What is wrong with you people talking about curves not being straight. Everybody knows that a curve is a slightly bent straight line. It can't get simpler than that.
Actually a line is just a straight curve. According to my modelling software.
I have been wanting to build an icosahedron for along time. It is a regular solid with 20 sides, each an equilateral triangle. The SMT makes it easy to whip out triangles, just set the fence to the 30 degree mark. But I never could crack the bevel angle. Today I was looking for something else and found a spreadsheet I did a few years back that found all the face angles for a square-based pyramid (another possible project) and it reminded me where to look for the answer.
Specifically:
delta = arccos(-sqrt[5]/3) = 138.1896851° (roughly ) gives you the dihedral face angle.
So the table saw bevel angle would come out to (180 - 138.1896851)/2 = 20.90515745°.
No one has a protractor that precise, so the usual method is to make a template by constructing a right triangle. With one angle set to 20.90515745°, the ratio of opposite to adjacent sides is tan(20.90515745° = 0.382.
So you draw a 1000mm line, and a perpendicular 382mm line, then draw the hypotenuse of that triangle on a piece of mat board or something similar that's suitable for templates. You cut that out (although you really just need the end opposite the 20.90515745° angle) and use it as a gauge to set your bevel angle. (I may just use my sliding T-bevel) No protractor required, yet it's more than precise enough.
Anyhow, I have a pile of plywood scrap that's earmarked to finding out whether the above is enough to get me to a reasonable icosahedron.
Comment