This study developed a horizontal two-dimensional static elasto-plastic FEM analysis program for rail tracks. The validity of the developed program was confirmed by comparing it with an existing one-dimensional analysis model. The study also found that the program was able to analyze the behavior prior to track buckling and the buckling temperature with the same level of accuracy as the existing buckling analysis model. Furthermore, corner breakage in curves was a problem in the analysis using GIS railway line data. Therefore, a smoothing method was proposed in which corner breakage was replaced with a circular arc of a set radius, followed by a moving average, and confirmed its effectiveness.

In order to optimize maintenance and management of railway bridges and mitigate the risk to railway safety of scouring disasters, assessing risk of localized scouring at bridge piers is essential. This study proposes an indicator to represent risk of scouring by deriving dimensionless tractive forces acting on the riverbed from hydrological statistics during disasters. The proposed indicator is validated by comparing and organizing the results between disaster-affected and non-affected conditions. As a result, we found a consistent linear relationship between the normalized value and the embedment ratio when disasters occur.

One way to reduce maintenance costs for ballasted track is to replace existing lines with grouted ballasted track. In the construction of grouted ballasted track, the voids within the ballast bed are filled with a cement-based grout to form a pre-packed concrete bed. In this study, we proposed a design method for installing grouted ballasted tracks on an existing 40-year-old asphalt roadbed. We confirmed that the safety requirements for preventing structural and fatigue failure in both the asphalt roadbed and pre-packed concrete bed were satisfied.

Rail corrosion, particularly in humid environments, accounts for around 20% of broken rails in Japan. However, traditional methods for predicting the fatigue strength of corroded rails, such as using corrosion thickness and S-N curves, have limitations. The study involved collecting corroded rails, scanning their surface profiles, and applying finite element analysis to evaluate stress at fracturing points. A new method using stress and stress gradients was proposed to predict the number of load cycles required for corroded rails to fracture in bending fatigue tests, with an estimated range of 0.5 to 2 times.

Repainting of coated steel bridges is performed manually, and construction management is required to prevent variations in coating thickness. However, the equipment used on site to measure coating thickness only measures it locally, which makes it difficult to measure a large number of points over a wide area. The purpose of this paper is to determine the applicability of a coating thickness measurement method using an infrared camera. To this end, we report the results of the evaluation of the optical properties of coatings applied to steel plates with various surface profiles and the results of a study on the material composition of coatings suitable for this coating thickness measurement.

For some high-speed railway bridges, substantial cost and effort are required for in situ measurements during train passage to manage the effects of resonance. This paper presents a method for estimating the deflection of railway bridges in resonance based on onboard measurements of the track geometry. Based on a theoretical analysis that explicitly considers the dynamic response component, the proposed method estimates the deflection from three peak differences in the track geometry between the first and last cars. Numerical simulations and field measurements demonstrated that the proposed method accurately estimates the maximum deflection with an error margin of <0.5 mm.

This paper proposes a new method for evaluating the running safety of railway vehicles in the event of flange-climb derailment, focusing on the contact conditions between the wheel and rail. The method uses two key variables: lateral contact position and normalized transverse creepage, defined as the ratio of transverse creepage to the wheelset angle of attack. Through vehicle dynamics simulations under various running conditions, the relationship between these variables and the running safety was investigated. The results show that the transit domains for the loci of these variables differ significantly between derailment and non-derailment cases. This paper provides fundamental insights into the development of advanced safety assessment techniques for railway operations that take detailed wheel/rail contact dynamics into account.

In the railway industry, AI-based automation using camera images is increasingly being introduced to tasks that were previously performed visually. For future application to safety-critical operations such as autonomous driving, it is essential to investigate the causes of AI decision-making errors. This study proposes a method to trace train forward surveillance and estimate the causes of decision-making errors, such as missed detections, at three stages: input images, AI architecture, and training data. This paper describes the proposed cause-estimation method and reports verification results using evaluation images.

To support dispatchers during disruptions, this paper describes a proposal for a metric to evaluate the “number of late-arriving passengers”, the development of a train rescheduling simulator; and presents a method for analyzing and evaluating the impact of actual train rescheduling plans on passenger convenience. The evaluation metric counts passengers whose delay in arrival at their destination station exceeds a threshold, enabling the comparison of train rescheduling plans. The simulator predicts timetables under different train rescheduling actions. Using the evaluation metric, the analysis identifies when and where passenger convenience changes due to changes in train rescheduling actions. We describe these methods in detail.

Two types of impact compression tests were conducted to investigate the impact characteristics of a double-skin aluminum alloy carbody structure used in a conventional railway vehicle. One of the two tests was an overall compression test in which the entire end face of the specimen was compressed. The other was a local compression test in which part of its end face was compressed. As a result of the tests, buckling was the main cause of failure in the overall compression test, whereas shear fractures were the main cause of failure in the local compression test. Finite element analyses were conducted in which several material fracture criteria were used to compare the analysis results with the experimental results. Comparison between the analyses and the tests showed that the fracture criterion based on the ESI-Wilkins-Kamoulakos model was applicable in the analyses of the compression tests.

A comparative study of driver gaze allocation during automatic and manual operation on a demonstration train was conducted to verify whether forward gaze allocation during automatic operation is equivalent to that during normal manual operation. Forward gaze allocation was measured for three drivers on test trains and 15 drivers on revenue service trains, and data from 11 drivers were analyzed. The results showed that 10 of the 11 drivers had the same or greater forward gaze allocation during automatic operation as during manual operation.