If you assume that the tension on the string is the same in front of the saddle and behind it, then the resultant force vector is along the bisector of the break angle. If the saddle is vertical then you can resolve that resultant on the saddle top into two vectors; one pushing downward on the axis of the saddle, and one tipping the top of the saddle toward the nut. That tipping vector has to be taken up by the wood in front of the saddle. If the saddle is back angled so that it bisects the break angle, then the resultant is all along the axis of the saddle; there is no tipping foce and no stress on the front of the bridge.
Rich Turner, who is pretty savvy, says that USTs work better under a back angled saddle, owing to the greater static down force. Like him, I use a 9 degree back angle.
I was taught to cut violin bridges so that the bridge centerline axis is on the break bisector. If you don't do this there is a tipping force on the bridge that eventually causes it to warp toward the nut, and this warp costs you some sound (note: not the angle, the warp). Because the violin bridge is itself tapered it often happens that the back face is perpendicular to the top, and that's the usual way the rule is stated.
On something like an archtop guitar it probably doesn't make much difference how you tip the saddle, just so the resultant is within the footprint of the bridge base. You'll know it if it's not!
There is a torque on the top of any fixed-bridge instrument, due to the saddle height off the top. There is some discussion about whether the break angle effects this, but within that I would not think that the saddle angle, whether it's along the bisector or not, would make any difference. I'm hoping to do some break angle testing some time soon, but don't have any plans to look at saddle angle in there: you've got to stop some place!
|
I know I have seen banjo bridges set this way.
)