May_June_AMP_Digital

FEATURE A D V A N C E D M A T E R I A L S & P R O C E S S E S | M A Y / J U N E 2 0 1 9 4 0 tained during the carburization process due to the increase in carbon content in the near-surface layer lowering themar- tensite transformation start and finish temperatures. When the carburized part is quenched, much of the surface layer is transformed to martensite, while some is retained in the form of austenite. Austenite content can be reduced by cold treating, additional tempering cycles, and cold working the surface material. ROLLING CONTACT FATIGUE Rolling contact fatigue (RCF) testing was performed using three identical test units specially designed to hold a sample with a simple cylindrical geometry to expedite test- ing by minimizing complex variables inherent in full-scale component tests [14] . RCF test units rotate a sample stressed in rolling contact by three balls separated by a bronze retain- er, which is radially loaded against the test sample by two tapered bearing rings thrust-loaded by three compression springs. Compression spring loads are calibrated prior to testing using a load cell. The primary failure mechanism of rolling contact fa- tigue is spalling—progressive damage induced by cyclic loading inwhich subsurface cracking causes pits to formand propagate as material drops from the surface. Failures are detected by microphones that monitor the increased test- ing noise due to a growing spall failure, automatically ter- minating the test when the noise level reaches a predefined threshold. Figure 5 shows an RCF tester. An SEM image of a typical spall failure is shown in Fig. 6. Fatigue results were analyzed using a Weibull proba- bility distribution [15] . Weibull distributions provide accurate predictions of failure probability and enable life cycle and reliability comparisons between different population groups such as materials and processes. The L10 life, the life at which 10% of the population is predicted to fail, is common- ly used for performance comparisons in bearings and other components. Weibull probability distributions for the RCF tests are shown in Fig. 7. The curved trend of RCF data for two-hour carburization indicates a minimum failure life of about two million cycles. The four, eight, and 12-hour carburization groups have comparable L10 lives of about 3.7million cycles. The 24-hour carburization shows the greatest performance life with a nominal L10 of 18.5 million cycles, an increase of 400%over the average L10 life of the four, eight, and 12-hour groups. This increase in RCF life performance is attributed, 8 Fig. 4 — Retained austenite distributions as a function of depth. Fig. 5 — Rolling contact fatigue testing unit. Fig. 6 — SEM image of a typical spall failure from rolling contact fatigue.