Wheel wear, contact noise, rail-climb derailment, and braking distance are directly affected by rail-wheel friction characteristics. More specifically, wheel wear, derailment tendency, and noise can be significantly reduced if friction can be moderated on turns. A friction coefficient of around 0.35 with positive friction characteristics is deemed to be ideal. While friction modifiers are proven to bring significant benefit to railways, there is currently no indigenous manufacturer in the country. Additionally, there is no easy way to decide “how much to use’’ and “where to use’’ the friction modifiers. This work aims to develop and characterise top of the rail friction modifiers with positive friction characteristics and use machine learning to decide “how much to use’’ and “where to use’’ the friction modifiers. The rail-wheel simulator built in Centre for Railway Research at IIT Kharagpur and ball on disc tester are being used to characterise the fabricated friction modifiers and test the machine learning models developed in this work. The work would have far reaching implications for India and the world in safe and economical operation of railways.
Accessibility Controls