An experimental investigation on fracture toughness predictions for carbon segregation metals using spherical indentation tests

The occurrence of carbon segregation in steam generator (SG) often leads to a decrease in fracture toughness and an increase in failure risk. To ensure the service safety of SG, a non-destructive testing method for quantitative evaluation of fracture toughness reduction with carbon segregation is necessary. To this end, this study provides an experimental investigation on whether the spherical indentation tests (SITs) are capable of revealing the fracture toughness reduction with the increasing carbon content. Solidarity of the existing fracture toughness prediction models has been extensively investigated through experiments on four carbon segregation samples with carbon content 0.21 %, 0.31 %, 0.35 %, and 0.39 %, respectively. It is found that both the critical strain and critical damage criteria, depending on phenomenologically summarized fixed critical values, failed in reproducing the decreasing trend of fracture toughness with increasing carbon content. For the critical stress criterion, the updated critical value, achieved by comparing the results of conventional fracture toughness and indentation tests on the steel with 0.21 % carbon content, can improve the prediction accuracy and successfully reproduce of variation of fracture toughness with carbon contents. However, consistency of three repeated predictions from the critical stress criteria is poor, which may hinder its engineering application. By contrast, the energy release rate model independent of phenomenologically summarized critical values can yield roughly well predictions, from viewpoints of both decreasing trend of fracture toughness with increasing carbon content and the repeatability of three tests. This experimental investigation can provide methodological guidance for nondestructive fracture toughness evaluation on SG facing carbon segregation. •Non-destructive SIT has been proposed for quantitative evaluation of fracture toughness reduction.•Solidarity of the four representative fracture toughness prediction models have been extensively investigated.•The critical stress model has been improved.•Source of errors in each model has been discussed in detail.