To increase the utilization of used tires, reduce carbon emissions and improve asphalt pavement performance, SBS/TB crumb rubber modified asphalt binder was designed, which was enhanced by SBS and terminal blend (TB) crumb rubber. SBS/TB crumb rubber modified asphalt binder was prepared by mixing 0 %, 10 %, and 15 % TB crumb rubber with 2 % and 3 % SBS, respectively. This study investigated the microstructure, thermal stability and rheological properties of SBS/TB crumb rubber modified asphalt binder. The Spearman correlation coefficient is introduced to analyze the correlation of microstructural, thermodynamic and rheological parameters. The results showed that SBS and TB crumb rubber were uniformly dispersed in asphalt binder without agglomeration phenomenon. In addition, the interaction between SBS and TB crumb rubber resulted in the formation of cross-links between the polymer and the asphalt binder, significantly improving the storage stability and the thermal stability of the modified asphalt binder. The pyrolysis mechanism of the modified asphalt binder is One-dimension diffusion or One-dimension phase boundary. With the addition of SBS and TB crumb rubber, the rheological, high-and-low temperature properties of modified asphalt binder are improved. Finally, microstructural, thermodynamic and rheological parameters have an extremely strong correlation by Spearman correlation coefficient analysis.

Timely detection and identification of rail breaks are crucial for safety and reliability of railway networks. This paper proposes a new deep learning-based approach using the daily monitoring data from in-service trains. A time-series generative adversarial network (TimeGAN) is employed to mitigate the problem of data imbalance and preserve the temporal dynamics for generating synthetic rail breaks. A feature-level attention-based bidirectional recurrent neural networks (AM-BRNN) is proposed to enhance feature extraction and capture two-direction dependencies in sequential data for accurate prediction. The proposed approach is implemented on a three-year dataset collected from a section of railroads (up to 350 km) in Australia. A real-life validation is carried out to evaluate the prediction performance of the proposed model, where historical data is used to train the model and future ’unseen’ rail breaks along the whole track section are used for testing. The results show that the model can successfully predict 9 out of 11 rail breaks three months ahead of time with a false prediction of non-break of 8.2%. Predicting rail breaks three months ahead of time will provide railroads enough time for maintenance planning. Given the prediction results, SHAP method is employed to perform cause analysis for individual rail break. The results of cause analysis can assist railroads to plan appropriate maintenance to prevent rail breaks.