The problem considered is the unidirectional growth of alloys close to monotectic composition and within the miscibility gap when the direction of gravity is down the temperature gradient (solidification upwards) or up the temperature gradient (solidification downwards). The systems Al-In and Al-Bi are taken as examples. With solidification upwards it is shown that bulk liquid composition adjusts to that of the monotectic while in the reverse situation, the bulk liquid gradually rises. In the former case it is possible to grow aligned fibrous structures of monotectic composition but in the latter the microstructures are irregular and globular.

The coalescence process has been studied in four different hyper-monotectic Zn-Bi alloys with 4–10wt % Bi . The alloys have been studied by isothermal treatments at a temperature just above the monotectic temperature and at different holding times. The alloys have been studied both at microgravity and at normal gravity. The change of the particle distribution as a function of composition and holding time has been investigated. The changing of the particle distribution has been compared with the theory of particle collision. The validity of the collision theory is investigated with respect to the theory of liquid phase sintering.

Unidirectional solidification experiments with monotectic and near monotectic Cu-Pb alloys have been performed with the growth direction both parallel and antiparallel to the gravity vector. It was found that a more regular composite structure was possible to achieve in the samples solidified parallel to the gravity vector. It was also found that the amount of lead rich phase, regularily incorporated in the structure, was less than one would expect theoretically. It is proposed that monotectic alloys can solidify to two different kinds of composite structures. One, which can be described by the theory of coupled growth of a rod eutecticum, and the other similar to the growth of primary rods.

The effects of thermal and solute gradient, gravity induced convection during vertical directional solidification, on longitudinal macrosegregation of Bi and Mn rich off-eutectic starting compositions, has been studied as a function of composition, growth velocity and gravity vector orientation. Since the morphology of these alloys is characterized by an aligned, rodlike permanent magnet composite when grown cooperatively, the magnetic properties were used to measure composition segregation and the transition from dendritic to composite growth. Severe macrosegregation was observed in all cases studied and the degree of convection inferred by modeling the observed composition segregation using a stagnant film approach. Morphological stability was found to follow a constitutional supercooling-type law for both Bi and Mn rich compositions.

We have made a complete calculation that extends the classical theory for crystal nucleation in pure melts to binary alloys. Using a regular solution model, we have developed approximate expressions for the free energy change upon crystallization as a function of solute concentration. They are used, together with model-based estimates of the interfacial tension, to calculate the nucleation frequency. The predictions of the theory for the maximum attainable undercooling are compared with existing experimental results for non-glass forming alloys. The theory is also applied to several easy glass-forming alloys (Pd-Si, Au-Si, Fe-B) for qualitative comparison with the present experimental experience on the ease of glass formation, and for assessment of the potential for formation of the glass in bulk.

This paper summarizes the results of a series of studies dealing with crystal growth from the melt and in particular, the growth of the narrow bandgap semiconductor PbSnTe by the Bridgman technique. Theoretical calculations have shown that for low Prandtl number fluids such as semiconductors, convection has no effect on the temperature field in the melt but has a significant effect on the compositional field. They also show that convective flow will exist in a Bridgman growth system on Earth, even in the “thermally stable” configuration. Measurements of the specific heat have been made up to temperatures of 722° C. The thermal diffusivity has been measured up to 1010° C which includes values for the melt. The interdiffusion coefficients for PbTe and SnTe in the melt have been determined. A technique has been developed and tested which provides a vacuum tight quartz ampoule with electrical feedthroughs for interface demarcation studies during the Bridgman growth. Analysis of crystals grown has indicated that the PbSnTe system is solutally unstable and the distribution of SnTe in the crystal, after an anomalous first to freeze region, follows that predicted by Pfann for convection controlled growth. Studies of the amount of supercooling in PbSnTe indicates that this may explain an anomalous composition distribution in the first to freeze region of an unseeded growth.

During dendritic solidification of metal alloys there normally occur temperature gradients in the liquid ahead of the solidification front. The gradient causes a natural convection ahead of the front. The natural convection causes a heat transfer from the interior of the liquid, which gives the possibility for crystals to grow ahead of the solidification front. The growth of these crystals is determined by the cooling rate and the kinetics of the solidification process. The paper will deal with the theory behind the growth of free crystals ahead of a solidification front. Some experimental results will be given. The effect of gravity on the growth process will. be discussed.

The effect of solid-liquid interface shape on lateral solute segregation during steady-state unidirectional solidification of a binary mixture is calculated under the assumption of no convection in the liquid. A finite element technique is employed to compute the concentration field in the liquid and the lateral segregation in the solid with a curved boundary between the liquid and solid phases. The computational model is constructed assuming knowledge of the solid-liquid interface shape; no attempt is made to relate this shape to the thermal field. The influence of interface curvature on the lateral compositional variation is investigated over a range of system parameters including diffusivity, growth speed, distribution coefficient, and geometric factors of the system. In the limiting case of a slightly non-planar interface, numerical results from the finite element technique are in good agreement with the analytical solutions of Coriell and Sekerka obtained by using linear theory. For the general case of highly non-planar interface shapes, the linear theory fails and the concentration field in the liquid as well as the lateral solute segregation in the solid can be calculated by using the finite element method.

Vacuum brazing under 1-g and microgravity (carried out on TEXUS II in preparation of an experiment for FSLP) conditions revealed the effect of various parameters (e.g. thermal history, hydrostatic pressure, buoyancy forces) on the microstructure of the joint and the gap filling phenomena.

The following results will be presented:

(a) Design of brazing specimens

(b) Variations of microstructure depending on gap width, gravity, and thermal history

(c) Flow patterns of molten braze as revealed by radioactive tracer experiments

(d) Evaluation of Plateau-type simulation experiments

The tin-15wt%lead and tin-3wt%bismuth alloys have been solidified in the low-gravity environment provided by the Space Processing Applications Rocket (SPAR), on the KC–135 airplane, and at high “g” levels in a centrifuge furnace. In each case the resultant cast structure was significantly different from that obtained in ground based experiments. Earlier low-gravity studies with the metal-model system NH4Cl-H2O presaged these results. This paper presents and discusses the influence of changes in the gravity force on the grain structure of these materials.

The thermal diffusivity of Ge has been measured over a temperature range from 300° C to 1010° C which includes values for the melt. Specific heat has been measured from room temperature to 727° C. Thermal conductivity has been calculated over the same temperature range as the diffusivity measurements. These data are reported along with the best values from the literature for the other parameters which are required to calculate the temperature and convective fields for the growth of germanium by the Bridgman method. These parameters include the specific heat, the viscosity, the emissivity, and the density as a function of temperature.