We have had the opportunity to apply the most widely used methods of inclusion rating; and, from our experience, it appears that all metallographic methods, including the use of the quantitative television microscope, give roughly equivalent results as far as heat to heat comparison is concerned.

In this respect, the ASTM Recommended Practice for Determining Inclusion Content in Steel (E 45-63), adopted by the French Bearing Steels Quality Standard (NF A 35-565), appears to be one of the simplest and most readily achievable. Nevertheless, the addition of more precise practical rules would be useful in order to improve laboratory to laboratory reproducibility.

From the point of view of service behavior, it should be emphasized that fatigue properties do not depend upon the overall cleanliness of the steel but, mainly, upon the chemical nature of each type of inclusions. The nondetrimental effect of sulfide inclusions is now widely accepted, but the case of oxide inclusions needs to be carefully considered.

The major problem is, therefore, to afford realistic values for the permissible limits of every type of nonmetallic inclusion.