Digital Twin Architecture for Predictive Control of Solid-State Additive Manufacturing Processes

Abstract

This study examined the effectiveness of a digital twin architecture integrated with predictive control for optimizing solid-state additive manufacturing processes. A quantitative experimental design was employed using 120 validated manufacturing runs, where process parameters such as tool speed (mean = 1185 rpm), applied pressure (mean = 5.10 kN), and temperature (mean = 421°C) were systematically varied and monitored through real-time sensor systems. The digital twin model was synchronized with physical operations to generate predictive outputs, which were compared with observed manufacturing results. The findings demonstrated strong predictive alignment, with correlation coefficients ranging from 0.89 to 0.94 across key variables including temperature, deformation index, and bonding strength. The implementation of predictive control resulted in significant performance improvements, with defect rates reduced from 14.2% to 6.3%, dimensional deviation decreasing from 0.48 mm to 0.33 mm, and temperature variability reduced by 29.0%. Process stability improved from 71% under baseline conditions to 92% with predictive control integration. Statistical analysis confirmed that these improvements were significant at the 0.05 level, with large effect sizes observed for defect reduction (d = 0.92) and dimensional accuracy (d = 0.81). Sub-group analysis revealed that moderate parameter ranges produced optimal performance, achieving the highest predictive accuracy (93%) and lowest deformation variability (9%). Regression analysis further indicated strong explanatory power of the digital twin model, with coefficients of determination exceeding 0.87 across all evaluated outputs. The results also highlighted the importance of real-time data synchronization, system scalability, and computational efficiency in maintaining effective predictive control. Overall, the study provided empirical evidence that digital twin-based predictive control systems significantly enhance manufacturing efficiency, process stability, and product quality, supporting their application in advanced data-driven manufacturing environments.