Many authors have studied the solidification process of cast iron and the effect of grain boundaries in austenite shell on the growth of spheroidal graphite. But, the studies on the melting morphology of cast iron are rare and the effect of grain boundaries in austenite shell on the melting procedure of spheroidal graphite cast iron is unknown. Therefore, in this work, the melting procedure of cast iron and the role of grain boundaries in austenite shell on the melting of spheroidal graphite have been studied. The main results are summarized as follows. 1. In white cast iron containing silicon, melting initiates at the interface between austenite matrix and temper carbon which was decomposed from $Fe_3C$ during heating. 2. In gray cast iron, melting initiates at the boundary of eutectic cell where elements with low melting temperature are condensed. The dissolution of kish graphite is difficult. 3. In spheroidal graphite cast iron containing little phosphor, melting initiates at the outer region of austenite shell in which silicon is condensed. In this case, grain boundaries in austenite shell give little effect on the melting procedure of spheroidal graphite. 4. In spheroidal graphite cast iron containing phosphor above 0.3 wt%, its melting phenomena are changed with heating rate due to the existence of steadite. In this case, it can be concluded that liquid phase of steadite, which segregated on outer region of austenite shell, moves to spheroidal graphite-austenite interface along the grain boundaries in austenite shell.

The structures and mechanical properties of undirectionally solidified Sn-Zn eutectic alloys have been examined over the growth range 7mm/h to 6,000mm/h. The structures of unidirectionally solidified Sn-Zn eutectic alloys were primarily broken-lamellar at growth rates below 760mm/h and became fibrous at higher growth rates above that. At a growth rate 3,084mm/h the structures were fibrouse only. There is no dendrites at any growth rates, but occasionally ribbon-like morphologies were seen. The under cooling increased parabolically with growth rate increase. The hardness of specimes increased with growth rates increase but heat-treated specimen decreased after growth rates 760mm/h and became constant value. The effect of heat-treatment was good at lower growth rate. At room temperature tensile strength increased with growth rates up to R=990mm/h and then tensile strength became near-constant value. The effect of heat-treatment was well at lower growth rate.

In order to investigate the mold reaction characteristics of Mn steel and Cr steel castings individually, the mold reaction products were examined by scanning electron microscopy, electron prove microanalyzer and X-ray diffractometer. From this experiment, the results were summarized as follows: 1) The mold reaction depth increased with increase of Mn content, while it decreased with increase of Cr content. 2) Mold reaction depth decreased with Mn content at $1200^{\circ}C$. 3) Mn, among the reaction products, forms a low fusion silicate, Mn $O.SiO_2$ while Cr forms a stable oxide, $CrO_3$ which hindering the reaction between FeO and $SiO_2$ thus the formation of $FeO.SiO_2$ was depressed.

The melt of highly purified Zn was poured by top pouring process into the open green sand mold, that was made by using the 3-plane L-sections pattern. After skin was formed, the unsolified melt was poured out by rolling-over. The thicknesses of skin for each different of castings were investigated with one dimension. The results obtained and could be summerzed as follows: 1) While the 3-plant L-sections castings were solidifying in the mold, solidification blocks of different section modulus in the castings were formed, i.e. 1-dimension divergency block, 2-dimension heat divergency block, 3-dimension heat divergency block, 2-dimension heat convergency block, and 3-dimension heat convergency block. 2) When the chill plate was set up to the mold in order to change section modulus artificially, heat divergency blocks and heat convergency blocks according to the shape of chill plate were revealed.

Thin sections in castings solidify faster than thick sections. Solidification rates increase because of an increase in heat transfer from molten meltals to molds through these solidified thin sections. The cooling fin effect in pure zinc casting was studied about the solidification time, superheat, and fin size by adopting the pourout test. The following results could be obtained from the study: (1) The fin effect could be represented by the effective cooling surface area ($A_f/S$) increased. (2) The fin effect could be obtained as functions of solidification time, superheat, and fin size.