CoatingIQ
← Course index

Wrap-up

100%

Lesson 13·Running forge dies

Records and traceability: what to log, what it tells you over time

The minimum die log per hit-block, what a year of those logs reveals that a single teardown cannot, and the standards (AIAG CQI-9, IATF 16949) that already require most of it.

8 min readLesson 13 of 13

Tying it together

What this means on the shop floor

The logbook is not extra work. It is the only instrument that sees the slow variables. The fast variables (preheat temperature, lube dose, cycle time) have their own instruments and their own readouts on every shift. The slow variables (lube batch quality, operator habit, ambient conditions, press alignment drift) have no other instrument. If the log does not capture them, nothing does, and the shop pays for the gap in dies that died early without explanation.

The shop that gets value out of logs is not the shop with the most fields. It is the shop with the discipline to fill the seven minimum fields, on every die, on every hit-block, on every shift, for a full year before drawing conclusions. Three months is not enough. Three months captures one season, two or three operators, and a handful of lube barrels. A year captures the full lube supply cycle, every operator the shop employs, all four seasons, and enough press-service intervals to see the slow drift.

Internal pushback questions

  1. On the most expensive die in current production, where is the log, who filled it out yesterday, and what are the seven fields recorded for the last hit-block? If the answer involves opening a cabinet and digging, the log is not a working instrument.
  2. What lube barrel batch number is in the spray system right now, when was it opened, and which dies have run against it? A shop that cannot answer in under two minutes cannot detect a bad batch when one ships.
  3. What does the die-life average on this press, on this job, look like over the last twelve months by month? If the answer is "about the same," the data is not in a form that would show a drift. If the answer is a graph, the shop is running the press as an instrument.
  4. When was the last time the night-shift and day-shift pull criteria were compared, on paper, for the same die on the same job? If the answer is "we trust the operators," the variance between shifts is invisible and uncorrectable.

Common confusions

A logbook is not a quality record. A quality record is the certificate that ships with the part. A logbook is an engineering record that ties the production history of the asset (the die) to the variables that ran against it. The two records overlap on traceability but serve different purposes. The certificate proves the part is good. The logbook explains why the die that made it is on its third repair.

Photographs are not optional. A column of free-text observations from twelve different operators across a year is a fraction as useful as twelve photographs taken against the same registration mark in the same lighting. The eye sees patterns in images that no text description captures. A phone camera mounted on a tripod with a ring light, used the same way every inspection, produces an image record that closes the gap between operator vocabulary and visible reality.

Logs that nobody reads are not logs. They are paper. A shop that fills out the log religiously and never reviews it is doing the work without collecting the benefit. The right cadence is a monthly twenty-minute review by the maintenance supervisor across all active dies, looking for the four patterns from Step 3. If the review does not happen, the log is theater.

Closing

Lesson 1 named the four operator-controlled levers (preheat, lube, inspection cadence, change-out timing) and put 90% of the spread in die life inside them. All four improve only with records. Preheat improves when the pyrometer reading on the die face is logged, not remembered. Lube improves when the barrel number and dose are logged, not assumed. Inspection improves when photographs and crack lengths are filed against the hit count, not described from memory. Change-out improves when the pull criterion is documented and reviewed against scrap-rate trend, not negotiated against the shift schedule.

Pick one die on the floor today. Write its serial number, hit count as of this moment, and the seven minimum fields on a clean log. Hand the log to the next operator with the instruction to add an entry at the next 500 or 1000 hits. The instrument starts running on the first entry.

Sources