The Doctor was recently called in to investigate a variable-hardness issue involving rotary cutting dies. The process used to diagnose what was wrong and how to correct the problem offers a number of important lessons. Let’s learn more.

The Problem

Areas of variable hardness (i.e., hard and soft spots) were identified along the blade length (i.e., cut edge) of 4150 rotary cutting dies (Fig. 1). The hardness problem had been occurring, off and on, for a number of years. Some parts came out very good, and others did not. No apparent pattern existed. For example, two mirror-image dies (same size, same material, same manufacturing method, same production time frame) yielded one good and one bad part.

The definition of “hard” and “soft” is the response of material to hand sharpening using carbide tools. Soft areas are where the tools “dig in.” The steel will form burrs and produce a serrated surface, which is virtually impossible to sharpen. Hard areas are those in which the tool “slides across” with minimal resistance. If the part is too hard, however, cut edges are brittle, causing cracks and resulting in scrap parts. The targeted hardness for the dies is 56-62 HRC. The part diameters vary from 50-200 mm (2-8 inches), and the length ranges from 140-635 mm (5.5-25 inches).

The heat-treatment process involved induction hardening and polymer quenching (Fig. 2) followed by induction tempering. After air cooling, the dies were visually inspected for cracks, a hardness check was done and life was tracked throughout the product’s service life in the field.


After thorough analysis, the main focus areas appeared to be: material chemistry (higher-than-normal certified sulfur levels were found) part geometry (given the raised areas on the dies); austenitizing temperature (not measured) and time; quench parameters (polymer concentration, flow, pressure, spray angle, spray area); reheating of a localized area after initial quenching; and tempering temperature (not measured) and time.

Action Plan

The following action plan was established to investigate the problem.

  • Obtain a “good” and “bad” sample for metallurgical analysis.
  • Monitor the induction heating parameters (power/frequency settings, scan speed, dwell time, etc.) for variation (part-to-part, operator-to-operator).
  • Determine hardening and tempering temperature (via Tempil® paint and infrared thermography).
  • Check the quench spray pattern.
  • Determine pressure/flow characteristics during quench.
  • Send parts out for neutral (furnace) hardening, oil quench and oven temper, then check for soft/hard spots.
  • Analyze the parts in the lab as well.

Metallurgical Analysis Work

The surface hardness of the “good” dies averaged 61 HRC. On the “bad” dies, a hardness difference of approximately 1.5 HRC was found between the hard and soft blade areas.

The microstructure of the hard area of the blade consisted of tempered martensite with some ferrite observed. The microstructure in the soft area revealed slightly more ferrite to be present.

The depth of hardening (visually) in the hard area was 2.39-2.41 mm (0.094-0.095 inch) beneath the surface, while the depth of hardening in the soft area was 2.34-2.36 mm (0.092-0.093 inch). The height of the blade above the surface on the dies was approximately 1.32 mm (0.052 inch).

The conclusion reached was that optimizing the process parameters (adequate soak time at temperature, proper quench concentration, spray pattern, quench pressure and flow pattern) would eliminate this hardness differential.

Field Analysis Work

The induction equipment being used was approximately 10 years old; a 150-kW, 10-kHz unit running 218-291 volts, one phase, 7,260-7,424 amps. The range of adjustment was as follows.

  1.     Scan distance:  0-915 mm (0-36 inches)
  2.     Scan speed:  0-150 mm/second (0-6 inches/second)
  3.     Dwell time:  0-99.99 seconds
  4.     Power level:  0-120%
  5.     Quench:  on/off
  6.     Rotation:  0-350 rpm


The questions raised prior to running the process were as follows.

  • Is the power level fluctuating?
  • Is the part rotation proper?
  • What is the (air) gap between the quench ring and the die? Is it adequate?
  • What are the characteristics of the quench?
       • Is the quench duration too long or too short?
       • What is the effectiveness of quench (i.e., analysis of quench variables – supply pressure, flow, polymer concentration, etc.)?
       • Spray quench versus dip quench – which is better?
       • Is 8% the proper variable polymer concentration?


An abbreviated outline follows. Full details are available online.

  • Output power regulation            
       • Constant output power is required.
       • The output power limit was not reached.
  • Austenitizing temperature
       • Temperature at existing settings varies.
       • Decrease scan speed to increase scan time.
  • Part wobble
       • Minor but noticeable.
  • Polymer quench
       • Less backwash and more cascading are required.
       • Refractometer readings were being improperly performed.
       • Some quench holes were plugged.
  • Residual part heat
       • The amount of residual heat present on the part varied.
       • Variation in temperature due to part mass.
  • Tempering
       • Recipes were not being used for tempering.
  • Barber-pole effect
       • Slight (visually)
       • Increase RPM to reduce effect
  • Other
       • Water overflows onto shop floor

The Findings

The root cause of the problem was poor quenching. Some of the quench holes were partially plugged with mineral deposits. The polymer being used had a definite odor to it, and the quench tank itself had a significant amount of scale buildup. The heat exchangers were also suspected of having mineral-deposit buildup and were in need of cleaning (i.e., acid flushing).

During the course of this investigation all of the process parameters were better understood, and the process recipe was optimized.

Conventional oven tempering was also recommended over induction tempering. The dies should be immediately placed in a tempering oven running at 150˚C (300°F) allowing 1.5 hours (minimum) heat-up time and one hour per inch for soak.


The lessons learned, through a combination of laboratory and shop-floor testing, show how to best solve heat-treat problems.



1. Herring, Daniel H., Atmosphere Heat Treatment, Volume I, 2014