We started by taking some splines of bass wood to pin out the curves. Once we established the least amount of pins to hold the curves (3 for the parabolas and 4 for the s-curves) we bent them till we felt like they were going to break or they did break.
The most prevailing thing about the 2nd iteration was the difference in the curves in relationship to the different thicknesses. The top one is 1/32 in thick rectangular cross-section; the bottom is a 1/16 in thick rectangular cross-section. Of course the bottom one broke.
Now based on our knowledge of structures a rectangular cross-section has two different strength axes. The problem we are finding is that the breakage (most force) occurs on the weak axis of the section. For the 3rd iteration we changed to a square cross-section (1/8 in) to improve the weak axis. Although it may not be as flexible, the strength improved. Also, as you can see on the top curve of the 2nd iteration, the spline begins to rotate (because of the force which ends with breakage) about the weak axis; the square spline helps it stay rigid about that axis.
To continue from here we have three different kinds of wood (walnut/cherry/mahogany) that we will attempt to curve (both rectangle and square sections) to breaking point; this should show us the possible pros and cons for other woods (a different aesthetic compared to bass). Along with that we are going to practice digitizing some of the curves we have. – David