The microstructural analysis of tool steels yields important information on how forging dies will perform in service. It is particularly important to keep retained austenite in the tool below 4%. Through a series of controlled tests on tool-steel slugs, it was found that cryogenic or oil quenching can help minimize retained austenite in the tool steel and enhance tool-steel performance.

Typical pre-hardened parts manufactured by APT.

Figure 1. Parts emerge from hardening and quenching. The furnace is equipped with an enriched atmosphere and an internal oil-quench system.

Until recently, the results of good heat-treating practices as applied to forging dies were only measured subjectively and experientially. If a tooling item failed prematurely, it was generally blamed on the material or the heat-treat process – and rightfully so. But as a backdrop to the success or failure of any forge tool, many questions are asked: What determines premature failure? How does one know if the metallurgical properties of forged parts are being optimized? What secrets in the heat-treating process help to assure a quality forging?

We know that most tool failures occur because of wear, thermal fatigue or cracking. If tools crack, we make them softer. If they wear too quickly, we make them harder. As principal at Ashland Precision Tooling (APT), a job shop that specializes in the manufacturing, heat treating and nitriding of tooling for the forging industry, I set out to scientifically define and evaluate the effect of heat-treating practices on the metallurgical quality of H-13 tool steel and draw conclusions as to how we may improve our processes.

Like most quality forge tool manufacturers, I have concluded that starting with a certified, high-quality grade of steel is imperative – especially with the uncertain composition of foreign steel flooding the market today. Given the starting point of a quality block of H-13 steel, there are many theories about its proper heat treating. However, most toolmakers agree that the most significant quality issue is minimizing levels of retained austenite in the material – a primary cause of premature tool failure.

Effect of Heat Treatments on Retained Austenite

When APT started cryogenically treating masters and gauges to achieve stability, we found that retained austenite could be measured in some metallurgical laboratories with X-ray diffraction testing. For the research we began this year, we developed a variety of heat-treating, quenching and tempering scenarios. With this in mind, we conducted a series of tests to gain fresh insight on potential best practices and to define the effects of specific thermal treatments. Our goal was to determine the best way to optimize quality, efficiency and cost-effectiveness in our heat-treatment process.

We conducted a total of eight tests. Test slugs of material (2 inches in diameter and 2 inches in length) were cut from the same bar of material and each piece was identified with a serial number. Each part was routed and processed as an individual job to reduce the probability of error. The parts were all the same size and were treated in the same furnace – an All Case furnace with a 0.65% carbon-enriched atmosphere at 1850°F. All parts were flash quenched in oil for 45 seconds (a standard practice at APT for any part with a section size greater than 1 inch). After flash quenching, all parts were allowed to cool to 200°F.

Until this point, all parts were treated identically. The effects of subsequent variations after the common treatment were then observed as follows:
  • Test 1: Establishing a Baseline – This test was aimed at establishing a baseline from which all other tests could be measured. The only post-heat-treat process that was conducted was a 1000°F temper. As expected, this test yielded the highest level of retained austenite, which measured 2.1%.
  • Test 2: Effects of a Second Temper – This test was aimed at measuring the results of a second temper. The part was allowed to cool to room temperature after the first temper and then subjected to a second temper at 1000°F. As expected, the results were better but not significantly so. The retained austenite measured 2.0%.
  • Test 3: Normal Oil Quench + 24-hour Deep Freeze – This trial measured the results of a normal oil quench, a cool to 200°F and then a continuation of the quench to -100°F in a freezer for 24 hours. After the deep freeze, the parts were warmed back to room temperature then tempered to 1000°F. Once again, the results were better. The retained austenite measured 1.8%.
  • Test 4: Normal Oil Quench + 48-hour Deep Freeze – Test 4 was the same as Test 3 except the parts were left in the deep freeze for 48 hours. The retained austenite measured 1.5%.
  • Test 5: Normal Oil Quench + 24-hour Deep Freeze + Second Temper – Test 5 was also very similar to Test 3. In this test, a second temper was added after a 24-hour deep freeze and the 1000°F temper. The retained austenite measured 1.6 %.
  • Test 6: Normal Oil Quench + 48-hour Deep Freeze + Second Temper – This test was designed to see if the quench time at -100°F had an impact. The test was identical to Test 5 except the parts were left in the deep freeze at -100°F for 48 hours instead of 24 hours. Once again, the retained austenite measured a little lower at 1.5%.
  • Test 7: Normal Quench + Cool to 200°F/Cool to -100°F + 4-hour Freeze – Test 7 measured the effect of a quench conducted in liquid nitrogen. In this test, the parts were flash quenched in oil, cooled to 200°F and then taken down to -100°F for six hours. After that, the parts were submerged in liquid nitrogen for approximately four hours. The parts were warmed to room temperature then tempered to 1000°F. The retained austenite measured 1.7%.
  • Test 8: Normal Quench + Cool to 200°F/Cool to -100°F + 4-hour Freeze + Second Temper – This test was designed to see if a second temper made a difference when the parts were subjected to a -300°F quench. It was exactly the same as Test 7 except the second temper was added at the end of the process. There was no measurable difference in the retained austenite from Test 7. They both measured 1.7%.
After the fifth test, the results started to indicate that a quench with deep freeze had more impact than the number of tempers. With the deep freeze, austenite levels fell from 2.0% to 1.5% – a total reduction of 25%. Also, second tempers had little effect on the reduction of retained austenite.

Interpretation of Test Results

Although it would be easy to conclude at first glance that cryogenically treated parts were better than those that were tempered once or even twice, we thought it made sense to bring in someone from outside our organization to evaluate our findings. We enlisted the help and opinion of Greg Denis, technical director of Ellwood Specialty Steel. We gave him our samples and asked several key questions:
  • How do the microstructures of our eight test slugs appear under a microscope?

    Denis: We found that all the slugs exhibited the same or very similar microstructures. All the slugs were very fine-grained, exhibited homogeneous structures and all represented austenitizing at the lower side of the typical hardening temperature range. All slugs exhibited an excellent quench with little to no grain-boundary carbide precipitation. No visual differences associated with the small range of retained-austenite levels in the martensitic matrix of the various test conditions were observed. Finally, all the slugs displayed microstructures that would be judged to represent very good to excellent heat-treat quality and response.
  • Would the cryogenically treated parts perform better in service?

    Denis: The material and heat-treated microstructures do not suggest any real economic support for the added cost of cryogenic treatment based on anticipated improvement of tooling performance. This is mostly influenced by the fact that the base heat treatment is effective enough that further cryogenic treatment is unnecessary.
  • What is an acceptable amount of retained austenite, and when does it start to affect performance?
    Denis: A level of stabilized retained austenite at 2% is very good for H-13 steel. Even higher levels, say 4%, should generally not present performance issues for most tooling applications. Higher levels of retained austenite could well suggest issues with improper heat treatment. Obviously, 1.5% of retained austenite is better, but not significantly so when already at a 2% quality level. All of these tests are somewhat pre-empted by the excellent base heat treatment of the test program, and at least a double temper is still recommended.

Figure 2. Parts as they emerge from hardening and quenching.

What the Research Tells Us

Most forge tool operations cannot identify the level of retained austenite in their products. For us it was encouraging, but more importantly insightful, to find out that our 2% austenite level was 50% below the acceptable standard. For any forge tooling operation desiring quality or cost-saving gains, getting a benchmark on its current parts is the first step. My recommendation is to send actual parts to a laboratory for an X-ray diffraction test and find out if the test sample is above or below 4% retained austenite.

If, on the one hand, tool-steel tests show retained-austenite levels below 4%, the tools may be over-treated and there may be an opportunity to examine processes for cost savings. If, on the other hand, forging tools show retained austenite levels in excess of 4%, exploring techniques like cryogenic quenching or oil quenching with your heat-treating provider may be in order.

Our research showed a 25% drop in retained austenite by using cryogenic quenches. If your initial heat treat and quench is done properly, however, it will probably be unnecessary.

Finally, for those curious on how APT attained 2% retained austenite, I believe the key factors are our oil-quenching method coupled with treating parts in similar batch sizes. The tests we conducted were on identical batch sizes, which allowed the development of optimal time and temperature settings that do not under-treat the large parts or over-treat the small parts.

Author Steve Englet is president of Ashland Precision Tooling LLC, Ashland, Ohio. He may be reached at: Ph. 419-282-5726; E-mail For additional information visit

SIDEBAR: About Ashland Precision Tooling

APT is one of North America’s major independent toolmakers. The company is distinguished by its total in-house control of the tooling process from the purchase of raw materials through to finishing operations, including heat treatment. The company’s 50,000-square-foot facility in Ashland, Ohio, has produced more than 30,000 hot-forge and cold-form tools. APT employs between 40 and 50 people, with the average experience of its key toolmakers in excess of 20 years.