Hi Art,
I have put together a few pages that explain the calculation of such a gnomon. http://riegler.home.cern.ch/riegler/sundial/tmpfiles/3Danalemma.pdf The Gnomon of a sundial can be shaped such that the Equation of Time is automatically corrected and the Sundial shows the Civil Time. Because the sun has the same declination twice a year, but the Equation of Time is different at these dates, the gnomon has to be changed twice a year, like it is e.g. the case for the dial of Martin Bernhardt: <http://www.praezisions-sonnenuhr.de/> http://www.praezisions-sonnenuhr.de/ In order to avoid changing the gnomon twice a year I developed a 'semi transparent' double gnomon some time ago that incorporates both gnomons so they don't have to be changed: http://riegler.home.cern.ch/riegler/sundial/mainpage.htm The absolute value of the equation of time at a given declination is not very different, so by approximating the analemma with a symmetric curve one can arrive at a single Gnomon that doesn't have to be changed, which is what you want to do. The error introduced by this approximation is at maximum 1.7 minutes, so it is quite good. As Roger points out, this gnomon is infinitely thin in the middle, so there one has to keep some minimum radius to keep it stable. The best way to avoid this problem is to really build the top and bottom piece separately and really have both pieces going to zero radius. Then you drill a hole along the axis and you insert a steel cable that you tention at the two extremities. Like this you don't have to worry about the stability of the thin piece. A steel cable of 2mm diameter is already extremely strong, so this will be a very good approximation to the 'zero' radius. Best regards Werner Riegler From: [email protected] [mailto:[email protected]] On Behalf Of Art Krenzel Sent: Friday, March 27, 2009 4:56 PM To: [email protected] Subject: 3D analemma I am seeking some assistance to construct a 3 dimensional analemma to be used on a gnomon of an equatorial sundial. Specifically I am seeking an equation which can be used to plot the analemmic error information in X and Y such that I might enter the data into a CNC machine to cut the curved surfaces. I want to average the time errors from each side of the figure 8 layout to make a symmetrical 3 dimensional object. Does anyone have a formula to plot analemmic timing errors where I can enter a value for X (not days) and get a timing error as Y for graphical purposes? Thank you for any efforts to help me solve my problem. Sincerely, Art Krenzel
--------------------------------------------------- https://lists.uni-koeln.de/mailman/listinfo/sundial
