Papers by Author: Bai Cheng Liu

Abstract: Large steel ingots are the important material for the equipment manufacturing industry. It is still difficult to predict and control the macrosegregation in ingot. In this paper, the cooling curves at the surface of ingot and temperature variation of the mold were measured. The carbon distribution was measured through the local region dissection of ingot. Then, based on the definite the heat transfer coefficient at the interface of mold/ingot, a two-phase model with consideration of the motion of equiaxed grains is applied for the prediction of macrosegregation in 160-t steel ingot formed during the solidification. The results indicate that the heat transfer coefficient at the interface of mold/ingot decreases sharply after starting solidification and then varies slowly. Negative segregation at the bottom of ingot forms due to the interaction of solidification interface and equiaxed grains deposition during solidification. The positive segregation appears in the riser with thanks to the solidification shrinkage and the floating enriched solute. Finally, the results of the predicted and the measured are in good agreement.

Abstract: Plastic deformation and recrystallization of a Ni-based single crystal superalloy were experimentally investigated. Compression and Brinell Indentation were utilized to cause plastic deformation, and thereafter some deformed samples received heat treatment. Surface topography around the indent confirms the anisotropic plasticity of single crystal superalloys. The influential distance below the indent is much larger than that on the indent surface. Microstructural observation by Electron Back-scatter Diffraction (EBSD) shows that it is easier for nucleation and grain boundary migration in the dendritic arms. In addition, the recovery has almost no effect on preventing recrystallization for deformed samples with small plastic strains (around 5%).

Abstract: As a new method, liquid-metal cooling (LMC) process is used in manufacturing industrial gas turbines (IGT) blades. Numerical simulation is an effective way to investigate the grain’s growth and morphology, and optimize the process. In this paper, mathematical models for heat dynamic radiation and convection boundary of LMC process is established to simulate the temperature fields. Cellular Automaton (CA) method and KGT growth model are used to describe the nucleation and growth. Simulation results and experimental results are compared. The mushy zone and microstructure evolution are studied in detail. This study indicates that simulation and experimental results agree very well with each other. The withdrawal rate has an important influence on the shape of mushy zone and growth rate of the grain directly. A concave mushy zone is formed and the grain tends to convergent under an excessive high of withdrawal rate. But, the mushy zone has a convex shape and the grain is divergent under a smaller withdrawal rate. A variation withdrawal rate (from 2mm/min to 9mm/min) is found to obtain smooth mushy zone, which improves the parallelism of grain and produces high quality IGT blades.

Abstract: In the present investigation, a physically based numerical model was developed to predict the yield stress of Al-7Si-Mg cast alloy during processing. It covered the integrated unit step models of the physical metallurgy of solidification, solid-state of homogenization, and structural hardening of precipitation. The as-cast microstructure of Al-7Si-Mg alloy was calculated based on the cellular automaton method and the evolution of the precipitated phase during aging process was achieved by a precipitation kinetic model involved nucleation, growth and coarsening. The yield stress prediction was achieved by a strengthening model including the effects of as-cast microstructure, solution strengthening and precipitate hardening. The predictions of this model were verified by comparing with experimental measured yield stress which shows that this model is successfully applied to predict the yield stress evolution of Al-7Si-Mg cast alloy.

Abstract: Macrosegregation predictions have been performed for the alloy solidification in regular and irregular geometries. The continuum model is used to describe the mass, momentum, species, and energy transportation. Based on the same conservation equations, different discretization forms are derived for the orthogonal grids and non-orthogonal grids. Comparisons are made between the results with orthogonal grids and non-orthogonal grids for different cases. Firstly, simulations are conducted for the Pb-48%Sn alloy solidified in a rectangle cavity, i.e., the H-H benchmark. Then, simulations are applied to a 53t steel ingot. Prediction results are compared with concentration measurements along different transverse sections. Although both prediction results are in good agreements with the measurements or results in references, some differences can be observed due to the different boundary fitness. It is advised to adopt the non-orthogonal grids in macrosegregation simulations instead of the orthogonal grids for complex computation domains.

Abstract: The mathematical model for microstructure formation and porosity prediction during solidification process of SG iron casting was established and applied to a practical brake housing casting. Quantitative microstructure analysis of specimens machined from the castings was compared with the simulation, and the two results are in acceptable agreement on nodule counts and size, pearlite fractions and hardness. It is indicated that the model can calculate the fraction of ferrite and pearlite more accurately, and specially can reflect the effect of both under-cooling during solidification and the nodules formed in eutectic period on the pearlite content. The present porosity prediction was compared with those of a former method and commercial software, which leads to that the current methods used for porosity prediction should be investigated and improved further.

Abstract: Ni3Al based superalloy has recently been used for the single crystal gas turbine blade. The grain selection behavior in grain selector directly determines the casting’s final microstructure and properties. A mathematical model based on the modified CA-FD method was developed for the three-dimensional simulation of directional solidification process of Ni3Al based single crystal superalloy castings. The microstructure evolution was simulated with the modified Cellular Automaton method. The grain selection process in the grain selector and final microstructure of casting were simulated. The results indicate that the stray grain is easy to nucleate at the middle part of the pigtail because of the discontinuous mushy zones formation. This agrees with previous published experimental results. Based on simulated results, a newly designed grain selector with optimized geometry was proposed to avoid stray grains.

Abstract: A three-dimensional (3-D) cellular automaton (CA) model for simulating the dendrite morphology of cast Mg alloys has been developed. In the model a technique based on two sets of mesh is utilized to perform the simulation to reproduce the texture of Mg dendrites. The CA calculation is performed using a set of mesh that is defined by the hexagonal close-packed (HCP) crystal lattice, and other computations are carried out by using a cubic mesh. The two sets of mesh are coupled by using interpolation method. The kinetics of the solid-liquid interface is obtained directly by the difference between local equilibrium composition and local actual composition given by the solute transport equation. The model was used to simulate 3-D columnar growth of sixteen grains and 3-D equiaxed growth of a single dendrite of AZ91D alloy. Permanent mold castings of AZ91D alloy were produced and sampled for optical metallographic examinations, and the simulated results were compared with the metallographic results.

Abstract: Hydraulic turbine band castings are susceptible to deformation in heat treatment process if their cooling is not well controlled. The coupled analysis of forced air flow and heat transfer in normalizing process of a heavy turbine band runner casting with outer diameter of 8000 mm was carried out by using ANSYS software. The band undergoes significant deformation because of uneven cooling resulted by uneven air flow around it during normalizing. The forced air flow pattern is a key factor which influences the cooling uniformity and efficiency. It is optimized by adjusting the cooling fans’ orientation relative to the casting to form cyclone around it. Consequently, cooling uniformity is improved to avoid deformation.

Abstract: Directional solidified turbine blades of Ni-based superalloy are widely used as key parts of the gas turbine engines. The mechanical properties of the blade are greatly influenced by the final microstructure. In this paper, a mathematic model was proposed for the three dimensional simulation of microstructure evolution in directional solidification. Based on the thermo model of heat transfer, the grain growth within the blade and the microstructure morphology were simulated via a Cellular Automaton method. Validation experiments were carried out. The simulated cooling curves and microstructures corresponded well with the experimental results.