CoatingIQ
← Course index

Wrap-up

88%

Lesson 07·Running Forge Dies

Inspection cadence: what to look for at 1K, 10K, 50K, and 100K hits

A hit-count inspection checklist that tells you what is normal at each milestone of a forge die's life and what is the early warning the operator must not miss.

6 min readLesson 7 of 13

Tying it together

Cadence runs on hit count, not the calendar

A forge die's life is measured in hits. A die that runs 200 parts a shift on a high-volume line reaches 50K hits in five weeks. The same die on a development line, running 30 parts a shift, reaches 50K hits in eight months. The wear state at 50K is the wear state at 50K. The number of weeks since the die was put in service does not matter to the steel.

Inspection cadence on a forge die is keyed to hit count. The four milestones below cover the working life of a typical H13 die under good preheat and lube discipline. Every shop adapts the exact numbers to its own die-life history, but the pattern is consistent. The numbers shift, the checklist does not.

What this means on the shop floor

For closed-die forge, the four inspection points are usually the flash land, the deepest internal corner of the impression, the parting line on a load-bearing axis, and the flow region with the highest material velocity. Pick them at build, photograph them, and inspect those four every milestone.

For upsetters and hot-headers, the punch face is the first point and is the one most likely to over-temper or pick up. The die cavity floor and the corner radius at the cavity entrance are the next two.

For ring-roll dies, the work-roll face is the first point. The mandrel sees the longest sustained contact and is the second. The face of the die that contains the ring is the third, and it shows the heat-check network earliest.

Across all of them, the cadence is the same. The points are different.

Internal pushback questions

  1. Which four points on each active die in the shop are the inspection points, and are those points written on the die's inspection sheet or on a sticky note in the foreman's desk?
  2. When was the last time two inspectors on different shifts inspected the same die in the same week, and did their findings agree to within one notch of the wear scale?
  3. Where are the photos from every inspection of every die for the last six months, and can any operator pull up the 10K photo of die serial 1247 inside two minutes?
  4. If a 2 mm crack shows up at a 50K inspection that was not in the 10K record, who has authority to stop production on the die, and is that authority defined in writing or is it whoever the line supervisor calls?

Common confusions

A heat-check network at 10K is not the same finding as a heat-check network at 50K. At 10K it is the early signal that the thermal cycle is running. At 50K it is the working state. The same word on the log entry is two different findings at the two milestones.

A signed log is not an inspection record. A log entry that reads "checked, no change" without a measurement, a photo, or a dye-pen result is a signature, and the next inspector inherits no information from it. Measurements on the log, photos on file, gauge readings to one decimal place, or the log is decorative.

A die hitting tolerance is not a die in good condition. Parts at the edge of the tolerance band, scrap rate creeping up, cycle time creeping up, and an inspection record showing dimensional drift across the last three milestones is a die telling you it is at end of life. The tolerance check is the last metric to fail. Every other metric is earlier.

Hit count is not calendar weeks. Two dies with the same hit count have the same wear state regardless of how long it took each die to reach the number. The hit count is the inspection trigger. The calendar is the scheduling convenience.

Up next: reading wear: crazing, washout, abrasion, adhesive pickup, plastic deformation.

Sources