Understanding Fatigue Mechanisms in Ancient Metallic Railway Bridges: A Microscopic Study of Puddled Iron

Abstract

Being the second oldest railway network in Europe, almost half of the metallic railway bridges in France were built before the 1900s. At the time, puddled iron was the building material of preference amongst most engineers, hot riveting was the preferred joining method.Considering the ageing of these constructions, as well as the increasing evolution of traffic in the network, the understanding of the behaviour of this type of structures to cyclic loadings is of utmost importance. However, high cycle fatigue characterization is an expensive and time consuming process. The use of a fast characterization method through the measure of self-heating of the specimens during cycling loadings is proposed and analysed.For homogeneous materials, like modern steel, the use of self-heating measurements to estimate fatigue-life related properties has shown good correlation to results obtained by the standard fatigue tests with the advantage of a lower time of characterization. In our case, non-metallic inclusions present in the puddled iron will have an important influence in the response of the material, and therefore the interpretation of the temperature's evolution needs a different approach.Due to its complexity, the integrity of the structure must be analysed from a multi-scaled point of view. For this study, we focused in the quantification of the influence that the population of non-metallic inclusions have on the mechanic and thermal behaviour of a puddled iron. For this purpose, a set of specimens previously tested by the self-heating method were prepared and observed using two different tools: a classic optical microscope and a scanning electron microscope. The relationship between geometric parameters of the inclusions found in the specimens and their temperature response to cyclic loadings was shown. The hypothesis of the presence of micro plasticity (dissipative phenomenon) near inclusions was also validated through SEM observations.