I'm not blaming the proposed FR5 loading yet. These are hand-loaded aluminium blanks, and the sheet only records how far each sticks out. No blank length alongside it. We kept the samples, thankfully.
A recent tray comparison taught me how quickly a few good examples become a claim about the whole setup. Here I want to compare the same identified blanks properly before anybody calls the stop unreliable. What should the setter and quality put beside those projection readings?
Six are linked by their tags. Two loose ones are not, so I've left them out of the old comparison. The sheet says front face for the reference, no drawing view.
Have the setter identify that face and the seating stop on a sketch. Then measure the retained lengths and compare the repeat seating of each blank. The loose two can be new samples, not magically recovered history.
We found mixed cut lengths in a box labelled one size. Everybody had stared at the projection because that was the column on the sheet. A second column was cheaper than the stop we were about to order.
Mostly length variation. One blank had a rough cut end as well, so we still had a seating problem to sort out. That's why I wouldn't subtract two numbers and declare the whole batch explained.
Setter has identified the front reference and stop. The six retained lengths form two groups, matching two saw batches. One blank also has a raised edge on its end; quality is checking that against the blank preparation requirement.
That raised edge belongs with its own sample identity. Don't correct it and then use the changed blank to explain the old reading without saying what changed.
The raised edge was recorded before the blank was corrected to the preparation requirement. The new check sheet now carries blank identity, length, end check and the agreed projection reference. Repeats are under way with both length groups.
Amara, the repeated readings for each checked blank are consistent with the agreed method. Sofia, the longer group projects further by the measured length difference. The corrected blank now repeats too, but its earlier condition stays in the notes.
Yes. Stores and the setter agreed the batch labels and the preparation check before issue. This gives the loading study a usable starting point; it does not authorise automatic loading.