Deformation Evolution Mechanism and Multi-Objective Collaborative Optimization of Large Crankshaft Machining

As a core component with weak stiffness in ship power systems, the large crankshaft undergoes dynamic coupling effects from multiple factors such as cutting force, centrifugal force, and clamping support during its turning process, making it difficult to accurately predict and control machining deformation. To address this question, this paper proposes a method for machining deformation prediction and process parameter optimization that integrates finite element simulation, empirical modeling, and multi-objective optimization. Taking the turning of the main shaft journal as the object, a three-dimensional cutting simulation model is established based on ABAQUS to obtain cutting force and deformation data under multi-factor coupling conditions. A high-precision deformation prediction model was constructed using an exponential empirical formula combined with nonlinear least squares method. The model's goodness of fit R2 was 0.9343, and the predicted results were in good agreement with the simulation values. Based on this, the NSGA-II genetic algorithm is introduced to minimize machining deformation, minimize total cutting force, and maximize material removal rate as optimization objectives. The Pareto optimal solution set is obtained to achieve the coordination and unity between machining accuracy, efficiency, and cutting load. By combining simulation, prediction, and optimization techniques, the inherent relationship between cutting parameters and deformation patterns can be clearly presented. This method makes the selection of process parameters more scientific and efficient, and also provides a replicable technical solution for controlling machining deformation of low stiffness components such as large crankshafts.