Strength of hardmetals

Abstract

A statistical analysis for describing failure probability based on defect-initiated fracture gives good quantitative agreement with the results of bend tests on hardmetals (cemented carbides). The variation in strength level in testpieces and components is dependent on the specimen volume, the stress gradient, a toughness parameter, and the defect size distribution, which under some circumstances can be determined from the size distribution of defects that caused failure in a sample of testpieces. By modifying the analysis to take into account an experimentally observed upper limit on the strength of hardmetals it is shown that there are optimum specimen sizes and defect size distributions for increasing the average strength of hardmetal specimens and components. by reducing the number of defect-initiated fractures. Sources of error in adopting a phenomenological approach to the statistical analysis of failures in hardmetals are pointed out.