Antonio Viscusi

Fe-Nd-B Rare earth magnets are widely used in high performance applications but suffer from poor corrosion resistance due to their multiphase microstructure and the preferential dissolution of rare earth rich intergranular phases. In this study, the corrosion behaviour of Fe-Nd-B magnets was investigated and the effectiveness of electroless nickel coatings as a protective solution was evaluated. Microstructural and compositional analyses were combined with potentiodynamic polarisation tests, comparing bare and coated magnets with low alloy steel. Coating adhesion was assessed by pull off testing according to ASTM D4541. The results show that Fe-Nd-B magnets are significantly more susceptible to corrosion than steel, while nickel exhibits a passive behaviour in the investigated environment. Electroless nickel coatings provide effective protection and display adhesion comparable to that measured on steel substrates, demonstrating their suitability for corrosion protection of Fe-Nd-B magnets.

This work introduces an initial finite element (FE) framework for modelling particle-substrate interaction during fluidized bed surface finishing of Laser Powder Bed Fusion (L-PBF) components. Due to the complexity of as-built surface morphology and the difficulty of experimentally observing high-speed particle impacts, the mechanisms governing material removal remain poorly understood. The proposed 3D explicit FE model simulates the impact of stainless-steel particles on representative AlSi10Mg asperities using Johnson-Cook plasticity model and damage formulations. Results show that erosion occurs mainly through localized brittle-like detachment rather than extensive plastic deformation. Sequential impacts and oblique trajectories significantly increase internal energy absorption, enhancing asperity fragmentation and the surface smoothing level. The framework provides a foundation for future optimization of Fluidized Bed Finishing (FBF) parameters for improved finishing of additively manufactured metal parts.

This study investigates the applicability of deep learning models for automated quality classification of cold spray coatings, focusing on three deposition categories: good, degraded, and poor deposition. Three state-of-the-art convolutional architectures, ResNet-50, EfficientNet-B0, and ConvNeXt-Tiny, were evaluated across two training phases designed to assess the impact of dataset balancing, data augmentation, and higher input resolution. In the first phase, models were trained on an imbalanced dataset using only class weighting; EfficientNet-B0 achieved the best performance (ACC 80%, F1 77%), while ResNet-50 showed notable instability (ACC 60%, F1 56%). In the refined second phase, oversampling, advanced augmentation, 380×380 resolution, and early stopping led to substantial performance gains for all models. ConvNeXt-Tiny achieved the most robust and balanced results (ACC 93.3%, F1 90.3%), outperforming EfficientNet-B0 and ResNet-50 particularly in sensitivity and specificity for minority classes. Grad-CAM analysis provided qualitative insights into the decision-making process: poor samples elicited strong, spatially extended activations corresponding to defective regions, degraded samples produced more localized responses aligned with mid-scale irregularities, and good samples yielded diffuse, low-intensity activation patterns associated with surface uniformity. These interpretable attention maps validated the physical relevance of the learned features and confirmed the suitability of ConvNeXt-Tiny for reliable and explainable cold spray quality assessment.

Aiming to minimize time, energy, and materials-consuming trial and error experimental analyses, a numerical modeling approach of vitrimer flow and cold spray deposition is proposed in this work. The characteristics of vitrimeric matrices were evaluated by elaborating data from previously performed differential scanning calorimetry and dynamic mechanical analysis. The pieces of information related to the transition temperatures and mechanical evolution after curing were exploited to feed the numerical models and to run sensitivity analyses. The flow model is based on prior evaluation of the dry reinforcement permeability at a micro- and meso-scale. The flow model has been implemented using a commercial simulative environment based on the control volumes approach. A single impacting particle was simulated in a finite element environment to analyze, in a focused way, the deposition mechanisms.The objective of this analysis is the integrated implementation of a numerical model for vitrimer flow through carbon fabric reinforcement in infusion processes and single particle deposition on vitrimer matrix composite substrates.