Investigation of Surface Integrity in High-Speed Machining of Hardened AISI 52100 Steel Using FEM and ANN–GA

Abstract

Surface integrity plays a critical role in determining the functional performance, fatigue life, wear resistance, and reliability of machined components used in aerospace, automotive, and bearing industries. The present research investigates the surface integrity characteristics during high-speed machining of hardened AISI 52100 steels through an integrated framework combining thermo-mechanical finite element simulation, experimental validation, statistical analysis, and artificial intelligence-based optimization techniques.
A thermo-mechanically coupled finite element model was developed in ABAQUS/Explicit using the Johnson–Cook constitutive model to simulate chip formation, cutting force generation, stress evolution, and temperature distribution during orthogonal machining. The effects of cutting speed (100–300 m/min), feed rate (0.05–0.25 mm/rev), and depth of cut (0.5–1.5 mm) on machining performance were investigated. Simulation results revealed that cutting force increased significantly with feed rate and depth of cut, while machining temperature increased with cutting speed. The maximum cutting force of 1450 N was observed at low speed and high feed conditions, whereas the maximum temperature of 940°C was recorded at a cutting speed of 300 m/min.
Experimental validation was performed using a CNC turning setup equipped with a piezoelectric dynamometer and infrared thermography system. The finite element predictions showed excellent agreement with experimental measurements, with an average prediction error of approximately 3%, confirming the reliability of the developed model. Analysis of Variance (ANOVA) indicated that feed rate was the most influential parameter affecting cutting force, contributing nearly 60% of the total variation, followed by depth of cut (20%) and cutting speed (12%).
Grey Relational Analysis (GRA) was employed for multi-objective optimization considering minimum cutting force, temperature, and surface roughness. The optimum machining condition was obtained at a cutting speed of 300 m/min, feed rate of 0.15 mm/rev, and depth of cut of 0.5 mm. Furthermore, a hybrid Artificial Neural Network–Genetic Algorithm (ANN–GA) model was developed to predict and optimize machining responses. The ANN model achieved a prediction accuracy of 97.2%, while the ANN–GA optimization resulted in an 18% reduction in cutting force, 22% reduction in machining temperature, and 15% improvement in surface finish.
The study demonstrates that the integration of finite element modeling, statistical optimization, and artificial intelligence techniques provides an efficient and reliable framework for predicting machining behavior and improving surface integrity. The developed methodology offers significant potential for industrial applications in smart manufacturing, process optimization, and sustainable machining of hardened steels.

Keywords