novocaine
Old School Grand Master
But the clamping pressure is still radial and there is no material on the other side of the clamping bolt to take said force. This means the forces are not balanced and you end up with a torsional force at 120 and 240 degrees. Look at you original and where it snapped. You've added more material yes but the stress riser is still in the same place.I don't agree. There is abigger cylindrical area effectively working in the perpendicular plane. But ill refine and ill do a finite element analysis.
Move your bolt down to below the lowest point on the angle of the seat tube. It becomes a ring and the forces become balanced.
Your FEA will show two stressed areas as indicated above. These will work hardener whilst the remainder stays ductile and it will fail at which ever of those points reaches yield first. Fatigue is a bitch.
It's still an arse to machine as well.
But as I have no horse in this race I'll bow out now, I hope it works.