A discussion about the limitations of the Eurocode’s high-speed load model for railway bridges

High-speed railway bridges are subjected to normative limitations concerning maximum permissible deck accelerations.For the design of these structures,the European norm EN 1991-2 introduces the high-speed load model(HSLM)—a set of point loads intended to include the effects of existing high-speed trains.Yet,the evolution of current trains and the recent development of new load models motivate a discussion regarding the limits of validity of the HSLM.For this study,a large number of randomly generated load models of articulated,conventional,and regular trains are tested and compared with the envelope of HSLM effects.For each type of train,two sets of 100,000 load models are considered:one abiding by the limits of the EN 1991-2 and another considering wider limits.This comparison is achieved using both a bridge-independent metric(train signatures)and dynamic analyses on a case study bridge(the Canelas bridge of the Portuguese Railway Network).For the latter,a methodology to decrease the computational cost of moving loads analysis is introduced.Results show that some theoretical load models constructed within the stipulated limits of the norm can lead to effects not covered by the HSLM.This is especially noted in conventional trains,where there is a relation with larger distances between centres of adjacent vehicle bogies.

A strategy for out-of-roundness damage wheels identification in railway vehicles based on sparse autoencoders

Wayside monitoring is a promising cost-effective alternative to predict damage in the rolling stock. The main goal of this work is to present an unsupervised methodology to identify out-of-roundness(OOR) damage wheels, such as wheel flats and polygonal wheels. This automatic damage identification algorithm is based on the vertical acceleration evaluated on the rails using a virtual wayside monitoring system and involves the application of a two-step procedure. The first step aims to define a confidence boundary by using(healthy) measurements evaluated on the rail constituting a baseline. The second step of the procedure involves classifying damage of predefined scenarios with different levels of severities. The proposed procedure is based on a machine learning methodology and includes the following stages:(1) data collection,(2) damage-sensitive feature extraction from the acquired responses using a neural network model, i.e., the sparse autoencoder(SAE),(3) data fusion based on the Mahalanobis distance, and(4) unsupervised feature classification by implementing outlier and cluster analysis. This procedure considers baseline responses at different speeds and rail irregularities to train the SAE model. Then, the trained SAE is capable to reconstruct test responses(not trained) allowing to compute the accumulative difference between original and reconstructed signals. The results prove the efficiency of the proposed approach in identifying the two most common types of OOR in railway wheels.