PSI - Issue 52

Ding Zhou et al. / Procedia Structural Integrity 52 (2024) 430–437 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

436

7

values that fall in three different temperature ranges, namely |ΔT 0 | < 8.7 °C, 8.7 °C < |ΔT 0 | < 15 °C, and |ΔT 0 | > 15 °C. Clearly, most of the data fall within the uncertainty range such that |ΔT 0 | < 8.7 °C, and the percentage of data in |ΔT 0 | < 8.7 °C increases with the number of data considered in a dataset. Only less than 1% of T 0 are such |ΔT 0 | > 15 °C if more than 9 data are used. Therefore, including low temperature data in the analysis seems appropriate for all cases where limited amounts of materials or specimens exist.

Table 2. Statistics of variations of T 0 with various number of data considered in each data set.

Number of data considered

8

9

10

11

12

13

| ΔT 0 | < 8.7

86.6% 11.8% 1.6%

88.9% 10.2% 0.8%

91.0% 8.4% 0.6%

93.0% 6.7% 0.3%

94.8% 5.1% 0.1%

96.3% 3.7% <0.1%

8.7 < | ΔT 0 | < 15

| ΔT 0 | > 15

Number of data considered

16

17

18

19

20

21

| ΔT 0 | < 8.7

99.2% 0.8%

99.6% 0.4%

99.9% 0.1%

100%

100%

100%

8.7 < | ΔT 0 | < 15

0 0

0 0

0 0

| ΔT 0 | > 15

0

0

0

Fig. 6. Percentage of deviation from the reference T 0 at -50.4 °C with the number of data considered in each data set.

4. Conclusions In the frame of the European project FRACTESUS, fracture toughness in the ductile to brittle transition of the JRQ ferritic steel at various temperatures was measured with sub-sized 0.18T CT specimens. The fracture behavior was found consistent with the Master-Curve approach. The reference temperature T 0 determined with the 0.18T CT specimens (-50.4°C ± 8.7°C) is in good agreement with T 0 previously obtained with larger 0.5T CT specimens on the same material (-42.7 ± 7°C). On the one hand, these results show that a reliable T 0 assessment, fully in line with the requirement of the ASTM-E1921 standard, can be realized with sub-sized 0.18T CT specimens. On the other hand, we demonstrated the possibility to relax a bit the restrictions of the standard by considering fracture data obtained at lower testing temperatures than those recommended. Indeed, including data obtained at T 0 - 50 °C, namely 20 °C lower than recommended, does not seem to affect T 0 from a statistically point of view.