Applicazione di tecniche DoE e RSM per la stima degli effetti del trattamento plasma e laser su giunti incollati

The path undertaken during these three years of Ph.D. focused on two main areas of study: - identifying the effect of low-pressure plasma (LPP) surface treatment parameters on the lap shear strength of adhesive bonded joints made using different substrates. - To determine the optimal laser treatment settings as a surface preparation method for CFRP before bonding. A ytterbium-doped fiber laser source was used in combination with a scanning system. Specifically, for the first point under investigation, four different polymer substrates were considered: 5- and 7-layer carbon fiber reinforced polymers (CFRPs), polyamide 6 and 6.6 (PA6 and PA6.6). To study the effects of the two input parameters, namely power and treatment time, two-level factorial designs were used for each substrate type. The analysis was carried out considering different types of processing gas. The objective function studied was the tensile strength (TSS) of bonded joints. For each set of joints, shear strength values were compared using the DoE approach in order to detect any systematic behavior between different substrates. Finally, it was possible to identify the set-up parameters that gave the best performance in terms of shear strength, considering any equivalent conditions from a statistical point of view. This aspect is particularly important in view of the optimization of production cycle processes; in fact, it allows the maximization of joint efficiency by limiting the energy cost for treatment. In the second part of the study, DoE was used to model the tensile shear strength (TSS) of glued joints and to study the effects of varying three parameters, namely power, pitch and lateral overlap. The analysis was performed considering different focal distances. For each set of joints, shear strength values were modeled using Response Surface Methodology (RSM) in order to identify the best performing set-up parameters, considering any statistically equivalent condition. Regression models also allow prediction of joint behavior for parameter settings not tested experimentally. This is particularly important in view of process optimization of the production cycle; in fact, it allows maximizing the efficiency of the joint while limiting the energy cost for treatment. In order to achieve the results obtained, I had the opportunity to join a proven working group composed of Prof. Lucia Cassettari and Chiara Mandolfino, among others, and to complete a Ph.D. course that, in addition to having contributed to the achievement of important goals in the field of research, has undoubtedly led to an increase in the know-how and skills I developed in my previous graduate studies.