A three-dimensional mathematical model of mould electromagnetic stirring (M-EMS) was established. Based on Maxwell's equations, the continuity equation and the momentum equation, the distribution characteristics of electromagnetic and flow fields with M-EMS were numerically simulated by the finite element software ANSYS and the finite volume software CFX. The influence of M-EMS on electromagnetic and flow fields was examined, and the process parameters of M-EMS were optimised by industrial plant trials. By the model verification, there was a good agreement between the calculated results and the measured data. The results indicate that the tangential electromagnetic force increases with the increasing current intensity, and increases at first and then decreases with the increasing current frequency. The tangential velocity increases with the increasing current intensity and current frequency (2-6 Hz). According to statistical results of the centre equiaxial crystal proportion and the macroscopic defects of round billet for different process parameters in industrial plant trials, the optimal process parameters of the M-EMS are as follows: the current intensity is 400 A, and the current frequency is 2 Hz.

A comprehensive mathematical model of the solidification structure during the process of electroslag remelting casting (ESRC) of low carbon martensite stainless steel ZG06Cr13Ni4Mo has been established. The change of metal pool profile and grain growth and the microstructure evolution process from the beginning to the steady stage of the ESRC process were investigated by using the moving boundary method and the coupled technology CAFE method (cellular automaton -finite element method). The transition from equiaxed grains at the lateral wall of the mould to columnar grains has been revealed. In addition, casting of this steel has been carried out and the microstructure of the ingot obtained after grinding and acid leaching. According to the comparison of the metal pool profile, morphology and growth direction of the dendrite and the secondary dendrite arm spacing (SDAS) between the experimental results and the simulation results, the validity of the model has been demonstrated, which can provide a favourable theoretical foundation to optimise the process parameters for the control of solidification structure of ESRC of low carbon martensite stainless steel ZG06Cr13Ni4Mo.

Electroslag remelting (ESR) hollow ingot process with T-shape current supplying mould is a new metallurgical technology. A mathematical model was developed to describe the interaction of multiple physical fields of this process for studying the process technology. Maxwell, Navier-Stokes and heat transfer equations have been adopted in the model to analyse the electromagnetic field, magnetic driven fluid flow, buoyancy driven flow and heat transfer using finite element software ANSYS. Moreover, the model has been verified through the metal pool depth measurements, which were obtained during remelting of 10 electrodes into W900/500 mm hollow ingots of P91 steel, with a slag composition of 50-60 wt-% CaF 2 , 10-20 wt-% CaO, 20-30 wt-% Al 2 O 3 , #8 wt-% SiO 2 . There was a good agreement between the calculated results and the measured results. The calculated results show that the distribution of current density, magnetic induction intensity, electromagnetic force, Joule heating, fluid flow and temperature are symmetric but not uniform due to the multi-electrode arrangement in two symmetric groups. Simulation of the ESR hollow ingot process will help to understand the new technology process and optimise operating parameters.

Mo macrosegregation was studied through the comparison of numerical simulation of the ingot pouring process and experiment on as-cast 500 kg high-Mo austenitic stainless steel ingot. The simulated results showed the evolution of temperature, melt velocity and the patterns of Mo macrosegregation, and revealed the effects of pouring temperature and cooling rate on macrosegregation. The predicted variation of Mo macrosegregation was compared with measurement values from an industrial ingot along the vertical centreline and horizontal direction. Severe normal and gravity segregation were observed. Although a basic agreement was obtained, the lack of a sufficiently fine numerical grid and the neglect of sedimentation for free equiaxed grains in the prediction brought about the absence of A-segregation and Vsegregation. Further investigation would be needed to perform this investigation. The predicted results also confirmed that Mo macrosegregation in the ingot could be effectively diminished by improving cooling rate and decreasing pouring temperature.

Isothermal aging treatment of 2507 super duplex stainless steel (SDSS2507) was conducted at 850°C after solution treatment at 1150°C. The characteristics of sigma (σ) precipitation kinetics in SDSS2507 were discussed and an improved JMA (Johnson-Mehl-Avrami) model had been developed. The results show that, the precipitation mechanism of σ phase in hot-rolled SDSS2507 was quite different to other ordinary DSSs. It can be found that σ phase precipitated via two mechanisms in this experiment. (1) The eutectoid decomposition δ → σ + γ 2 . It dominated the precipitation kinetics of σ phase in aging 0-25 min, and at this time the volume fraction of σ phase increased rapidly. (2) After 25 min, the precipitation of σ phase absolutely relied on the transformation of γ → σ which is controlled by the diffusion of Mo and Ni in γ. In this period, the precipitation rate of σ phase was significantly decreased. In this case, the separate modeling on the basis of each transformation mechanism was adopted and then an improved JMA model was developed for the precipitation kinetics of σ phase in the whole aging process. As a result, a good agreement between the experimental data and this developed JMA model can be obtained.