A novel evolutionary approach in modeling wear depth of laser engineering titanium coatings

Abstract

In this study, a new evolutionary approach, Akaike information criterion–based multi-gene genetic programming, is proposed for formulating the functional relationship of wear depth of the laser engineering titanium coatings. The carbon nanotube–reinforced titanium coatings were fabricated with 0, 10, 15, and 20 wt% carbon nanotubes. Six main input process variables such as specimen temperature, friction coefficient, contact potential, gap, carbon nanotube reinforcement composition (%), and sliding distance were considered. The laser cladding process was performed, and a total of 21,600 samples are collected, randomly divided into 17,280 and 4320 sets, and then trained and tested in the proposed algorithm. The performance of the proposed model is compared to that of the artificial neural network model. Statistical evaluation of the models concludes that the proposed model outperformed the artificial neural network. To validate the robustness of the proposed model, sensitivity and parametric analyses are conducted, and the impact of each input variable on the wear depth is studied. Analysis reveals that carbon nanotube reinforcement composition (%) plays a significant role in reducing the wear depth of the laser engineering titanium coatings.