Step 2 of 4
43%
Diagnostic workflow when a die starts losing tolerance
The triage order that keeps a shop from replacing a working die because the spray bar moved. Part first, die surface second, process third, press last.
Step 2 of 4The die surface, without a pull
If step one points at the die or does not resolve the cause, the next step is a visual inspection of the die face on the press. The point of step two is to confirm or rule out the die before any tooling decision, not to perform a full inspection.
Open the press guard. Wipe the die face with a clean cloth if it is safe to do so. Photograph the impression at the same angle and lighting as the last inspection record. The photographic record is what makes step two diagnostic rather than impressionistic. A side-by-side at the same angle catches a localized washout that a fresh look does not.
What you are looking for on the surface:
Heat-check pattern progression. Has the network changed since the last photo? A stable network is normal at 60K hits. A network that has spread into regions that were clean at last inspection is a finding. Individual cracks propagating out of the network are a different finding (and a more serious one; see Lesson 8).
Lube coverage pattern. Wipe a clean cloth across the die face after a lube cycle. The pattern of graphite pickup should match the engineered spray map. A region that comes back clean while neighboring regions are heavily loaded is a spray pattern problem, not a die problem.
Washout signature. Smooth, matte erosion at high-velocity flow regions. Compare the depth and extent to the last photo. If the part drifted by 0.05 mm and the washout region looks unchanged, the die cavity is not the cause and you move to step three.
Adhesive pickup. Smeared or welded material on the die face. Often visible without magnification. Lube failure is the usual cause and that points at step three.
Surface temperature distribution. A quick IR scan across the die face. A 100°C delta from one corner to another is a finding, not a normal state, and it points at cooling-spray pattern (step three) before it points at the die.
Step two does not require a pull. The pull is reserved for when steps one and two have produced findings that justify the downtime impact. Pulling a die has a downtime impact; replacing a die has a larger one.
Quick check
You photograph the die face and find that the heat-check network looks the same as the last inspection photo from 10K hits ago, the impression dimensions look unchanged at the wear-region by visual reference, and the lube coverage pattern on the wipe test shows a clean band 5 mm wide running across one corner. The part is 0.05 mm thick on that corner. Does this rule the die in or out?