Casting of metals and alloys is concerned with the production of shaped castings (for example, an automotive cylinder block) or of simply shaped billets intended for subsequent metalworking into mill products (for example, into bars, plates, sheet, tubing, etc.).
In the production of shaped castings, a molten metal or alloy of desired composition is poured into a mold or die, which contains a cavity in the form of the component to be cast. The molten material is conveyed into the cavity via a gating system (Figure 1). Adjacent to the casting—often at the point or points (gates) where the gating system enters the casting cavity—are placed additional cavities called risers or feeders, the object of which is to provide a reservoir of molten metal. The function of these reservoirs is to compensate for shrinkage. (The majority of alloys shrink during solidification, grey cast-iron and type-metals being notable exceptions.)
Figure 1. A sand mold for a wheel which is to be cast in steel. The line shown indicates where the two halves of the pattern are parted to facilitate their withdrawal from the mold formed around them.
The mould is most often of silica sand bonded with a bentonite type clay, which has been activated with water. However, many shaped castings are made in metallic molds by processes called
Gravity die-casting, known as permanent molding in the USA.
Pressure die-casting, known as die-casting in the USA.
A hybrid of the above methods, low pressure die-casting, is widely used for the production of automobile wheels. As the nomenclature implies in the above metallic-mold processes, the delivery of the molten alloy may be by the action of gravity or pressure, the latter generated by a piston or by a gas.
Metal casting is most often chosen as a production method when the component concerned is of complex shape, although many mill products are initially cast into simple rectangular-sectioned billets by continuous casting machines (Figure 2). Such billets are subsequently worked plastically into the mill product concerned by rolling, drawing, extrusion, etc. In shaped-casting production, the component usually receives no further shaping operation other than finish machining.
Figure 2. A continous caster for producing steel billets for subsequent working into mill products. The mold enclosure is an internally-cooled hollow copper box type structure.
Although the beginnings of metal casting stretch back to the Chalcolithic period (5000-3000 B.C.), which preceded the Bronze.Age, attempts to develop science-based models of pouring and solidification only began in the present century. Early researchers were limited by the sheer complexity of the multidimensional coupled problems of fluid flow, heat transfer and solidification encountered. In spite of such limitations, a number of basic rules evolved, which permit practitioners to pour a wide variety of complex shapes of good reliability. For example, the US metal casting industry currently produces annually castings worth $ 23,000 million, which range from humble valves and fittings to sophisticated gas turbine and prosthetic hardware.
The production of defect-free shaped castings depends on the design of the gating and feeding arrangements. Prior to the middle of the present century, such arrangements were left almost entirely to skilled craftsmen. Ruddle (1956) reported that the earliest published scientific work appeared in the 1930s. He quoted Harding (1949) as stating that 50% of all rejected castings directly resulted from poor gating. Gating design should provide systems that deliver molten metal of good quality at the appropriate rate and velocity to the mold cavity. The metal should be free from entrained gases, slag and dross, etc., and should not undergo gross oxidation in pouring.
The propensity of most molten metals to form oxide films almost instantaneously complicates this design process and can lead to inconsistent mechanical properties if not taken into account. Campbell (1991) suggests that careful control of surface turbulence in the molten metal stream must be practiced. He suggests this may be accomplished by ensuring that the Weber Number is confined to the range of 0.2 to 0.8. This number relates the action of inertial pressure in the stream to the effects of surface tension, to unit:
The design of the feeding system owes much to Chvorinov (1940), who demonstrated that, assuming planar freezing, the solidification time for casting could be related to its volume to surface area ratio:
The mold constant, B, is a function of the thermophysical properties of the mold and the metal concerned, and the average temperature of the mold-metal interface during freezing, etc. Good thermal contact is assumed between the casting and the mold.
where h1s is the latent heat of fusion, cp1 is specific heat of the molten metal, k is the thermal diffusivity of the mold, and To denotes the initial temperature of the mold.
Conductivity, diffusivity and other thermophysical properties are assumed to remain constant during freezing. It is also assumed that no temperature loss occurs during pouring. The subscripts one and two denote, respectively, the metal and the mold while the letter i refers to the interface. Equation (2) has found extensive use in feeder design, where the freezing time of the feeder must always be greater than that of the casting. Chvorinov validated Eq. 2 by measuring the freezing times of a variety of low-carbon steel castings. It was later shown that the divergent heat flow, which occurs with curved surfaces and also at comers and edges of rectangular castings, should be taken into account. (See Berry, et al., 1959.)
Chvorinov also showed that the progress of freezing in planar solidification of a pure metal or other congruent freezing material is governed by:
The thickness of the frozen layer is thus related to the square root of the elapsed time. The so-called solidification constant is given by:
It is noted that the equation for skin thickening holds true only for minimal superheat in the molten metal.
At present, casting technology is poised at an important juncture in its history. This has been brought about by the impact of computer modeling. The earliest examples of computer modeling have employed the analogue computer. The contributions of Pehike and his co-workers (see Pehike et al. 1976) have provided a major insight on the possibilities of model experiments. However, the application of digital computer to the solution of metal casting problems has opened a new epoch for casting producers. Progress over the last decade using various computational techniques now renders possible a variety of options. Among them are the prediction of solidification profiles, the design of gating and feeding systems, the prediction of microstructure and finally the linkage of microstructure to mechanical properties.
Many of these developments are described in reports of conferences held periodically under the title The Modeling of Casting, Welding and Advanced Solidification Processes (1981 onwards).
Berry, J. T., Kondic, V., and Martin, G. (1959) Solidification Times of Simple Shaped Sand Castings in Sand Moulds, Trams. Amer. Foundrymen's Soc., 67:449-476.
Brody, H. D. and Apelian, D, (1981) Modeling of Casting and Welding Processes I, The Metallurgical Society of AIME, Warrendale, Pa., 1981 and subsequent volumes in series.
Campbell, J. (1991) Castings, Butterworth-Heinemann, Oxford, 1991.
Chvorinov, N. (1940) Theory of Casting Solidification, Giesserei, 27:(10)177-186, 11:201-208, No. 222-225.
Harding, E. W. (1949) Foundry Trade Jnl., 69:343.
Pehlke, R. D., Marrone, R. E., and Wilkes, J. O. (1976) Computer Simulation of Solidification, Amer. Foundrymen's Soc.
Ruddle, R. W. (1956) The Running and Gating of Sand Castings, Inst. of Metals, Monograph and Report Series No. 19, London.