PSI - Issue 68

Cainã Bemfica et al. / Procedia Structural Integrity 68 (2025) 1188 – 1195 Ludovic Vincent et al. / Structural Integrity Procedia 00 (2025) 000–000

1191

4

Figure 1 Comparisons between experimental (continuous lines) and simulated (dashed lines) tensile curves

One can notice that the temperature dependence of the material parameters allows predicting fairly well the evolution of the macroscopic constitutive behaviour of each material, with an exception on MM material at -110 °C and IM material at room temperature. The two model materials have similar yield stress, significantly larger than the one of the IM. The values of the parameters introduced in the constitutive equations (1)-(4) are given in Table 2.

Table 2 – Material constants used in Eqs. (1)-(4) to describe material behavior for FEA. Material E [GPa] ν σ a [MPa] σ T [MPa] T y [°C] p 1 [MPa/°C]

p 2 [MPa]

Q 2 [MPa]

b 2

MM

205 205 205

0.3 411 0.3 420 0.3 358

13.3 25.8 32.4

50.0 66.7 71.4

0.39 0.62 0.69

279 230 230

900 900 900

0.37 0.37 0.37

HMM

IM

The DBT, both in impact toughness and fracture toughness, are represented in Figure 2. One can observe that the DBTT of the HMM material is significantly larger than the ones of the two other materials. Microhardness maps revealed the presence of very limited microsegregation regions in the materials. The histograms of hardness values were similar (Figure 3a) even though the standard deviation for the MM is larger than for the two other materials. The PDF of misorientation angles of the three materials are also very close to each other (Figure 3b). The MM has a slightly larger R ratio value than the two other materials, but all the three values are representative of similar upper bainite crystallographic microstructures. However, the measure of average ferritic grain size obtained from the EBSD maps is significantly different, with a value for the HMM material equal to more than twice the one of the two other materials (MM and IM). This significant difference is a key microstructural element to be accounted for in the MIBF model presented later on. Fractographic exams performed on HMM and IM materials identified cleavage as the dominant brittle fracture mechanism. In only one case over thirteen ones, cleavage initiated from an inclusion located inside a ferritic grain, far from any grain boundaries. In all other cases (12/13), cleavage initiated from grain boundaries. In half of these cases (6/12), a brittle broken particle could be identified as the initiator and always as a carbide, enriched in Mn and/or Mo (Figure 4). This result is consistent with previous investigations that reported carbides as the main initiators of brittle failure for this class of materials (Chekhonin et al., 2023; Delattre, 2023; C. Li et al., 2016; Naylor, 1979; Zhang and Knott, 1999). In the rest of cases where cleavage initiated from grain boundaries, no particles could be identified as a starter for the global failure. The carbide population distribution has been characterized for the three materials and the identified values of the surface density Ns and Lee-Weibull distribution parameters are reported in Table 3. No major difference is observed between the values of the parameters of the three materials. However, the slightly smaller value of the parameter α for the IM material illustrates a larger probability to find larger carbides in this material than in the two other ones.