Technical Reports

Roller Bearing Failure Mechanisms Research

  • 01
  • Aug
  • 1994
AUTHOR: Robert L. Florom
OFFICE: RPD
REPORT NUMBER: DOT/FRA/ORD-94/21
SUBJECT: Facilities & Test Equipment
KEYWORDS: Wayside Detection, Hot Bearing Detection, Acoustic Emissions, Grooved Axle Journal, Brinnel, Spall
ABSTRACT: Mechanisms that cause roller bearing failure are not well understood and were the subject of a research program conducted by the Association of American Railroads (AAR), Transportation Test Center, Pueblo, Colorado. The program was funded by the Federal Railroad Administration under Task Order 45. Additional support for the program was provided by Brenco Incorporated, Servo Corporation of America, Timken Company and RASTECH. The program included three separate tests: (1) Roller Bearing Failure Mechanisms, (2) Cone Bore Growth, and (3) Raceway Defect Growth Rate. The Roller Bearing Failure Mechanisms test results demonstrated that bearings with grooved axle journal defects develop measurable temperature gradients across the cup surface, which can be detected using wayside Hot Bearing Detector systems configured to scan both the inboard and outboard bearing raceways to identify wheel sets with this defect. Audible acoustic emissions generated by bearings having this defect did not generate any Acoustic Bearing Detector (ABD) system alarms. The Cone Bore Growth test results demonstrated that while the initial rate of cone bore growth is higher for bearings having a 0.0050-inch interference fit than for bearings having a 0.0025-inch interference fit, the rate of bore growth slows significantly with accumulated mileage. The interference fit of bearings with an initial interference fit of 0.0050 inch remained near the maximum allowed by current AAR specifications over the 20,000 mile test. The Raceway Defect Growth Rate test results showed that measurable growth occurred for the cup spall defect, while no measurable growth occurred for the cup Brinnel and cone spall defects during the 20,000 mile test. The ABD system consistently identified the bearing with the cup Brinnel as defective, but did not consistently identify the bearings with the cup and cone spalls as defective.

####