PSI - Issue 68

Julie Papin et al. / Procedia Structural Integrity 68 (2025) 727–733 Julie Papin et al. / Structural Integrity Procedia 00 (2025) 000–000

729

3

hours. Among the eighteen microstructures of Delattre et al., the fracture toughness tests focused on eleven microstructures (Table 1). = ! "'#)$)%). ' ( (20 + ln( )) (1) Table 1. Description of the eleven heat treatments and reference temperatures and standard deviations obtained following ASTM E1921. Quenching rate ( °C/h ) Tempering condition Tempering parameter TP Reference temperature T 0 (1T) ( °C ) Standard deviation ( °C ) 150 640°C – 6h 19.90 -51 7.2 150 640°C – 20h 21.00 -31 6.7 800 640°C – 6h 19.90 -84 6.7 800 640°C – 20h 21.00 -41 6.7 2000 640°C – 1h 18.26 -104 6.7 2000 610°C – 6h 19.25 -110 7.2 2000 640°C – 6h 19.90 -104 6.7 2000 660°C – 6h 20.33 -69 6.7 2000 640°C – 20h 21.00 -56 6.7 8000 640°C – 6h 19.90 -131 6.7 8000 640°C – 20h 21.00 -83 6.7 2.2. Testing conditions and observation of fractured specimens For each heat treatment condition, twelve fracture toughness tests were performed on C(T)0.5T specimens, at three different temperatures (4 tests/temperature), and the Master Curve reference temperature T 0, for which the median value of fracture toughness K Jc (1T) equals 100 MPa.√m , was then identified following ASTM E1921 recommendations (listed in Table 1). In order to select the appropriate values of temperature at which the fracture toughness tests should be performed, the formula proposed in ASTM E1921 standard was used to compute a first estimate of T 0 : T ) ∗ = T41J − 34 °C , where T41J is the temperature for which the absorbed energy of the same microstructure in Charpy impact tests (Delattre et al., 2024) was equal to 41J. The three testing temperatures were then chosen as T ) ∗ −25°C , T ) ∗ and T ) ∗ +25°C . When considered necessary, they were adjusted during the tests. A comprehensive fractographic study of the fracture toughness specimens broken in the range 80 − 120 MPa√m was conducted, namely, 46 specimens from 8 heat treatments up to now (this part of the study being still ongoing). The fractographic analysis method was based on Delattre’s methodology (Delattre et al., 2024, 2022a, 2022b); by following cleavage rivers, the fracture initiation site was identified, then characterised using point energy dispersive spectrometry (EDX) analysis at 10kV voltage on each half of the fractured specimen at high magnification. 3. Results 3.1. Fractographic analysis: the three brittle fracture initiator types Brittle fracture initiation can generally be classified in two types: transgranular fracture by cleavage or intergranular fracture. In this study, the few cases (2/46) of intergranular fracture observed on the initiation sites were rather ‘rough’, isolated surfaces (no extended intergranular area) that seemed to trigger the subsequent dominant brittle fracture crack propagation by cleavage. In all other cases, cleavage did not initiate from intergranular fracture. When found during fractographic analysis the determination of the nature of the particle responsible for crack initiation was sometimes delicate. There were two types of location for cleavage initiation sites, namely, either from an inclusion located far from the grain boundaries, or close to a grain boundary. In this last case, it could be triggered by the fracture of a non-metallic inclusion (Fig. 1a) or of a large carbide (Fig. 1b), but in several cases, no particle could be detected with enough certainty at the initiation site. These observations suggest that the grain boundary played a significant role in the brittle fracture initiation, with or without a defined local carbide or non-metallic inclusion particle playing the role of cleavage initiator. The inclusions identified were Ti(C,N), sometimes associated with MnS and (Al,Mg)-rich oxides (Fig. 1c). Regarding carbides at initation sites, EDX analysis revealed an enrichment in molybdenum, manganese and carbon compared to the matrix (Fig. 1d).