This study examines the mechanical and tribological performance of pulse-deposited multilayer Co–P coatings on low-carbon steel, combining experimental characterization with AI-based predictive modeling. Multilayer structures containing 64–1600 layers were fabricated via pulse electrodeposition using alternating 10% and 90% duty cycles to precisely control composition and thickness. FESEM/EDS, XRD, microhardness, fracture toughness, and ball-on-disc tests revealed marked improvements in hardness (up to 840 HV), reduced friction (COF: 0.93–0.46), and significantly enhanced wear resistance (minimum mass loss: 1.5×10 -3 g). To extend experimental insights, an Adaptive Neuro-Fuzzy Inference System (ANFIS) accurately predicted mass loss and identified load as the dominant wear-controlling parameter, with layer thickness and layer count exerting secondary effects. A Genetic Algorithm (GA) optimization further suggested near-optimal conditions (14.66 N load, ∼1590 layers, ∼592 nm thickness) achieving the same minimal mass loss observed experimentally. Overall, the integrated AI framework not only reproduced experimental behavior but also provided predictive and optimization capability, offering a practical decision-support tool for designing advanced wear-resistant coatings.
Ahmadi et al. (Wed,) studied this question.