I found the failed bracket's first inspection and its corrected pass. Different support arrangements, same dimension heading. Our proposal subtracts one reading from the other and calls the difference weld movement. Can we use either result that way?
Not as a paired movement value without resolving the measurement conditions. Put the actual supports and drawing requirement beside each result. The corrected pass also follows an operation that the first reading did not.
Quality confirms the first check used the welding fixture as support. Final check used the inspection fixture after correction. Neither sheet records the first freely released condition on the inspection fixture.
Then that missing condition stays missing. You can retain both historical observations without subtracting them. A fresh identified set needs the incoming fit-up, intended location, first released check and any correction tied together.
Removed the movement figure from the proposal. Fabrication and quality agreed the new observation sheet, including the correction operator's entry. Historical pair retained with the reason it cannot support that calculation.