Rolling Contact Fatigue

Rolling contact fatigue (RCF) tests were conducted using the Caterpillar style test machine. A test roller and load roller were placed in contact under load through a lever arm and dead weights. Using a gear set that drives the roller and test specimen, a -43% slip was induced (surface velocity of the load roller was 43% faster than the test specimen) as compared with pure rolling where the surface velocities of the roller and test specimen are equal. The specimen rotated at 1330 rpm in a bath of 195 °F (90 °C) transmission fluid. A 10 micrometre ceramic element filter removed wear particles from the fluid and the entire fluid system was changed after 1,000 hours of operation. Failure was determined by formation of a pit, spall or surface crack using a vibration transducer to stop the test when such an irregularity was indicated. The test was considered a run out after 7 million cycles. At least 6 tests were run at the failure stress to provide sufficient data for a Weibull analysis of the test results.

A series of 19 test materials were manufactured by industrial sources representing a variety of materials and processing conditions, using densities from 7.1 g/cm3 to near pore-free density. A wrought steel is included to serve as a reference.

The RCF test specimens were machined from blanks provided by 11 different sources. Final dimensions were 2.3 inch (58.4 mm) outer diameter, 0.948 inch (24.1 mm) inner diameter and a stepped face width (loaded width either 0.302 or 0.202 inch [7.67 or 5.13 mm], depending on tensile strength). The blanks were rough machined and ground, then returned to the source for final treatment (such as heat treatment, surface densification, etc.) then final ground and polished.

The load rollers were machined from the same wrought steel (AISI 4620M) that was used for the wrought steel test specimen. The rollers were 3.75 inch (95.25 mm) outer diameter, 1.75 inch (44.45 mm) inner diameter and 0.50 inch (12.7 mm) wide, the width selected to ensure complete coverage of the entire test specimen width while under test. The load rollers were case carburized to 58-63 HRC at a depth of 0.075-0.105 inch (1.90-2.67 mm).

The RCF life (in millions of cycles) for the given stress is listed in the columns identified as “G50” and “G10” where G50 indicates the cycles to achieve a 50% failure rate and G10 indicates the number of cycles to achieve a 10% failure rate, statistically analyzed at a 50% confidence level. The column entitled “contact fatigue stress” for a G50 failure rate at 10 million cycles lists the calculated contact stress for a 50% failure rate at 10 million cycles. The Weibull slope value is the slope of the line for the Weibull distribution of the data; typically a higher number is favored.