I am mostly into the theoretical side of these wonderful valve gears, and am working on designing one for the sake of a 3D model of a steam locomotive I am working on.

I have been using D.L. Ashton’s “Design Procedures for Walschaerts and Stephenson’s Valve Gears” as a guide, but there a few parts which I still find unclear.

Assuming an inside-admission WVG, what determines where engineers put the trunnion of the expansion link? I have seen examples of valve gears where the trunnion is higher than the valve spindle, higher than the top of the radius rod even, or lower that the valve spindle. I’ve also seen examples with long a long radius rod and short eccentric rod, and the reverse.

Is all of this fairly arbitrary, and determined by constraints elsewhere in the engineering? Or does certain placement reduce the inherent error that crops up in various places (converting between angular/linear motion, etc) for certain purposes. Does one trunnion placement favor movement in forward gear, while another is less optimal in forward, but still equally decent in reverse?