PSI - Issue 33

Costanzo Bellini et al. / Procedia Structural Integrity 33 (2021) 824–831 Author name / Structural Integrity Procedia 00 (2019) 000–000

828

5

3. Results and discussion After carrying out the experimental tests and the numerical simulations, the results are reported in graphs for comparison purposes. As concerns the flexural behaviour, the graph of Fig. 4 shows that the four tested specimens presented good repeatability, especially in the loading phase, since all the loading displacement trends were almost overlapping. The numerical model was able to predict in a good manner the loading phase; in fact, both the stiffness rigidity, that is the slope of the curve, and the maximum load were correctly calculated. Moreover, also the descending trend after the maximum load was predicted in an acceptable manner by the numerical model. As concerns the short beam test, whose results are presented in Fig. 5, also in this case the repeatability of the experimental results, relevant to four specimens, was quite good. The numerical model presented a load increase that was not perfectly linear; in fact, some sudden slope changes can be noted. On the contrary, the maximum load was correctly predicted. Finally, the descending trend, after the maximum load, was not well predicted, since after the first load drop a new increment was found. However, for design purposes, this portion of the curve is not used, so this outcome does not represent a problem.

Fig. 4. Comparison between numerical and experimental results for the long beam specimens.

Fig. 5. Comparison between numerical and experimental results for the short beam specimens.