I'm reviewing gusseted steel support-bracket records for an FR5 proposal. Old notes say aligned while restrained, outside the drawing requirement after release, then corrected. I want fit-up, fixture location and final inspection separated before that correction becomes an unexplained instruction in the new cell. Paperwork review only, no live welding adjustments.
Find one identifiable bracket with its sequence intact. Fit-up record, restrained check, released measurement, any correction, final acceptance. Keep the drawing reference and measurement supports beside each stage; otherwise similar words can describe different checks.
Only a tick. The released sheet has a number against the final drawing requirement, but the restrained sheet names a fixture alignment check. I've been calling both dimensional inspection, which wasn't helping.
That clears up one comparison. It does not explain why the released bracket missed the requirement. Can you link that released reading to the correction record and the later accepted reading for the same bracket?
Yes, one bracket is traceable through all three. Correction record says straightened at bench, but doesn't identify the method. The final reading is after that operation, not evidence of an acceptable untouched weldment.
Ask the process owner to explain that bench operation and whether any original method record survives, while quality checks the supports and references used for the two released readings.
It was in the proposal summary. I've corrected the summary to say final acceptance followed bench straightening. Quality confirmed matching released-check supports; the missing straightening method is now with fabrication, not filled in from memory.
Keep that one complete chain separate from any wider correction-rate claim. It establishes what happened to this bracket. A pile of unlinked final passes will not establish how often the earlier failure occurred.