The effects of the crosshead speed, hydrogen content and temperature on fracture strength and fracture surface morphology were investigated using a tempered martensitic steel containing 1.67 mass% of Si (H-Si) and one containing 0.21 mass% of Si (L-Si). When L-Si specimens were charged with a small amount of hydrogen, fracture surfaces showed a transition from quasi-cleavage (QC) to intergranular-like (IG-like) to intergranular (IG) at room temperature. In contrast, when H-Si specimens were charged with a small amount of hydrogen, fracture surfaces showed a transition from QC to IG-like at room temperature. This transition in the fracture surface morphology can be explained by the magnitude relationship between intergranular and transgranular strengths under hydrogen charging. At a temperature of − 196°C, hydrogen did not lower the fracture strength nor did it change the fracture surface morphology. Hence, hydrogen embrittlement at room temperature was presumably caused by hydrogen accumulation and lattice defect formation during stress application as well as by hydrogen trapped before stress was applied. Fracture strength decreased and converged to a constant value (lower critical stress) with decreasing crosshead speed. The crosshead speed for obtaining lower critical stress decreased as the fracture surface changed from IG to IG-like to QC. Therefore, the crosshead speed for obtaining lower critical stress should not be treated as a constant but should be determined experimentally for each type of fracture surface.

The effects of the crosshead speed, hydrogen content and temperature on fracture strength and fracture surface morphology were investigated using a tempered martensitic steel containing 1.67 mass % of Si (H-Si) and one containing 0.21 mass% of Si (L-Si). When L-Si specimens were charged with a small amount of hydrogen, fracture surfaces showed a transition from quasi-cleavage (QC) to intergranular-like (IG-like) to intergranular (IG) at room temperature. In contrast, when H-Si specimens were charged with a small amount of hydrogen, fracture surfaces showed a transition from QC to IG-like at room temperature. This transition in the fracture surface morphology can be explained by the magnitude relationship between intergranular and transgranular strengths under hydrogen charging. At a temperature of-196 °C, hydrogen did not lower the fracture strength nor did it change the fracture surface morphology. Hence, hydrogen embrittlement at room temperature was presumably caused by hydrogen accumulation and lattice defect formation during stress application as well as by hydrogen trapped before stress was applied. Fracture strength decreased and converged to a constant value (lower critical stress) with decreasing crosshead speed. The crosshead speed for obtaining lower critical stress decreased as the fracture surface changed from IG to IG-like to QC. Therefore, the crosshead speed for obtaining lower critical stress should not be treated as a constant but should be determined experimentally for each type of fracture surface.

Particularly in the transition temperature range, reliable use of toughness data, from small specimens of structural steel can be helped by improved understanding of the initiation of cleavage. For this purpose, fracture details in local regions were studied. Methods of examination were scanning electron microscopic (SEM) fractography, microstructural analysis, hardness tests, and topographic examination of fracture surfaces. The results indicate that examination of fracture surface morphology through SEM fractography is essential for improved understanding of cleavage fracturing in the transition range. Fracture surface details reflect influences of both ferrite and prior-austenite grain size. In addition, for some specimens, fracture morphology is also influenced by a large size inhomogeneity due to dendritic solidification. In heavy section A533B steel, explanations for large scatter of small specimen KIc estimates may be closely related to dendritic inhomogeneity.