Each aluminum block has a pair of drilled passages to direct pressure oiling from the pivot ball to each end of the roller wheel axle. Metering of this oil is copied from rocker-pivot oiling scheme of the original '55 Pontiac V8 engine- a drilled passage through a cam journal, thus a "spurt" every revolution, plumbed to all eight pivot balls. The test spinning showed the oiling system functioning as planned.
Regarding acceleration forces in the valvetrain- yes, of course the mass of all moving parts was part of the calculations. The 10,000 RPM goal is not a "blue sky" number. It's based on years of experience with my pushrod hemi V8, with its massive (literally) valvetrain. It runs reliably to 9,000 RPM in competition. The four cylinder being developed uses one of the same hemi heads, same valve sizes, etc., but with much less reciprocating mass in its OHC valvetrain- so the 10,000 RPM capability is not at all an unreasonable expectation.
Again, thanks for the suggestions about redesign, alternative materials, etc. However, if it weren't for the one neglected stress riser in that one follower (oops!), I'd still be trying to make the valvetrain fail.
No FEA. Purely old-school approach. [Although a pocket scientific calculator beats the heck out of the old slide rule...]