Maintenance options: polish, repolish, weld repair, re-nitride, re-coat, scrap

Scrap is the right call when one of four conditions is met. The substrate has propagating cracks that re-grind cannot remove without taking the impression below dimensional tolerance. The die has accumulated two or more prior repairs and the metallurgy is now a patchwork of original heat-treat, weld deposits, prior nitride cycles, and re-tempered HAZ. A process change has made the impression geometry obsolete. The economic comparison favors a new build, defined as: the sum of remaining viable repair paths each comes in at more than roughly 50% of new-die effort in engineering, machining, and treatment time, with lower expected post-repair life than a new build.

The last condition is the one that gets skipped on the shop floor. A die that has already been re-nitrided once, repolished twice, and is now showing washout that would require either a weld repair or a substrate re-grind followed by another re-nitride is past the point where keeping it alive is rational. Each repair drops the expected post-repair life, and the cumulative engineering time across the maintenance history is approaching what a new die from drawing would have taken. The maintenance shop holds onto the old die because the original build effort feels sunk, but the right framework is forward-looking: what does the next 50,000 hits cost on this die versus on a new build, and which path delivers them with lower variance.

Scrap also covers the catastrophic case: cracks that have propagated into the bulk, fracture in service, or damage from a press incident. These dies stop being repair candidates and start being failure-analysis specimens. Section the die, document the failure mode, feed the lesson back to the next build, and order the replacement.