This one is an example of a fracture of one of the pivot bolts. This problem was less widely reported than the weld failures, mainly because they were usually noticed before both bolts broke, so the bike (like this one) did not break completely into two parts.
I think the most likely reason these bolts broke is that the upper and lower bolts were not completely coaxial, and the transition from threaded to unthreaded portions acted as a stress riser. The revised joint has a single pivot rod with no diameter changes.


This looks like a weld defect, but probably not one that is leading to failure, for two reasons. First, most of the published images of the frames that broke showed evidence of corrosion at the lowest part of the joint. This suggests that the cracks started there, where the tensile stress is greatest, and propagated upward until the remaining weld could not support the load. There was one image on the Tern Forum of a small fracture line at the bottom of the joint, which is what would be expected; unfortunately, this was probably the last place anyone would look.
Secondly, there is paint *over* the defect, suggesting it was applied over an existing crack and also hasn't progressed since. I wonder why it wasn't picked up during inspection at the factory though.
Steve