complex leads to breakdowns. Talking of breakdowns, what happens if you want to modify something on a big rig ? Call in construction crews ? Crane truck ? The small-moving-mirror system seems to be favored by all the systems that are in current use or experimentation. Even if several mirrors are removed or faulty the array continues to work. Peter -- http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist