I couple of weeks ago, I received an e-mail from Starsailor asking what my thoughts were on solar trackers. As it turns out, I had recently gotten my hands on some solar cells and was planning on putting them together and seeing if I could track the sun myself to help increase their output.
I won't go into the details of the various emails that we sent back and forth, but I will show some of the experiments I did in Sketchup to test out the polar axis sun tracker that I had wanted to build.
If you don't know what a polar axis sun tracker is, go to this Wikipedia link and scroll down. http://en.wikipedia.org/wiki/Solar_tracker
I'll start by showing a picture of a simple sun tracker that I drew in Sketchup. Notice how there is a rectangular block that sticks out of the front of it.
With Sketchup, it is possible to accurately simulate the shadows cast by the sun during any time of the year or any location on Earth. This means that it is pretty easy to tell if the sun tracker is pointing at the sun or not by looking at the shadows cast by the rectangular block.
To open the shadow window, go to the menu bar, click window, and go to shadows.
To set your latitude and longitude, go to the menu bar, click window, and go to model info.
Notice that there isn't any shadow in this picture. This means that the sun tracker is pointing directly at the sun.
In this picture though, you can see a shadow, which means that the sun tracker isn't pointing at the sun.
I first simulated the sun tracker's behavior during the winter solstice, the shortest day of the year. In order to do so, I needed to make the angle of the sun tracker the same as the sun's declination at this time of the year. Here is a table of the sun's declination values. http://www.wsanford.com/~wsanford/exo/sundials/DEC_Sun.html
Here is the sun tracker lying flat.
The latitude where I live is about 39 degrees, so I first rotate the sun tracker forward 39 degrees to compensate.
The sun's declination during the northern hemisphere's winter solstice is about -23.5 degrees, so I will next rotate the sun tracker forward 23.5 degrees more.
The basic idea for this sun tracker is for it to rotate 360 degrees every day along the polar axis, or 15 degrees per hour. This means that it will end up in the same spot at noon every day. This picture shows the axis that it turns on from above.
In this picture, you can see that it is facing directly at the sun at noon on the winter solstice, which is good.
I moved the sun forward 3 hours and rotated the sun tracker 45 degrees. You can see how a polar axis sun tracker is doing a pretty good job of tracking the sun on this day. There is a little bit of a shadow up at the top, but, for the most part, it is pointing right at the sun.
I next tried the experiment out during the summer solstice by first changing the declination angle to +23.5 degrees.
At noon, there wasn't any significant shadow, which meant that the sun tracker was pointing directly at the sun.
I then moved the sun forward 6 hours and turned the sun tracker 90 degrees. Here you can see that there is a fairly long shadow.
The fact that the shadow was so much longer confused me for awhile. I had thought that the polar axis sun tracker more accurate than that.
I decided to try another approach. I next drew lines from the sun to the sun tracker using the Sketchup Sun Position plugin. http://cerebralmeltdown.com/projects/sunplugin/default.htm
This made it a little bit more obvious why the sun tracker wasn't pointing at the sun. It might be kind of hard to tell from the picture, but the sun's azimuth has increased so much that it is basically behind the sun tracker.
Here is another angle. The line from the sun to the sun tracker is highlighted in blue. The line also looks like it agrees perfectly with the shadow.
It looks like a polar axis sun tracker is only so accurate. I still think that it will work well with solar panels, so I'm still planning on using, especially because it is so simple. It doesn't look like it would work very well for everything though. A solar concentrating parabolic dish, for example, needs to be more accurate than that for it work correctly.
Once I actually get the chance to build this thing, I will have to put up some pictures. It could be awhile though before I have the time for it, so, for now, I'll just have to play around with this cad model to keep myself occupied.