Empirical comparison of FDM additive manufacturing process parameters by application grey relational analysis and multi-linear regression

Abstract Fused deposition modelling (FDM) additive manufacturing is influenced by a combination of controllable and uncontrollable factors that significantly impact the mechanical properties of the final product. This study introduces a novel application of multi-linear regression (MLR) for optimizing process parameters to achieve superior mechanical properties through multi-response optimization. The MLR results are compared with prior outcomes derived from grey relational analysis (GRA) to assess their relative effectiveness. Six critical process parameters—density ratio, layer height, nozzle speed, print temperature, and shell thickness—were analyzed to optimize four mechanical performance measures: hardness, flexural strength, compressive strength, and ultimate tensile strength (UTS). While GRA optimization favored lower parameter values, MLR identified higher parameter values as optimal for enhancing mechanical performance. Key differences include variations in nozzle speed (120 mm/s), nozzle temperature (210 °C), and bed temperature (80 °C) for specific performance measures. The findings demonstrate that MLR offers greater flexibility by predicting optimal parameter settings beyond predefined experimental conditions, whereas GRA remains confined to pre-tested configurations. This hybrid optimization approach effectively addresses research gaps by integrating ranking-based optimization (GRA) with predictive modelling (MLR), providing a more comprehensive and generalizable framework for optimizing FDM process parameters. The study highlights the potential of combining these techniques to improve mechanical performance, process efficiency, and predictive accuracy, making it highly relevant for advanced applications such as biomedical implants and high-performance industrial components.