The first step was to model the tracker in Solidworks.
We knew from a study of historical weather data that the tracker did not need to swing from east to west more than 60° each way - the sun is not high enough in the sky or powerful enough to meet the requirements of the testing standards beyond ±60° azimuth angle. Given this, it was possible to use a simple design that is similar to a swinging gate. At the base of the gate is an actuator that pushes and pulls enough for ±70° of azimuthal motion. The motion about the horizontal axis is from facing straight up to about 70° rotated towards the horizon. The tracker needed to support substantial mass as the exact size and weight of collectors varies a lot, we would also test unknown new collector designs. One peculiar requirement of the testing requires the use of large fans to provide the necessary 2m/s to 4 m/s wind speed. We used large air curtains aimed sidewise to blow across the collector face. These fans and the structure to support them added considerable weight. Constraints of limited funding and inexperienced labor, required material and design choices allowing assembly with simple, inexpensive and available tools. The desgn addressed a number of constraints: limited funding requred spending as little as possible, the need to use inexperienced and unskilled labor, and the need to use available materials and tools allowing construction and assembly with simple, inexpensive and available tools. We had access to cutoff and band saws, drills, a MIG welder, wrenches and so on. I designed the structure from 4 and 6 inch square tubing, and several sizes of angle iron. Some simple laser cut and brake bent parts made easy transitions between the structural tubes. These parts were simple to clamp and weld accurately in place. Having worked with sheet metal in the past I knew this would be an economical fabrication method. One surprising thing, obvious in retrospect, was how willowy the large platform made from angle stock turned out to be. Flat structures of this sort want to droop into a saddle (hyperboloid) shape. This drooping happens because the angles are able to twist along their long axis - they have little torsional strength. Adding a 3 inch tube along both sides was a light weight solution that effectively solved the problem. I chose polymeric bearings from IGUS Inc. instead of ball or roller bearings as this application requires only very slow, limited motions, and it is out in the weather. The vertical (azimuthal) axis used large spherical flange bearings, and thrust bearings to take the weight. The horizontal (zenithal) axis used smaller spherical flange bearings with shaft collars to keep all the components aligned properly. Each axis has an absolute encoder installed to report the direction of the tracker platform. We chose a magnetic rather than optical encoder as these are not degraded by condensation or other environmental inputs. Welding up the platform. The tracker was welded up on the ground then disassembled and moved to the lab roof. |