Reversal Methods - Absolute Measurement

One of the simplest places to see reversing is in calibrating a level.

if I = indication - I_a first measure I_b second measure with level reversed.
and e_l = error in level
and e_t = error in table..

Then
I_a = e_t + e_l
I_b = e_t - e_l

The rest is just simple math..

Dan Gelbart briefly mentions it..

https://youtu.be/cwdoUjynpEk?t=1355

A related process happens in the 3-plate method to get a perfectly flat plate..

I found this..

https://zero.sci-hub.se/399/ac570088aaf6156507eabea4d19740b7/evans1996.pdf#view=FitH

Generational delinquency - degredation of straightness, flatness and measure

There is a problem in machining - the error in the tool transfers to the error in the work - so if you use a tool to create another tool - things just get worse - reversal methods are the cure..

Using 2 plates instead of 3

This trick - lets one use 2- plates... But I would make it nicer.. <grin>

https://www.youtube.com/watch?v=Z0jCycLpXSM