Thanks to expert consultation from VM inventor Jim-in-UK, I belatedly figured out that the 10mm-stroke magnetic induction VM recoil sensor I was initially using...
...was inadequate for light-weight, high power, long-recoil (>1cm) springers like my Webley Tomahawk producing ~2fpe or more per pound of gun. So on his recommendation, I cobbled together a 30mm long-stroke VM...
...that's giving far more sensible results that are consistent with previously published springer interior ballistics, as shown in the acceleration plot below.
PS: I calibrated the LSVM by removing the spring and leaving the back end of the tube open. Then, holding the tube vertically with the coil end down, I dropped the magnet down the tube and used its free-fall velocity when it passed through the coil (calculated to be 2.13m/s) to create a reference calibration signal: 305mV/m/s.
An interesting detail about springer internal ballistics is revealed in this acceleration plot. In Trigger to Target, the Cardews discovered that the violence of piston movement causes the back of the mainspring to momentarily lift off the rear guide, to then crash back into place after piston bounce.
I think that's what's causing the three recoil "bumps" in the plot below, providing a bit of experimental evidence that the Cardews got it exactly right in their (what some claim is) "dated" research.
