Multi-objective tribological and energy optimization of an automatic valve lapping machine using a hybrid RSM NSGA-II approach

Optimizing valve seat reconditioning requires balancing sealing performance, surface integrity, energy consumption, and component wear within practical workshop constraints. This study presents the design, development, and multi-objective optimisation of a low-cost automatic valve lapping system using a hybrid Response Surface Methodology (RSM) and Non-dominated Sorting Genetic Algorithm II (NSGA-II) framework. A prototype automatic valve lapping rig was developed by integrating a DC-motor-driven spindle with adjustable spring loading and an Arduino-based control and data-acquisition system, enabling controlled variation of spindle speed (300–600 rpm) and axial load (60–140 N). Leakage time, surface roughness (Ra), electrical energy consumption, and valve wear volume were measured using a three-level factorial design. Quadratic response surface models with satisfactory statistical adequacy were established for all responses. The RSM models were employed in NSGA-II to maximise leakage time and minimise surface roughness, energy consumption, and wear, subject to practical operational constraints. The optimisation results reveal clear trade-offs between sealing quality, energy efficiency, and component life, and identify an optimal operating window of approximately 430–470 rpm and 90–110 N, providing a robust compromise solution and a practical operating map for workshop valve seat reconditioning.