Analysis of field data to develop rail wear prediction model

Rail wear rate is important for budgeting rail replacements and reducing operational risks. It acts as a performance indicator for rail-wheel lubrication. Executive judgement based on experience and historical data is current practice for ad-hoc decisions. Wear rate depend on operational conditions. These are train speed, axle load, rail-wheel material type, size and profile, track construction, characteristics of bogie type, Million Gross Tonnes (MGT), curvature, traffic type, weather and environmental conditions. Lubrication and rail grinding are generally used at curve sections. It results in balanced wear rate leads to controlled rolling contact fatigue and enhanced rail-wheel life. There is no international standard available for rail-wheel lubrication capable of accurately predicting rail-wheel wear. It is important to study the factors behind these and develop a model based on field data for predicting rail wear. This paper focuses on collection and analysis of field data for developing an effective rail wear prediction model.