ANN-NSGA-assisted investigation of tribo-performance of lead-free nickel boron coating

Abstract

In recent times electroless Ni-B coatings free of lead are being investigated with an aim to replace the lead stabilized bath. Traditional formulations of Ni-B coatings often rely on the use of heavy metal/lead-based stabilizers. However, the environmental and health concerns associated with these stabilizers have prompted the exploration of stabilizer-free formulations. The tribological behavior and wear mechanisms of heat-treated heavy metal-free Ni-B coatings were investigated systematically by varying load (10–50 N), speed (0.5–1.5 m/s) and distance (400–1000 m) in accordance with Taguchi's L27 orthogonal array. An artificial neural network (ANN) combined with genetic algorithm is proposed to investigate and optimize wear rate and coefficient of friction (COF) of heat-treated Ni-B coatings. Globular morphology was obtained for the coatings with crystalline peaks of Ni, Ni3B and Ni2B for coatings heat treated at 350°C for 1 h. The mean absolute percentage error of ANN model for wear rate and COF was 5.2846% and 2.0572%, respectively. The multiobjective optimization showed 30.325 N load, 0.50007 m/s speed and 759.2959 m distance corresponding to simultaneous minimum wear rate and COF of 0.005297 × 10−6 g N−1 m−1 and 0.25238 respectively. The worn surface of optimized conditions coatings exhibited oxidized flattened nodules and debris that act as load bearing area leading to lower wear rate and COF.