The aim of the current study is to predict the crack appearance as a function of different parameters, such as material composition and thermomechanical path, applied to Continuous Casting (CC). The mean is to use a damage model that predicts the crack appearance at the grain scale (mesoscopic scale) and to validate it on peritectic steel grades in CC conditions.
At first this document contains the literature review in the field of fracture prediction in CC. First, the description of the peritectic steel family is done, giving its properties and explaining its different characteristics in terms of micrographic morphology and mechanism of precipitates appearance. Then, this chapter presents the different defects reported in the CC process and especially those detected in peritectic steels. The mechanisms thought to induce the fracture are enumerated through a literature review. Their role is explained by analyzing the experimental tests used to characterize the ductility loss. This phenomenon is also detailed in order to understand the mechanisms of cracking in steels. Finally, the different alloying elements present in the steels are enumerated and their respective effects on fracture are presented.
vspace{4mm} The damage model, its identification method and how to use it on CC case is then explained. The implementation of the fracture model in the Lagamine code is presented and explained. The different damage law parameters are listed and the experimental ways to identify their values are exposed. The specific treatment of the samples applied in order to have a representative experimental test reflecting the real process is described.
Finally this chapter contains the explanation of the construction of a numerical cell representative of the steel microstructure, and the way the constraints are applied to it.
The following part describes the experimental test campaign to determine the rheological parameters. Their results are analyzed in regard to the ductility loss at a critical range of temperature.
A macroscopic behavior rheology law is then identified. The mesoscopic cell and the experimental test campaign to determine the damage law parameters are defined according to the experimental results as well as the industrial information about the macroscopic stress, strain and temperature histories.
One part of this study is focused on the damage model parameters identification and analysis.
It describes the different choices required to obtain an accurate damage parameters set. The final set is then discussed trying to link the numerically identified values to the physics underlying the parameters and the chemical composition.
Last chapter presents the application on CC process. All the steps of the determination of the damage in the C case are presented in details in this chapter. The macroscopic results are analyzed. The data transfer from macroscopic to mesoscopic simulations is explained and the problems which have been experienced are presented there. Finally, qualitative results of the mesoscopic study are given and conclusions are drawn from the results for the further industrial use of the model.
The conclusion strikes the balance of the thesis work. Its originality, lacks and successes are reported. Perspectives over development of the model toward other steel grades or numerical improvement for both, the model and its identification methodology are presented.
![[Public/Internet]](/ETD-db/images/unrestricted.gif "[Public/Internet]")