The additive manufacturing (AM) field is rapidly expanding, attracting significant scientific attention. This family of processes will be widely used in the evolution of Industry 4.0, particularly in the production of customized components. However, as the complexity and variability of additive manufacturing processes increase, there is an increasing need for advanced techniques to ensure quality control, optimize performance, and reduce production costs. Multiple tests are required to optimize processing variables for specific equipment and processes, to achieve optimum processing conditions. The application of digital twins (DTs) has significantly enhanced the field of additive manufacturing. A digital twin, abbreviated as DT, refers to a computer-generated model that accurately depicts a real-world object, system, or process. A DT comprises the complete additive manufacturing process, from the initial conception phase to the final manufacturing phase. It enables the manufacturing process to be continuously monitored, studied, and optimized in real time. DT has emerged as an important tool in the additive manufacturing industry. They allow manufacturers to enhance the process, improve product quality, decrease costs, and accelerate innovation. However, the development of DT in AM is an iterative and continuous process. It requires collaboration between domain experts, data scientists, engineers, and manufacturing teams to guarantee an accurate representation of the process by the digital twin. This paper aims to provide a comprehensive analysis of the current state of DT for additive manufacturing, examining their applications, benefits, challenges, and future directions.

This work aimed to study one of the most important challenges in orthopaedic implantations, known as stress shielding of total shoulder implants. This problem arises from the elastic modulus mismatch between the implant and the surrounding tissue, and can result in bone resorption and implant loosening. This objective was addressed by designing and optimising a cellular-based lattice-structured implant to control the stiffness of a humeral implant stem used in shoulder implant applications. This study used a topology lattice-optimisation tool to create different cellular designs that filled the original design of a shoulder implant, and were further analysed using finite element analysis (FEA). A laser powder bed fusion technique was used to fabricate the Ti-6Al-4V test samples, and the obtained material properties were fed to the FEA model. The optimised cellular design was further fabricated using powder bed fusion, and a compression test was carried out to validate the FEA model. The yield strength, elastic modulus, and surface area/volume ratio of the optimised lattice structure, with a strut diameter of 1 mm, length of 5 mm, and 100% lattice percentage in the design space of the implant model were found to be 200 MPa, 5 GPa, and 3.71 mm⁻¹, respectively. The obtained properties indicated that the proposed cellular structure can be effectively applied in total shoulder-replacement surgeries. Ultimately, this approach should lead to improvements in patient mobility, as well as to reducing the need for revision surgeries due to implant loosening.

Secondary dendrite arm spacing (SDAS) is one of the most important factors affecting macrosegregation and mechanical properties in solidification processes. Predicting SDAS is one of the major parameters in foundry technology. In order to predict the evolution of microstructures during the solidification process, we proposed a simple model which predicted the secondary dendrite arm spacing based solely on the tip velocity (related to the tip supersaturation) and cooling rate. The model consisted of a growing cylinder inside a liquid cylindrical envelope. Two important hypotheses were made: (1) Initially the cylinder radius was assumed to equal the dendrite tip radius and (2) the cylindrical envelope had a fixed radius in the order of the dendrite tip diffusion length. The numerical model was tested against experiments using various Pb–Sn alloys for a fixed temperature gradient. The results were found to be in excellent agreement with experimental measurements in terms of SDAS and dendrite tip velocity prediction. This simple model is naturally destined to be implemented as a sub-grid model in volume-averaging models to predict the local microstructure, which in turn directly controls the mushy zone permeability and macrosegregation phenomena.

This study investigates the optimization and prediction of mechanical properties in 3D-printed PLA composites reinforced with graphene nanoplatelets (GNP). The effects of GNP content (0, 2, and 5 wt.%), nozzle temperature (190–210 °C), print speed (20–60 mm/s), and layer thickness (0.15–0.35 mm) on tensile strength, Young’s modulus, and hardness were analyzed using a central composite design, at three print orientations (0°, 45°, and 90°). Compared to pure PLA, the incorporation of 5 wt.% GNP led to a 67% improvement in tensile strength, a 205% increase in Young’s modulus, and a 44% enhancement in hardness. Advanced machine learning models, such as XGBoost and Gaussian Process Regression, were employed for prediction, with R² values exceeding 0.99 and MAPE below 4%. The models were validated using K-Fold Cross-Validation (K = 5), ensuring reliable and robust predictions while preventing overfitting. SHAP (Shapley Additive exPlanations) analysis indicated that GNP composition and layer thickness were the most influential factors, with SHAP values ranging between ±0.75. The Gaussian Process model outperformed both Linear Regression and XGBoost, achieving the highest R² of 0.9900 ± 0.0021, the lowest MSE (0.6593 ± 0.1054), RMSE (0.812 ± 0.323), MAE (0.6755 ± 0.1123), MAPE (3.157% ± 0.320), and RRMSE (3.409% ± 0.513), highlighting its superior predictive accuracy and stability. This integrated methodology, combining experimental optimization, ANOVA, and interpretable machine learning, presents a promising and potentially robust strategy for optimizing the mechanical performance of GNP-reinforced PLA composites, emphasizing their potential for high-performance engineering applications.

The wear loss and frictional characteristics of magnesium-based hybrid composites reinforced with boron carbide (B₄C) particles and graphite filler were the main subjects of the investigation. Key parameters, including reinforcement content (0–10 wt%), applied load (5–30 N), sliding speed (0.5–3 m/s), and sliding distance (500–3000 m), were varied. Data-driven machine learning (ML) algorithms were utilized to identify complex patterns and predict relationships between input variables and output responses. Five distinct machine learning algorithms, Artificial Neural Network (ANN), Random Forest (RF), K-Nearest Neighbor (KNN), Gradient Boosting Machine (GBM), and Support Vector Machine (SVM), were employed to analyze experimental tribological data for predicting wear loss and coefficients of friction (COFs). The performance evaluation showed that ML models effectively predicted friction behavior and wear behavior of magnesium-based hybrid composites using tribological test data. A comparison of model performances revealed that the Gradient Boosting Machine (GBM) provided superior accuracy compared to other machine learning models in predicting both wear loss and the coefficient of friction. Additionally, feature importance analysis indicated that the graphite weight percentage was the most significant influence in predicting the coefficient of friction and wear loss characteristics.