Derailment Analysis for Prediction of Damage and Probability of Release for Novel Railroad Tank Car Designs

Abstract

Over the past 15 years, there has been substantial research performed to improve the safety of tank cars carrying hazardous materials. This has included a series of full-scale tank car impact tests and the development of impact/puncture models capable of evaluating the performance of existing and novel tank car designs under specific impact conditions. In parallel, statistical information was developed on tank car performance based on data from past accidents. However, such data cannot predict the performance of novel tank car designs involving new materials and configurations. This paper presents an innovative analytical framework to address this gap by simulating railroad tank cars behaviors under various train accident and rolling stock features. The finite element method-based simulation model is developed to generate impact forces on railcars after the initiation of train derailments. The results of the modeling were compared, validated, and calibrated with the real-world train derailment to assure the practicality of the analytical tool. This research is novel as it is the first quantitative, analytical methodology that can accurately and practically assess the physical behaviors of railcar movements and impact forces in train derailments. The results of this study provide new and important information for tank car performance in train derailments, which contributes to the literature in structural modeling of rail tank cars. It also specifies an approach to building a bridge between analytical methods and real-world scenarios and data for estimating tank car conditional probability of release (a frequently used metric for hazardous materials transportation safety).

Statistical Analysis of Train Derailment Severity for Unit Trains Versus Manifest Trains

“…One of the concerns identified by this result is that there are different types of behaviors in the derailment data and the variation cannot be represented well by this type of model. Thus, we analyze train derailment data through a physical model ( 33 ) that can model typical derailments. For example, the derailment data include a significant number (approximately 20%) of single-car derailments (severity = 1) with a wide range of derailment speeds.…”

“…Therefore, we developed a method to exclude these outliers and then used the TG model again to estimate derailment severity. Considering that we covered most of the derailment data and improved the TG model’s performance, we determined the threshold for excluding outliers by inferring from previous research ( 14 , 15 , 33 , 34 ) and several trials. Derailment data outside the range from 0.1 × derailment speed to 1.6 × speed are treated as outliers and are excluded from analysis (Figure 7).…”

Simplified Dynamic FEA Simulation for Post-Derailment Train-Behaviour Estimation through the Enhanced Input of Wheel–Ballast Friction Interactions

“…Moreover, 3D FEA simulations have been conducted for specific objectives, such as for verifying the operation stability between a track and vehicle, designing walls, and conducting detailed simulations of specific accident scenarios. However, because these simulations are performed for particular objectives, they cover a limited range of train behaviours following derailment in various situations [9][10][11][12][13][14]. Moreover, using full-scale simulations increases analysis times.…”

“…Moreover, using full-scale simulations increases analysis times. behaviours following derailment in various situations [9][10][11][12][13][14]. Moreover, using full-scale simulations increases analysis times.…”