CoatingIQ
← Course index

Wrap-up

86%

Lesson 06·Forge Die Building

Distortion budget: how it propagates through the build

Allocating dimensional change across rough machining, heat treat, finish, EDM, and nitriding so the cavity tolerance survives the build.

6 min readLesson 6 of 13

Tying it together

What this means on the shop floor

For a hammer die at 44 to 46 HRC with a deeper compound layer, the budget is more forgiving. For a press die at 48 to 50 HRC with a 0.05 mm flatness call on the parting face, every step matters and the heat-treat fixturing has to be specified on the print. For an aluminum forging die going to PVD over nitride, the nitride growth eats some of the PVD adhesion budget, and Lesson 8 covers the polish that reclaims it. For an Inconel 718 forging die, the dimensional change at solution and aging is different in magnitude but follows the same propagation logic.

Pushback questions for the vendor

  1. What is your quench fixturing for this geometry, and can you provide the cycle trace with thermocouples on the part rather than only in the furnace?
  2. What distortion has your shop measured on similar geometry in similar steel, and can you ship a witness coupon machined to the same finish?
  3. If the part comes back outside the distortion tolerance, what is the rework path and who owns the finish grind?
  4. For the nitride cycle, what is your typical dimensional growth on H13 at the case depth you are targeting, and is it documented per batch or per quote?

A heat treater who can produce a cycle trace and a measured distortion log is running a process. One who quotes "guaranteed under 0.05 percent distortion" without specifying geometry, section thickness, or quench medium is naming a number, not delivering one.

Common confusions

The 0.05 to 0.15 percent heat-treat figure is not a tolerance. It is a typical range, and a symmetric block with high-pressure gas quench lands at the low end while an asymmetric block with oil quench lands at the high end or worse.

Nitride growth is not zero on H13. The 0.02 to 0.05 percent is real growth on a real part, and on a 500 mm dimension it matters.

Stress relief between operations does not reset the bulk temper if it runs at least 20°C below the lowest prior temper temperature. A 550°C relief on a block tempered at 580°C is safe. A 600°C relief on the same block is not.

EDM does not warp the part, but the recast layer redistributes its stress at the next thermal step. Stress relief after EDM is the operation that prevents the problem from showing up at nitride.

Up next: EDM after heat treat.

Sources