PSI - Issue 68

T. Fekete et al. / Procedia Structural Integrity 68 (2025) 687–693

690

4

T. Fekete et al. / Structural Integrity Procedia 00 (2025) 000–000

Fig. 3. (a) Specimen before the test, (b) Specimen right before failure

The experiments were carried out on a Gleeble 3800 universal testing machine. The crosshead was operated at a speed of 1 mm/min. The machine provided Force–Crosshead displacement data and was also equipped with a mechanical extensometer to obtain strain data. A camera was utilized to capture images of the gauge section at a rate of 50 images per second, with a resolution of 2560 × 2048 pixels. Figure 3a and 3b present photographs taken before and during a test, which document the occurrence of necking well away from the center of the gauge length. 3. Digital Twin The implementation of the new SIC methodology into the DT of the safety-critical Large-Scale Pressure System of an NPP is envisaged to occur in the near future. The DT , based on a state-of-the-art physical model, paves the way to interpret the system sensors’ data in a consistent manner, thereby enabling the overcoming of the limitations of pure empiricism (Cellucci 2018). This, in turn, provides strong support for a more reliable prediction of the system’s state of structural health (Rios, Bolander 2023). The rationale behind the development of the DT of the measurements was that if the SIC methodology is included in the System’s DT , it follows that this DT also encompasses the methodology of material science measurements and their evaluation, thereby providing support for the SIC s. It is widely recognized among materials science experimentalists that tensile testing results can vary significantly between specimens, particularly regarding the specimen’s elongation at initiation of necking, the location and extent of necking, the elongation and stress at ultimate failure, and the force-displacement curves. The physical quantities determined in the evaluation of these measurements are subject to a varying degree of uncertainty. In experiments conducted in accordance with the relevant measurement standards, test specimens with dimensions within the specified manufacturing tolerances shall be utilized. This assumes that measurements made on specimens that comply with the standard are not significantly influenced by geometric variations of the specimens. Currently, investigations into fission and fusion energy are increasingly requiring the utilisation of smaller and smaller test specimens (Zheng, Chen, Liu, Chen, Zhang, Liu, Shen 2020). Empirical evidence indicates that the uncertainty associated with measurements conducted on small test specimens is considerably greater than would be anticipated from measurements taken on those of a standard size. The question is what factors have contributed to the significant rise in uncertainty associated with measurements on small test specimens. It is well documented that measurement standards (e.g. ASTM E8/E8M-24 2024) are founded upon relatively straightforward lumped parameter models for the assessment of experimental outcomes. The basis for the DT of measurements is the fact that the theory utilized in the novel SIC methodology can describe geometric and material nonlinearities. Additionally, it is understood from the stability theory of time-evolving nonlinear systems (Bažant, Cedolin 1991) that they may be sensitive to variations in initial conditions. In many experiments, asymmetric necking was observed. This phenomenon can be attributed to either geometric imperfections or material inhomogeneities. The present study focuses on investigating geometrical imperfections, while the issue of material inhomogeneity will be addressed as part of future research. The tests were simulated using Hexagon Marc / Mentat Finite Element software. Three-dimensional models were developed (see Figure 4) using quadratic 20 -node hexahedral elements with an edge length of 0.5 mm at the gauge section. This resulted in the gauge section element size aligning closely with the printed grid on the specimen. The material was assumed to be homogeneous, isotropic, and elastoplastic with isotropic hardening. The Young’s modulus was set to 210 GPa, and the Poisson's ratio was 0.3 . In the models, von Mises plasticity was used as the constitutive model. The yield curve was calculated using the Chung-Cho method (Choung & Cho, 2008) using data from previous tensile tests. The boundary conditions, namely the grips, were implemented by Marc/Mentat’s ‘glue’ contact. In the model, large deformation theory was applied, and the solver used the updated Lagrangean method.