Issue 51

C. Bellini et alii, Frattura ed Integrità Strutturale, 51 (2020) 442-448; DOI: 10.3221/IGF-ESIS.51.32 445 tests. The presence of the adhesive improved the ILSS of the material, but it must be remembered that it was detrimental for the flexural strength of the laminate, as found in previous works [17,20]. Figure 3: Results relevant to the three-point bending tests carried out on the produced FML. The consistency of the data obtained from the experimental tests was assessed by statistical analysis; in particular, an ANOVA (Analysis of Variance) was carried out, and the results are reported in Tab. 1. This statistical tool asserted that the contribution factor of the layer thickness was quite low, less than 5%, while the contribution of the interface solution was high, more than 75%. The interaction contribution was calculated too, but the result was negligible and so it is not present in Tab. 1. In order to confirm these findings, the ANOVA was completed with the estimation of the p-value, another statistical parameter: the studied factor is not influencing the results if its p-value is higher than 5%, that is the value commonly used as limit. As visible in the table, the adhesion factor had a p-value equal to 0%, so it was influencing, while the number of sheets had a value of almost 13%, so it was unaffecting. Therefore, it can be concluded that among the studied factor, the only one that conditioned the ILSS value was the presence of the structural adhesive at the interface between metal and composite material. Source Contribution p-value Number of sheets 4.85% 12.9% Presence of adhesive 75.84% 0.0% Experimental error 19.31% Total 100% Table 1: ILSS value measured for all the tested specimen. The factor influence on the short beam strength was investigated also through the main effects plot, another statistical instrument, and the results of this analysis are presented in Fig. 4. As it can be seen from the chart in the figure, the ILSS increased with the adhesive presence, while it decreased with the increment of the number of sheets, even if this decrement was lower compared to the variation induced by the former factor, as stated also by the ANOVA. The trend of the shear strength as a function of the loading nose displacement is presented in Fig. 5. For each kind of laminate, a single curve is reported, that is representative of the relevant group because the data scatter was narrow. The shear stress trend was found comparable for all the specimen type; in fact, at first a shear stress increment was found, then, after the attainment of maximum load, each curve showed a fluctuating trend, due to the presence of various failures. The first part of the curve, till the maximum load, showed a linear trend for the laminate without the adhesive, while the other two types presented a slight knee before attaining the maximum value. This behaviour was probably caused by the plasticization of the adhesive layer, that in the former kind of laminate was not present. The slope of the first part of the