When fracture is by micro-void coalescence and the fine-scale microstructure of a steel is held constant the toughness of the steel is determined by the characteristics of the inclusions in the steel. Three inclusion characteristics which influence toughness are the inclusion volume fraction, the inclusion spacing and the resistance of the inclusions to void nucleation. A measure of the inclusion spacing which is often used in assessing the effect of inclusion spacing on toughness is the average nearest-neighbor-distance between the inclusions in the volume. One can also determine an inclusion spacing by determining the average spacing of inclusion nucleated voids on the fracture surface. The first objective of this work was to determine the spacing of the inclusion nucleated voids on the fracture surface and compare it to the average nearest-neighbor-distance between inclusions in the volume. The spacing of such voids on the fracture surface is about 2.5 times the average nearest-neighbor-distance between inclusions in the volume. The fraction of the fracture surface covered by voids which actually contain inclusions is very small, about 0.03. That this number is so small suggests the possibility that many inclusion particles are simply not retained on the fracture surface. To examine this possibility the sizes of all voids containing inclusions were measured and the size of the smallest void which contained an inclusion was determined. It was then assumed that all voids which were of this size and larger and which did not contain an inclusion were actually nucleated at an inclusion. The number of such voids was then determined. Thus, the number of voids which were nucleated by inclusions was taken to be this quantity plus the number of voids actually containing inclusions. Based on this assumption one can show that the fraction of the inclusions actually retained on the fracture surface is very small, that the spacing of such voids on the fracture surface is very close to the average nearest neighbor spacing of inclusions in the volume and that the fraction of the fracture surface covered by such voids ranges from 0.36 to 0.64 for the four steels investigated, which is considerably larger than the fraction of the fracture surface covered by inclusions which actually contain inclusions, about 0.03 for the four steels examined.

When the fracture mode is ductile the toughness of steel depends on the fine-scale microstructure and the characteristics of the inclusions. The characteristics of the inclusions which influence toughness are volume fraction, spacing and resistance to void nucleation. The purpose of this paper is to discuss the effect of inclusion volume fraction on toughness. Results obtained from 9 nickel steel in which the inclusions are MnS suggest that for a constant fine-scale microstructure and fixed inclusion spacing the crack tip opening displacement at fracture scales as f−1/3 where f is the inclusion volume fraction. This dependence of the crack tip opening displacement on volume fraction has been evaluated for the low alloy steel investigated by Birkle et al. and the scaling seems to hold for this steel. When the inclusion volume fraction is changed at constant inclusion spacing the number of inclusion particles per unit volume remains the same and thus changing inclusion volume means that the sizes of the inclusion particles change. If the crack tip opening displacement scales as f−1/3 then this implies that the crack tip opening displacement would scale as the inverse of the particle size. That toughness would decrease with increasing inclusion size could be due to two factors. First, the Rice and Tracey equations for void growth predict that the rate of void growth increases with increasing particle size. Second, for a constant particle type, one would expect the void nucleation resistance of the particle to decrease with increasing particle size.

Studies of commercial heats of AF1410 steel suggest that under appropriate conditions additions of rare-earth elements can significantly enhance fracture toughness. This improvement in toughness is not due to an extremely low inclusion volume fraction but is apparently due to the formation of larger and more widely spaced inclusions. The purpose of this work is to discuss our experience in using rare-earth additions to laboratory scale vacuum induction melted and subsequently vacuum arc remelted heats of ultra-high strength steels to achieve inclusion distributions similar to those observed in commercial heats modified with lanthanum additions. The results indicate that lanthanum additions of 0.015 wt.% to low sulfur steels which have been well deoxidized using carbon-vacuum deoxidation can result in lanthanum rich inclusions which are similar in size, volume fraction and spacing to those obtained in commercially produced heats of ultra-high strength steel to which lanthanum has been added. The heat of steel to which lanthanum additions of 0.015 wt.% were made had significantly higher toughness than did the heat of the same steel in which the sulfur had been gettered as small and closely spaced particles of MnS and which had an inclusion volume fraction similar to that of the heat modified by the addition of 0.015 wt.% lanthanum. This improvement in toughness was attributed to an increase in inclusion spacing. An addition of 0.06 wt.% lanthanum was excessive. Such an addition of lanthanum resulted in a huge volume fraction of large cuboidal inclusions which primarily contain lanthanum and oxygen and which are extremely detrimental to toughness.

Void generation has been compared for two heats of AF1410 steel in which the sulfur has been gettered either as particles of MnS or as CrS particles. Voids are nucleated at lower plastic strains when sulfur is gettered as CrS than when sulfur is gettered as MnS. Diffraction results indicate that the CrS has a monoclinic structure.