Z. Guo

SURFACE TENSION OF UNDERCOOLED LIQUID COBALT

This paper provides the results on experimentally measured and numerically predicted surface tensions of undercooled liquid cobalt. The experiments were performed by using the oscillation drop technique combined with electromagnetic levitation. The simulations are carried out with the Monte Carlo (MC) method, where the surface tension is predicted through calculations of the work of cohesion, and the interatomic interaction is described with an embedded-atom method. The maximum undercooling of the liquid cobalt is reached at 231 K (0.13Tm) in the experiment and 268 K (0.17Tm) in the simulation. The surface tension and its relationship with temperature obtained in the experiment and simulation are σexp = 1.93 − 0.000 33 (T − T m) N m−1 and σcal = 2.26 − 0.000 32 (T − T m) N m−1 respectively. The temperature dependence of the surface tension calculated from the MC simulation is in reasonable agreement with that measured in the experiment.

Thermophysical properties of undercooled liquid Au–Cu alloys from molecular dynamics simulations

The density and the specific heat of liquid Au–Cu alloy above and below the melting temperature are investigated in a wide composition range via constant temperature and constant pressure molecular dynamics simulations. The atomic interaction of the alloy is described with the embedded-atom method (EAM). The equilibrium melting temperature is evaluated from the change in the growth direction of a crystal–liquid sandwich structure under annealing. The simulated density of the Au–Cu alloy increases linearly with decrease of the temperature, whereas the specific heat remains constant over the entire temperature range of 900–1900 K. The excess volume is calculated according to the predicted density of Au–Cu alloy. The negative value of the excess volume and the exponential concentration dependence of the specific heat indicate that the Neumann–Kopp rule does not apply to the Au–Cu binary alloy system.

Two- and three- dimensional studies of dendritic morphology in magnesium alloy by means of synchrotron X-ray microtomography and cellular automaton modelling

Magnesium is the lightest structural material. As one of the dominant microstructure features, dendritic pattern determines the mechanical behaviour and performance of magnesium alloys. Dendritic topological observation was carried out on Mg-based alloy using synchrotron X-ray micro-tomography and the microstructure pattern of α-Mg dendrite was obtained. It was found that the α-Mg dendrite grew with eighteen primary stems, of which six lay in the (0001) basal plane, and the other twelve in the (1010) plane. An according numerical model based on the cellular automata method was developed. By defining a specific capturing functional mechanism, simulation of α-Mg dendrite in 3-D with eighteen branches was successfully achieved. The simulation results show that the model could reasonably describe the evolution of the dendritic microstructure and the subsequent dendrite morphology agrees well with that observed in the synchrotron X-ray tomography experiment.

Modelling and simulation for die casting mould filling process using Cartesian cut cell approach

Abstract A three-dimensional simulation model based on the Cartesian cut cell approach was developed to improve the computational accuracy and grid configuration at the domain boundary for die casting mould filling process. The finite volume method and Gaussian transformation were used to solve the governing equations for cut cells at the boundaries, while the finite difference method was used for internal orthogonal cells. The surface areas of the cut cells of all of the topological structures were sorted and calculated to discretise the equations. The model was verified by simulating the filling process in two die casting moulds. Results of the numerical simulation agreed well with those of water analogy experiments. By matching the physical boundary with the cut cell approach, the proposed model significantly improved the accuracy of the simulation as well as the smoothness of the domain boundary description.