Development of predictive material models

In this project, models are being developed to predict how the properties of steel change during and after manufacturing. The models will be used to predict how the properties change during welding.

The models being developed include material models, microstructure models, and diffusion models. Material models are used to predict how the material deforms under an applied load. These models can be more or less advanced depending on the required accuracy. In more advanced models, known as mechanism-based models, factors such as dislocation density and grain size are used to describe the material behaviour.

The material behaviour also depends on the microstructure of the steel. The microstructure is formed during steel manufacturing but also changes during processing, such as welding. Microstructure models are used to describe phase transformations that occur during heating and cooling of the material. The microstructure model can predict the phase changes for an arbitrary temperature history. The phase transformations occuring in welding can be predicted by using the temperature history obtained from the welding simulation developed in work package 1.6. The figure below shows how austenite, formed at a high temperature, is being decomposed into different structures such as ferrite, pearlite, bainite, and martensite when the steel is cooled down.

These microstructure models are then linked to the material model to provide a more accurate description of the material behaviour. The presence of hydrogen can also drastically affect the material behaviour, making diffusion models necessary to describe how hydrogen interacts within the steel. In the worst case, this can lead to crack formation due to hydrogen embrittlement.