The important parameters in all descriptions of the kinetics of crystallization behaviour are the differences in free energy Δgv between the liquid (glassy) phase and the various stable and metastable crystalline phases. The various quenching techniques can be subdivided into those in which the cooling rate is determined by the heat diffusivity and those in which the heat transfer coefficient is important. The temperature gradient and the solidification rates are different in the two cases. An additional difference is the nucleation rate which is dependent on the degree of undercooling.

A novel technique was used to measure directly the interface temperature and growth rate during solidification of pure gallium. The technique utilizes the Seebeck emf generated across two solid-liquid interfaces to determine the interface supercooling and an optical microscope to measure the solidification rates. The growth rate measurements were made in the range of about 0.08 to 1400 μm/s for 1.59 to 3.6°C interface supercoolings (ΔT) on the (111) interface of high purity gallium in glass capillaries. A plot of logv vs. 1/T for dislocation-free interfaces followed two straight lines, one for higher than about 1 μm/s growth rates and the other for the lower rates. The data at the lower rates correspond to the lateral growth mechanism for which a single nucleus forms and spreads across the interface before formation of the next nucleus, while those at the higher solidification rates correspond to the lateral growth by the multi-nucleation and spreading mechanism.

Three estimates of the solidification rates required to obtain a fully homogeneous structure during rapid solidification processing (RSP) have been made. One is given by the “absolute stability” criterion and another obtained from a new analysis for dendritic solidification. The third estimate, also derived from the above analysis, requires that “hypercooled” conditions be maintained after nucleation. A mechanism for the formation of “featureless” segregation-free zones during melt spinning and atomization processes is suggested and expressions for the critical cooling rate and thermal gradient required to produce such structures have been obtained.

The processes responsible for solidification, e.g., nucleation and growth, are influenced to various extents by the currently employed quenching techniques. The microstructures produced by these techniques are discussed and compared.

Small nickel-tin alloy samples (3.2 g) were bulk undercooled in a Pyrex-brand borosilicate glass medium using a levitation melter. Recalescence was directly observed utilizinq a rapid thermal measurement system (resnonse time less than 10 us) consisting of two silicon photodiodes, signal amplification system, and a digital storarie device.

Results of this research include: (1) measured undercoolings ranqed up to about 250 K; (2) both total solidification time and recalescence time become shorter with increasing initial undercooling; (3) in the case of hypoeutectic (Ni-25wt%Sn) alloy, a range of recalescence times was observed from 500 ms for a low undercooling (about 35 K) to 2 ms for a hiqh undercooling (about 230 K); (4) thermal profiles upon solidification of hypoeutectic comnosition alloys showed the existence of two distinct nucleation events; (5) although cooling rates before nucleation were generally low (<50 K/s), structures of fineness expected in rapid solidification processing were typically observed; (6) fineness of microstructure increases with increasing undercooling; and (7) uniform distribution of second phase was achieved by increasing undercoolina.

The metallurgical aspects of laser melted and laser alloyed surfaces are discussed. It is shown how varying the laser parameters can influence the microstructure and properties.

The effects of electron beam surface melting on a high strength Ni-Cr-Mo-V steel are first described with particular reference to the refined microstructure and the increased surface hardness. Similar experiments on an Fe80P13C7 alloy are described, in which surface amorphous layers were produced leading to hardnesses around 910VPN at the surface up to 1250VPN further in and falling to 650VPN in the as-cast crystalline substrate.

Metallic glass cladding of the steel was achieved by coating the steel surface with thin layers of powdered Fe80P13C7 alloy, and traversing with an electron beam sufficient to fuse the powder layer. Process variables including beam power and thickness of powder layer were examined and optimum conditions for glass formation in the surface layer were determined. These layers were examined by optical and electron microscopy, X-ray diffractometry and energy dispersive X-ray analysis. The microhardness variation with distance from the surface was also determined and found to be in the range 880–1150VPN.

The mechanism of ultrafast cooling (1012K sec−1) by picosecond pulsed laser irradiation is reviewed, and it is demonstrated that only partitionless transformations can occur. The glass formation range in a number of binary systems is reviewed, demonstrating that many glasses can be formed below the To-line. The formation of metastable crystalline alloys in the Fe-B, Ni-Nb, Cu-Co and Au-Co systems is reported.

Glass formation by ns and ps laser quenching is investigated for the systems Au-Ti, Co-Ti, Cr-Ti and Zr-Ti. Glassy phases are found in all but the last system. These results (as well as earlier ones) show little correlation with published rules to predict glass forming ability, but are consistent with the kinetic analysis presented. Simple Physical parameters related to glass forming ability are discussed. Comparison with results for Cr-Ti allows a determination of the nucleation time-lag.

Linear heat source surface melting studies have been carried out using a 1 to 15 kH oscillating beam incident on a steel substrate to produce a 3 to 35mm wide rapidly solidified strip of material at rates ranging from 1 to 100 cm.s−1. A simple relationship betweenmelt depth, power, and substrate velocity (V) has been determined at low values of V; limited results at higher velocities appear to follow the same relationship. A serendipitous discovery was the effect of prior surface homogenization treatment on the geometrical and microstructural uniformity of the subsequently formed rapidly solidified layers. This pre-treatment resulted in the formation of thin layers of uniform thickness under the most rapid processing condition (V = 100 cm.s−l). Omission of pre-treatnent resulted in irregular and sometimes discontinuous surface layers. In the case of a high speed steel of near peritectic composition (M2)pre-treatment resulted in the almost complete elimination of massive austenitic region within the dominant ferrite phase.

Electron beam glazing has been used to reprocess hardfacing deposits of conventional Stellite 6 alloy, and TiC, TiB2 particle strengthened Stellite 6. By varying the scanning mode and power density, a wide variety of microstructures have been developed, with significant increases in hardness. Single-pass line source scanning produces a distribution of moderately sized carbides (2–5μm) in a dendritic matrix, whereas multi-pass processing gives a more homogeneous refined microstructure with submicron carbides. Point source scanning, in contrast, produces the most refined microstructures. Effective control of the carbide distribution, including both fine and coarse carbides, may be the key to tailoring the microstructure for specific wear applications.

Influence of laserglaze on the microstructure and wear resistance of a cast nickel-base superalloy has been studied, which contains very high contents of refractory metals (Mo 33 wt%, W 17 wt%). After surface treatment by laserglaze the alloy structure changes extraordinarily. This change of microstructure enhances the alloy properties. Hardness of the laser melted region is evidently increased, and wear resistance and and corrosion resistance are also improved.

Amorphous phases are observed in laser glazed Fe82 Bl2C6, Fe 82 C18 alloys by Transmission Electron Microscopy. Differential Scanning Calorimetry is used to monitor the amorphous to crystalline transformation. Metastable crystalline structures are also found, such as e-phase in Fe82 C18 and Fe 70Mo10 C20 alloys and, respectively, very fine bainite and retained austenite.

Fourier series solutions have been obtained for the temperature distributions in melt spinning cooling belts and drums. These solutions have been used to determine drum and belt designs for producing 0.050 inch steel sheet. This paper discusses thermal design parameters for cooling drums used in melt spinning rapid solidification.

Although during melt spinning the ongoing interactive thermal and mass transfer processes are affected by both materials properties and process variables, a simplified resolution for this complex system can be obtained by first identifying and characterizing those substrate and melt material combinations that provide useful, steady state production performance. A previously described standardized falling droplet/inclined plane test procedure that simulates the dynamic wetting and solidification behavior during melt spinning is being used to generate a range of phenomenological information for combination of melt (droplet) and substrate (plane). Real differences in dynamic wetting/solidification patterns are being correlated with melt spinning performance. Phenomenological and process characteristics from some melt spinning experiments for representative material combinations have been analyzed to verify the validity of these projected correlations as well as to provide complementary information about the interrelation of melt puddle/substrate dynamics with product dimensions and quality.

The correlation between ribbon thickness and the length of the melt puddle residing on the surface of a melt-spinning wheel was established for a Ni-base superalloy. Since the melt puddle length defines the solidification time in which a ribbon with a certain thickness is formed, the above correlation allowed a direct derivation of the propagation velocity of the solid-liquid interface. The solidification rate V (mm/s) so obttined as a function of ribbon thickness S (mm) is V = 3.54S−1. Further, the above solidification correlation was analyzed using heat transfer considerations to yield information about the ribbon-wheel interfacial thermal conductance, the solid-liquid interfacial temperature, and the local cooling rate through the ribbon thickness. These thermal results are compared to those deduced from the secondary dendrite arm spacing measurements. Finally, there is a discussion on the ribbon microstructure based on our rapid solidification kinetic result.

A commercial aluminium alloy (3004) has been melt-spun on both an aluminium and a copper substrate. The meltspinning conditions have been varied in order to determine the optimum spinning parameteis for the apparatus used. The resulting ribbon has been examined by optical microscopy, transmission electron microscopy and anodization to show differences in grain structure and second phase distribution between the as-received rolled sheet and the melt-spun ribbon. X-ray diffractometry has been used to obtain information about the different intermetallic phases present in the sheet and ribbon, and peak heights of X-ray diffractometer traces have also been used to indicate textures present in the ribbons spun on aluminium and copper substrates.

A recently proposed model of shock consolidation of powders quantitatively predicts regimes of input energy and shock duration required to produce well-bonded compacts. A growing data base from shock experiments in which the shock wave and powder parameters of importance are controlled allows evaluation of the model.

Rapidly solidified crystalline AISI 9310, and microcrystalline Markomet 3.11, as well as amorphous Markomet 1064 and crystalline Mo powders, have been consolidated by shocks up to 2 μsec duration. The formation of amorphous layers on Marko 3.11 particle surfaces indicates that surface melting and rapid solidification occurred. Decreasing amounts of amorphous structure are retained in Marko 3.11 and 1064 powder compacts with increasing shock energies. Significant improvement in Mo particle bonding is achieved by reducing surface oxides prior to shock consolidation.

The Dynamic Compaction of powdered materials becomes a particularly attractive technique for consolidating hard materials, because of the ease with which high pressures are achieved, and for metastable materials, because of the possibility of avoiding large-scale, high temperature excursions. The shock-wave conditions necessary for achieving high densities and interparticle bonding are evaluated experimentally. The bonding and melting taking place can be identified by post-consolidation studies, and the inter-relation between compact integrity and melting is shown. An attempt is made to develop a “process window” analogous to that of explosive bonding, to describe the conditions necessary for consolidation. Finally, in view of the necessity for localisation of plastic flow during consolidation, an indication of the influence of material deformation parameters is given.

A range of Fe-Mn and Fe-Ni steels have been processed by melt extraction to give fibres and particulate. The consolidation and sintering performance of these alloys are presented. Densities of up to 95% can be achieved with hardnesses in the range 400–500 Hv while possessing grain sizes less than 20μm. Salient features of the sintered structures are described and it is concluded that sintering of pre-annealed short fibres compares favourably with a more conventional approach using powders.