A basic three dimensional thermal model has been developed to simulate the solidification sequence for gravity die casting process. The finite difference method was used to analyze the solidification process during all the casting cycles. The prediction of die temperature in the quasi-steady state was analyzed by the boundary element method. The influence of die cooling on the heat flow in the cast/mold system was also investigated. Predictions of the computer simulation on temperature profiles and location of shrinkage defects were in good agreement with those observed in experimental die castings. Models of computer simulation which is developed by this work can be useful for the design and process control of die casting.

The effect of heat treating temperature and ${\alpha}$ phase In the ${\beta}$ phase matrix were investigated for ${\beta}-CuZnAl$ shape memory alloys by tension test, fatigue test, and shape memory effect test. After heat treatment at $677^{\circ}C$, $750^{\circ}C$, $800^{\circ}C$ and $850^{\circ}C$ for 10 min. respectively, static fracture stress(${\sigma}_f$), fatigue fracture stress(${\tau}_{max}$) at $10^6$ cycle, and elongation(${\epsilon}_f$) was $24.2kg/mm^2$, $17.21kg/mm^2$ and 11.8%, respectively. As heat treating temperature decreased, fracture surfaces of the specimens were changed from the intergranular to the transgranular fracture mode. Especially, the a phase precipitated in the ${\beta}$ phase matrix had an effect on crack propagation and the fracture surface was characterized by dimple-like pattern when crack propagated in transgranular cracking mode. Precipitation of the ${\alpha}$ phase in the ${\beta}$ phase matrix lowered the transformation temperature by $10^{\circ}C$, and about 2.5 vol.% ${\alpha}$ phase did not affect the shape memory effect examined by the bending test.

Solidification analysis was conducted on large steel castings and ingots by a modified finite difference method. Auto-mesh generation system was developed for improving the application of the computer analysis system to casting disign. Combined use of the prediction parameters, solidification time and temperature gradient, and an auxiliary parameter, shrinkage potential, were used to predict the formation of shrinkage defects. Several examples on the prediction of shrinkage cavity by this method were campared with the experimental reslts. It was found that a quantitative design of large steel castings and ingots can be made by the computer aided analysis of solidification process.

Microstructural characteristics of Al-3.63wt.%Pb hyper-monotectic alloy rapidly-solidified by melt spinning were examined. Possibility of forming a planar liquid -solid interface during rapid solidification of this alloy was also considered with a morphological stability theory, and a mechanism of forming banded structure observed at the bottom parts of melt-spinned specimens was considered as well. Application of the absolute stability criterion predicts the liquid-solid interface of the primary aluminium phase to be able to maintain a planar interface during the early stage of rapid solidification. Formation of banded structure was supposed to be resulted from the release of latent heat during solidification, which affect the stability of a planar liquid-solid interface.

SGCI(Spheroidal Graphite Cast Iron), CVGCI(CV Graphite Cast Iron) and FGCI(Flake Graphite Cast Iron) having different contents of Mn($0.25%{\sim}0.85%$) and Ni($0.3%{\sim}1.2%$) were produced, respectively. The thermal expansion and thermal conductivity of the cast iron were investigated in the temperature range of $50^{\circ}C{\sim}300^{\circ}C$. As the graphite nodularity of the cast iron increases, thermal expansion coefficient increases, thermal conductivity and electrical conductivity to thermal conductivity ratio decrease. The thermal expansion coefficient of the cast iron increases with increasing Mn content and decreases with increasing Ni content. The thermal conductivity of the cast iron decreases with increasing Mn and Ni contents.