A study of the track degradation process related to changes in railway traffic

Abstract: This licentiate thesis presents the results of a research project concerning railroad maintenance. After acquiring the state-of-the-art knowledge in relevant areas a prediction model for railroad track degradation has been developed. The model is named DeCoTrack (Degradation Cost Of Track). The research study has been performed in a close collaboration with track experts at Banverket. The model simulates changes in degradation rate of the track due to changes in traffic characteristics. The inputs include for example parameters such as axle load, annual tonnage, speed, the mix of vehicle types and vehicle maintenance conditions. Outputs from the model are both track life length and the estimated degradation cost. When developing the model, results from research studies reported mainly from Europe and North America were combined with classical mechanical engineering theories, and empirical data from the last 20 years of railway transports in Sweden. By establishing compatible interfaces between the different input sources, information became easy to adapt to the model which was gradually implemented into an easy-to-use software. The model DeCoTrack has been verified mainly based on an extensive literature survey trying to find other comparative models and to relate their output to that from DeCoTrack. Out of 900 abstracts, 40 research works were found of interest and studied in detail. They led to three existing models: TMPM/ITDM - Track Maintenance Planning Model / Integrated Track Degradation Model TRACS - Total Right of Way Analysis and Costing System The Damage Exponent Heavy Axle Load Analysis These models turned out to be more complex than DeCoTrack and therefore it is more difficult to implement and use these models in real situations. In addition these models are also mainly focused on the modelling of uniform freight traffic while DeCoTrack can model and describe the track degradation due to mixed traffic. A direct comparison of outputs from these models has been done on traffic data where quantitative data were available. The correlations between outputs from DeCoTrack and the above mentioned models were good. A search for more external research work on freight traffic leads undoubtedly to U.S. and the AAR/TTC test centre in Colorado. All of the three models above refer in some way to data from that centre. When it comes to high speed goods or passenger traffic and related maintenance issues, the major sources of knowledge can be found in Germany, France and Japan. Future research is suggested to concentrate on vehicle classification. The domestic situation in Sweden with deregulation in traffic and mixed traffic on same routes highly prioritise such a decision. From an international perspective, that effort might bring a useful complement to the other models. They are focused on uniform freight traffic.