Various melt spinning embodiments have been developed over the past 30 years and have been widely used, particularly at bench-scale for the production of metallic glasses. Much of the processing science has been developed 'as-needed' by the metallurgical community. Commercial scale-up has been limited, primarily because of the difficulty in maintaining good product quality. As yet, pilot-scale, and larger, single stage melt-spinning machines have not become industrially significant. The broad aim of this work is to develop a scientific basis for some technological objectives.

One obvious processing restriction is the presence of surface defects on the product. A common span-wise defect, referred to as the cross-wave, is related to processing parameters. The key physics underlying the defect are found to be natural oscillations of the liquid metal puddle, akin to the oscillations of a plucked sphere of liquid. This discovery, not only provides a key piece of information about the defect and ways to eliminate it, but also ties this highly applied process to a classical fluid dynamics problem.

The planar-flow melt spinning process is known to become unstable for various combinations of control parameters. As with any manufacturing process, a wider range of control and outputs is desirable. In this spirit, a theory, based on capillarity constraining the liquid metal, is developed which predicts a range of stable operating conditions. This theory also suggests a technology to extend the range of stable operation. This technology has been developed and successfully reduced to practice. The theory is further developed to provide more intimate details of the flow structure and pressure drops in the process.

Another technologically driven element of this work is the examination of the cooling rates in the process and their manipulation using local heat transfer disruptions on the liquid metal puddle. This work demonstrates how product properties may be manipulated on a sub-millimeter scale.

Finally a detailed description of the experimental operating procedure and hardware is presented, in order that future researchers may not have to 're-invent the wheel' regarding many subtle aspects of the process.