Where the engineering decisions live

Some decisions move die life by more than others. The leverage is not proportional to the work the decision affects.

Steel grade and melting practice (material zone). The largest single lever on thermal-fatigue-limited dies. Standard H13 to premium ESR is roughly 2-4x on heat-check life. ESR H13 to W360 ISOBLOC or Dievar is another 1.5-2x at the same hardness. The decision lives in the material zone and constrains every downstream zone, but the consequence is bounded by the failure mode the die is dying from. Premium steel on a wear-limited die that scraps on cavity wash buys no service life because the substrate properties are not the binding constraint.

Heat treat path (thermal and case zone). Double versus single temper, vacuum versus atmosphere quench, and the cycle trace as a deliverable. A relaxed heat treat does not show on the hardness certificate. It shows in service as early cracking from retained austenite and brittle fresh martensite, plus dimensional growth as the austenite transforms over the first few thousand cycles. The leverage on die life is asymmetric. Holding the full spec is the baseline, and a small relaxation often costs 20-40 percent of cycle life on a die targeting 150K hits.

Internal radii (geometry zone). Opening an internal radius from R 1.5 mm to R 3 mm cuts Kt from roughly 4 to roughly 2.5 and typically doubles or triples fatigue life at that feature (Lesson 11). The decision lives in design, but it propagates through machining accuracy. As-built radius verification is required because finish operations and electrode wear undercut the print radius. The lever is large and the verification is cheap. Skipping the verification negates the design decision.

Case depth and compound layer composition (thermal and case zone). A 0.30-0.40 mm diffusion zone with controlled gamma-prime phase supports the surface contact load and prevents eggshell failure under PVD. Insufficient case depth (under 0.15 mm) transfers contact load to the substrate, the substrate yields plastically, and the cavity washes. Wrong compound layer phase (epsilon-dominant on an impact die, gamma-prime-dominant on a sliding-wear die) shifts the failure mode against what the application needs. The lever is roughly 2-3x on dies where the failure mode aligns with the case.

The lower-leverage decisions are still real. Polish method, EDM finish settings, masking strategy, and inspection artifact format all affect die life, but the magnitude is closer to 10-20 percent than 2-4x. The high-leverage decisions earn extra design review attention. The lower-leverage decisions are where shop-floor practice and vendor selection settle in.