For several years the Royal Navy has used hydraulic fluids which differ significantly from the lubricants used by bearing manufacturers when establishing load to life characteristics. The performance of relevant fluids has been assessed in conjunction with standard EN31 (AISI 52100) steel test specimens using two standard test machines (the rolling four-ball and Unisteel) and a rolling contact fatigue rig designed and built in the author's department.

In addition, a high speed rolling four-ball machine is being commissioned. The test results show that wide variations in performance occur with different lubricant-material combinations, and it is necessary to run tests at several different levels of severity in order to predict service performance reliably. The standard rolling four-ball test did not give good correlation with other test data, and a metallurgical investigation of some failed specimens suggests that an abnormal failure mechanism could be operating.