The objective of the investigation was to study the feasi-bility of applying continuous casting steel 100Cr6 (SAE 52100) for bearing balls. It was found that two of three continuous casted steel batches have longer or at least similar rolling contact fatigue lifetimes comparec to one ingot casted batch. For one of the continuous casted batches, the rolling contact fatigue lifetime was 30% less.

The micro- and macrostructure and the residual stresses below the surface of the balls were comparable. There is also no obvious difference between the four batches in the metallurgical parameters like contents of oxygen, titanium and sulphur as well as in the distribution of carbides and their network.

The reason for the shorter lifetime of one batch finally was found to be due not to the continuous casting process itself. There was a difference in the cross section of the different casting moulds, and by this in the speed of solidification. As a consequence an increased segregation of alloying elements was observed. So the primary carbides of this batch were of bigger size and higher amounts have been detected within a larger cross section of the bloom. Finally the inferior distribution and size of carbides in the layers of maximum equivalent stress below the surface were found to be crack initiation points. The reduced lifetime of one batch therefore could be correlated to these metallurgical differences.

It has been demonstrated that bearing balls made from continuous casted steel 100Cr6 (SAE 52100) have similar or better rolling contact fatigue lifetime than bearing balls made from ingot casted material. However, it is necessary to have minimal amounts of micro- and macro-inclusions as well as a homogeneous distribution of carbides, particularly in the main segregation areas. For that reason, the steel quality of every type of continuous casting machine has to be evaluated separately.