Some gear rating specifications, such as the popular the AGMA 2001, require a minimum percentage of lower transformation products (martensite, lower bainite)1 to a depth of up to 1.2 times the gear tooth length. It is believed that by maximizing the percentage of lower transformation products results in a higher achievable hardness, thus in turn enable higher gear rating calculations. While this seems to be a good idea, in reality there are problems associated with this requirement, especially for through-hardened gears, which are often heat treated as simple gear blanks before the gear teeth are cut. The requirement for AGMA 2001 Grade 2, through hardened gear material, is 10% maximum upper transformation products (upper bainite, pearlite) for sections up to 10 in., and 20% maximum for sections greater than 10 in.
I conducted an experiment using an AISI 4340 steel ring gear blank with a cross section of 9.4 in. wide x 10 in. high that was austenitized at 1600°F, polymer quenched, and tempered at 1050°F. With a surface hardness of 321 HBW, the 90% lower transformation products is not achievable even at 1.2 times the gear tooth length for most tooth sizes for a ring of this diameter. To quantify the transformation products present, I ran computer simulations of the expected transformation products in this ring cross-section after quenching (see Figure 1 and Figure 2). The pearlite formation at/near surface is negligible; therefore the result is not shown. The simulation essentially showed that there will be upper transformation products at or near the surface, and the amount will exceed 10% at depths less than 1.2 times typical gear tooth lengths for a ring of this size. The results may be surprising compared to the general assumption that AISI 4340 steel is a reasonably hardenable alloy. Other gear grades with leaner chemistry and a lower hardenability would certainly get even lower amount of martensite. This suggests that the majority of the time gear materials have not been meeting the % lower transformation requirement, and yet the gear seems to perform as designed.
“Why?” you ask. This may be that 90% martensite/lower bainite is not required to obtain the hardness values needed for the gear rating calculations for through-hardened gears. As long as the hardness used for gear rating calculations are met, whether the percent martensite/lower bainite requirement is met has no direct correlation with the calculation. So why should the gear be considered non-conforming to the standard when the non-conformance does not result in any measurable deficiency? Furthermore, in commercial production, gears rarely get destructively analyzed to check whether the percent martensite/lower bainite requirement is met. One may even question whether the 90% martensite rule even has any published data to support its rationale.
Another difficulty with this requirement is how to differentiate reliably upper from lower transformation products, in specific, distinguishing between the upper and lower bainite. Most through-hardened gear materials are quenched and then tempered at relatively high tempering temperatures, typically 900°F minimum. After such a high tempering temperature, the carbide coarsening that occurs during tempering makes it very difficult or impossible to differentiate between upper and lower bainite. Without a clear separation of the phases, it would be virtually impossible to arrive at any meaningful measurement of each phase—and therefore demonstrate conformance to the requirement.
Between the metallurgical limitation of the material grade, section size, and heat treatment to achieve such a high percentage of lower transformation products, and the challenges associated with phase measurement, it is unlikely that the upper transformation products requirement is enforceable for many of the large gears on the market. Therefore, consideration should be given to revise (or remove entirely) the 90% lower transformation products requirement to reflect the intended benefit without creating any standard conformance/certification dilemma for the gear suppliers.
1Or, alternatively, a maximum amount of lower transformation products (upper bainite, pearlite).