As the most important structural materials, steels with good properties and low costs have been mostly used in modern industry. The properties of steel products are directly raltated with their final microstructures. As two kinds of the most commen physical metallurgical phemomena taking place in the processing and heat treatment of steels, recrystallization and austenite-to-ferrite transformation have been used widely in steel industry to refine the microstructure and then to improve the mechanical properties. Consequently, it is of great importantance to develop a quantitve relationship between the processing of steels and their microstructure in order to produce tailor-made materials by designing the composition and controlling the processing. In this dissertation, an alternative methodology of mesoscopic modeling is devolped to investigate the microstructural evolution during hot deformation in low carbon steels.
The austenite static recrysatllization following a hot deformation is simulated by coupling a cellular automaton (CA) model with a crystal plasticity finite element model (CPFEM). With a purpose to describe the influence of pre-deformation on the microstructural evolution and transformation kinetics of austenite recrystallization, the local stored energy of deformation is firstly simulated by CPFEM. These results are then mapped onto the CA lattice as the initial states for the subsequete recrystallization simulation. The effect of the inhomogeneous distribution of the stored energy on the recrystallization kinetics is addressed using this coupling method. The simulation results reveal that the space distribution of the mechanical responsed of plastic deformation is very heterogeneous. The inhomegeous distribution of the deformation results in non-uniform nucleation with a cluster pattern as well as a remarkable deviation in the recrystallization kinetics with the classical JMAK theory.
In order to investigate the austenite-to-ferrite transformation occurring in hot deformation, a synchronous intergration CA model which addresses the simultaneity between deformation and microstucture evolution is also developed. The dynamic strain-induced transformation (DSIT) in a low carbon steel is investigated by using this model. It permits the mechanisms for the refinement of the DSIT ferrite to be studied, together with the effects of some important parameters, e.g. the the prior austenite grain size (PAGS) and strain rate. The simulated results indicate that the refined ferrite grains derived from the DSIT are achieved by “unsaturated” ferrite nucleation and “limited” growth. Furthermore, ferrite dynamic recrystallization is occurring to subdivide the DSIT ferrite grains formed at the early stage of transformation and maintain their equiaxed morphology. The simulated microstructure in the DSIT is in good agreement with the quenched dual-phase microstructure observed in the optical micrograph, which consists of fine-grained ferrite and fine martensite islands/flakes dispersing in the matrix.
With a purpose to validate the industrial usability of the mesoscopic microstructure modeling, an intergrated CA model is developed to predict the microstructural evolution of the austenite recrystallization during multi-pass hot rolling of steel strip. In contrast to the macroscopic empirically-based model, the mesoscopic model enables both quantitative and topographic predictions of the microstructure evolution, rather than mere average microstructural features. The use of practical operational parameters inputted for each pass ensures a realistic model capable of being tested against actual microstructural results. The results are compared with the predictions by the in house software ROLLAN and are found to be in good agreement.
修改评论