Assessment of the effectiveness of lubrication of Ti-6Al-4V titanium alloy sheets using radial basis function neural networks

The aim of the research presented in this article was to determine value friction coefficient using a simple tribological test and build an empirical model with use radial basis function artifi-cial neural networks. tests were carried out on specially designed simulator that allows sheet metal strip be drawn between two fixed dies. materials sheets Ti-6Al-4V titanium alloy thickness 0.5 mm. variable contact forces counter-samples rounded surfaces various lubrication conditions. Mineral oils bio-degradable addition boric acid (5 wt %) tested. Based results investigations, models built RBF artificial good properties network 2:2-35-1:1 confirmed by high determination R2 = 0.9984 low S.D. ratio equal 0.0557. It found COF highest for average values both nominal pressure kinematic viscosity. Over entire range pressures applied, SAE10W-40 engine oil ensured most effective reduction COF.