PSI - Issue 53
Costanzo Bellini et al. / Procedia Structural Integrity 53 (2024) 227–235 Author name / Structural Integrity Procedia 00 (2019) 000 – 000
232
6
numerical one. In particular, the experimental data is slightly shifted towards the right. This is due to the settlement of the specimen on the test fixture, and therefore does not represent a problem. It is evident that the slope of the two curves, representing the stiffness of the test specimens, is similar, and therefore the value of Young's modulus used for the simulation is realistic. On the contrary, there is a difference at the end of the linear part of the load curve. In fact, the load corresponding to the deviation from linearity was about 2600 N for the numerical test and 2400 N for the experimental one. This value depends on the yield strength of the material, and it must be remembered that the material properties adopted for the numerical simulation were taken from bulk specimens. Therefore, the determination of the effective yield strength of the truss material was needed.
Fig. 5. Comparison between experimental and numerical (not-enhanced model) results.
The effective local yield strength σ y can be calculated from hardness tests, as proved by Magarò et al. (2023), thanks to the following equation: = (1) where H represents the local hardness, and k is a constant. This latter parameter was calculated by determining the local hardness through micro-hardness tests on the bulk specimens, and then by dividing the yield strength relevant to those specimens by the hardness value. The micro-hardness measured on the bulk specimen was equal to 371 HV; therefore, being the yield strength 857 MPa, the k value was determined as 2.3. Micro-hardness tests were carried out on the produced specimens too, and a hardness of 393 HV was found for the core material, while the hardness of the skin material was 360 HV. Finally, employing these latter two data, a yield strength of 904 MPa was estimated for the titanium alloy of the core and 828 MPa for the skin. The numerical model was run by considering the redetermined value of yield strength, and the obtained load displacement curve was compared with the experimental one and that got from the previous model, with the yield strength value from the bulk specimen. As denoted in Fig. 6, the first linear portion of the curve calculated by the enhanced model is coincident with that determined by the previous model, but the deviation from linearity happened at a lower load, that is 2470 N. Therefore, the aim of reducing the gap between the experimental and numerical curves was gained by redetermining the yield strength. Finally, SEM micrographs were taken on the fracture surface and subjected to image analysis in order to define the damage mechanism more clearly. The damage behaviour indicated a ductile fracture, as indicated by the dimples visible in Fig. 7, relevant to the fracture surface. Additionally, the aforementioned magnification revealed pores, which
Made with FlippingBook Ebook Creator