That would cut a piece out from the center of the dowel that looked like some kind of weird drill bit or screw, yet the rest of the cylinder was still intact and the two pieces could be reassembled. With the right combination of pitch and center-line offset, the two paths can be the same path - the blade would enter and exit the wood along the same helix line drawn on the surface. I reasoned that the path of the blade along the surface of the cylinder where the blade entered should be parallel to the path created along the helix where the blade exits the cylinder. It was difficult to maintain a constant pitch without anything to go by, but the wider kerf of the spiral blade provided enough clearance between the two pieces that they would usually come apart easier. If I maintained a constant pitch, the two pieces could be separated like a nut and bolt. With the table tilted at roughly 45° to the left and pushing the dowel downhill along the fence I could see that the blade was going to make first contact with the dowel on the top corner and that I would have to twist and push the dowel a distance roughly equal to the dowel diameter before the blade would be cutting through the entire diameter. That got my curiosity going and after experimenting a while I wondered what would happen if I twisted a dowel through the blade? I rigged up a fence to keep the center of the dowel a constant distance away from the blade.

With the saw table tilted and changing the feed direction of the wood, I found that a random shape could be cut out of a board and still be trapped within the wood. In 2003 I was playing around with my scroll saw with a spiral blade.

Helix shape