Optimization of Metal Forming Process Parameters Using Computational Modelling and Simulation

This study presents a comprehensive framework for optimising metal forming process parameters through integrated computational modelling and advanced optimisation algorithms. A hybrid multi-objective optimisation approach combining gradient-based sequential quadratic programming with evolutionary algorithms was developed and applied to complex shape extrusion operations. The methodology incorporated systematic parameter sensitivity analysis, surrogate modelling techniques, and Pareto-optimisation strategies to navigate the multi-dimensional parameter space while maintaining computational tractability efficiently. The optimisation framework successfully identified optimal parameter combinations for extrusion ratio, ram speed, friction conditions, and geometric constraints through systematically exploring parameter interactions. Surrogate modelling using Kriging interpolation, radial basis functions, and polynomial response surfaces achieved computational efficiency improvements of up to 75% compared to direct finite element optimisation. The modified NSGA-II algorithm demonstrated superior convergence characteristics, generating well-distributed Pareto frontiers that revealed fundamental trade-offs between competing objectives, including force minimisation, energy efficiency maximisation, and process robustness. Parameter sensitivity analysis revealed that second-order interactions between extrusion ratio and friction conditions contributed 23% of total response variability, emphasising the importance of considering parameter coupling effects. The validation schemes indicated a high level of agreement between computational estimates and experimental observations, with the average error of prediction of less than 8 in the parameter space. Multi-objective optimisation identified process improvements of 15-30% in energy efficiency while maintaining force requirements within acceptable limits. This research provides a systematic framework for industrial parameter selection and similarly offers a foundational understanding of computational optimisation of complex manufacturing processes. The framework that has been developed provides a methodological approach to process optimisation, which can balance competing measures of performance against the manufacturing and process uncertainties.