Modeling of phase transformations and cyclic plasticity in pearlitic steels

Licentiate thesis, 2021

The low rolling resistance in railway transportation is a key factor for its high efficiency, but it comes at the price of very high contact pressure between the rail and wheel. Due to the high stresses and the large number of load cycles during railway operation, both the rail and the wheel can be susceptible to fatigue crack initiation and propagation. Furthermore, thermal loads due to frictional heating can generate high temperature in the material surface layers leading to gradual or drastic changes in material behavior. Other events that may cause such high temperatures include welding or grinding of the rail.

In this thesis, a modeling framework for phase transformations and cyclic plasticity of pearlitic steel is developed. The framework allows for modeling of the thermo-mechanical behavior of the individual phases. The implemented phase transformation kinetics in heating events include austenitization and possible tempering of martensite, and in cooling events the formation of pearlite, bainite, and martensite. Cyclic plasticity is incorporated in the model by using a Chaboche plasticity model with a von Mises yield function, non-linear isotropic and kinematic hardening. The capability of the modeling framework is demonstrated by studying the development of residual stresses during a double wheel flat scenario on the tread of a railway wheel followed by rolling contact loadings.

Further, the modeling framework is extended and improved by accounting for transformation induced plasticity (TRIP). To further improve the model, the influence of the choice of homogenization method is evaluated. Four methods are considered; iso-strain, iso-stress, self-consistent method and the linear mixture rule. These show different behavior during the multi-phase stages in simulations of a laser heated rail surface, which in turn affects the residual stress states. Although the conclusions are not entirely clear, comparison with experimental data indicates that the iso-strain and the self-consistent method are the most promising, with a slight advantage to the latter.

The thesis also presents an attempt at using the developed material model and the linear mixture rule in a simple butt-weld simulation, which can be seen as a first step towards simulation of repair welding of rails. The simulation includes a moving heat source and continuous addition of filler material. Preliminary results show that residual stress fields found for similar examples in the literature can be reproduced. Therefore, it is believed that the simulation methodologies developed in Papers A and B can be used as a basis for future developments towards simulations of repair welding of rails.