For those who are interested in the details of making quality parts, I HIGHLY recomment Carrol Smith's books. "Engineer to Win" specifically addresses the metalurgy aspects of this axle problem.
www.CarrollSmith.com -- The Official Internet Site for Carroll Smith Books
There are some excellent examples - photos, etc of both catastrophic and fatigue failure of half-shafts. These **were** top-notch halfshafts used in high dollar race cars. They don't screw around with this stuff, yet failures happen anyway.
I guess I should qualify my "joining two different axles" statement. Yes, it's frequently done by OE's, who seem to have all manner of technology available. I don't.
In our case, taking two OE shafts from different vehicles would require:
a) tempering the shafts from the *** unknown *** heat treat state
b) making a possibly dual inner diameter sleave to join them - don't screw around here and start with something like a welded steel tube!!! You must exactly know the stress you will be dealing with and start with material that doesn't have a built-in stress riser like a welded joint.
c) welding the sleaves to the tempered axle stubs
quadruple checking your weld quality - have a certified welder do the job in the first place, then use various means to determine if there are voids, inclusions, etc. An inclusion or void due to incomplete weld penatration are perfect stress risers, and are the place where a fatigue failure is most likely to start.
d) retemper the weld joints
e) heat treat the assembly back to the point where the OE splines and the weld won't be the failure point.
f) have a driveshaft shop balance the new axle
I've welded a few moderately stressed parts, the motor mounts on my truck's engine swap for example. However, they were specifically designed NOT to be highly stressed critical parts like halfshafts are.
- E*clipse