The Mirror AssemblyMirror selection and mounting are very important for heliostat performance. From a mechanical perspective the mirrors and their mounting frame must be strong enough to withstand wind loads and yet remain light enough be within the load capacity of the supporting frame and mount. From an optical perspective, high reflectivity and the ability to concentrate the light and desirable. From a practical perspective, readily available, and cheap materials are important.
My first prototype used a wooden frame and 12" mirror tiles for practical availability and price, I decided it would be 4 foot square to minimize lumber waste. I soon realized that something would have to be done to account for the space between the 12" mirrors. The solution was to trim the mirror tiles on the top and bottom rows.
Due to the sun's optical width of approximately 0.5 degrees the sun's reflection normally is magnified as distance increases, which decreases the energy concentration. A full parabolic mirror is an ideal solution, but this is harder to make. The compromise solution is to have a multi-faceted mirror with each facet angled slightly to simulate a parabola. At 50' a flat 12" mirror will reflect a solar image that is 12" + sin(0.5deg)*50' =~ 17.24" wide, or about 22.5" at 100'. So an idealy adjusted array of 12" mirrors would concentrate the light down to about an 18" circle at 50' or 24" at 100'. With an array of about 15 sqft that translates to solar concentration factors of approximately 8.5 and 4.8 respectively
In order to create flexible mounts for the mirrors I used springs on screws threaded into T-nuts in the 3/8 plywood. Each screw has a bracket made from two pieces of aluminum extrusion. The bracket holds the mirror edge and is held at the desired height by the spring and adjusting screw. I got lucky and found all these parts at a surplus store so they just cost pennies each. All except for the brackets which I must say were fairly problematic. First I glued to pieces of 1/4 aluminum together, one L and one U channel, using JB-Weld epoxy. Then I marked it in one inch lengths and cut it. The two pieces of metal broke apart from every cut and I was forced to sort them all and match them back up together again so that they fit perfectly. Each piece was drilled and filed down by hand.
During a thunderstorm with 1" hail the advantages of this approach became apparent when none of the mirror tiles suffered damage. The smaller tiles seemed to be stronger than a single large piece and the spring supports seemed to absorb some of the hail impact and prevent damage.