PSI - Issue 61

Onur Ali Batmaz et al. / Procedia Structural Integrity 61 (2024) 305–314 Onur Ali Batmaz et al./ Structural Integrity Procedia 00 (2019) 000 – 000

312 8

Figure 5 . Comparison of reference experiment (top) and FE simulation (bottom) results in terms of transverse shear strain fields γ xy . γ xy data acquired along the centrelines (shown with white dashes) is plotted on the right. In addition to the comparisons based on displacement measurements, transverse shear strain (γ xy ) predictions of FE models with fixed and proposed BCs are compared with the experimental measurements in Figure 5. γ xy contours are obtained from the experiment (top) and the FE simulation with proposed BCs (bottom) at the same impactor translation. These contours show good agreement, with only a few localized discrepancies in the experimental data. These discrepancies, as previously reported by Topac et al. (2017), are attributed to the stochastic distribution of elastic properties within the tested specimen, resulting in localized relaxed strain responses. Modeling these specific local variations is beyond the scope of this study. Furthermore, the variation of γ xy along the centerline of the beam (indicated with white dashes) are plotted for the experiment and FEAs with fixed and proposed BCs. The FE model with proposed BCs captures the overall trend observed in the experimental measurements. However, the FE model with fixed BCs tends to overestimate γ xy values along most of the test section and underestimates them near the boundaries due to the fixed constraining. Figure 6(a) and Figure 6(b) respectively show the influence of proposed BCs approach on the global impact response through impactor force-time history and force-displacement curves, in comparison with the results of FE model with fixed BCs. In both figures, the material response remains elastic until the sudden load drop, which occurs when delamination propagates. The periodic oscillations in the elastic regime occur due to the dynamic coupling of impactor, specimen and BCs (Lopes et al., 2009b). In Figure 6(a), the total failure accompanied by a sudden load drop is postponed from 257 µs to 376 µs when proposed BCs are employed, where this value was reported as 336 µs in the reference experiment. Correspondingly, as observed in Figure 6(b), employing the proposed BCs modelling approach instead of fixed BCs leads to a decline in global stiffness response from 11903 N/mm to 8994 N/mm, where a linear fit is used for the curves up to the load drop for comparison purposes. The impactor displacement at failure is postponed from 0.85-mm to 1.04-mm when proposed BCs are employed. In the reference experiment, this value was reported as 0.95-mm.

(a) (b) Figure 6. Influence of boundary conditions on the impactor (a) force-time history and (b) force-displacement response.