ABSTRACT

Spray forming is an alternative to conventional metal-working technology for the production of material preforms or near-net-shape components. However, the non-uniform droplets and coupling of process parameters associated with gas-atomized spray not only make process control difficult, but also severely limit the range of attainable microstructures. For this reason, the uniform-droplet spray (UDS) forming process was developed. The uniform droplet size and uncoupled process parameters allow for simplified modeling and precise microstructural control. 

The microstructure evolution in the UDS process can be divided into three stages: droplet solidification in flight, droplet impact, and post-impact evolution. The droplet thermal states, characterized by the degree of undercooling in undercooled droplets or the liquid fraction and morphology of solids in partially solidified droplets, have a great influence on droplet impact behavior and post-impact microstructure evolution. It is the objective of this work to research how droplet solidification evolves during the UDS process. A droplet thermal model was developed to study the effects of various process parameters such as droplet charge, initial droplet velocity, and droplet size on droplet solidification. Experiments were also performed to investigate the effects on droplet solidification of flight distance, droplet size, and oxygen concentration using a Zn-20 wt% Sn alloy. The effects of droplet thermal state on the deposit microstructure in spray forming were studied using a Sn-5 wt% Pb alloy. 

The droplet thermal model assumes Newtonian cooling in the droplets and simultaneously computes the droplet flight trajectory and droplet heat transfer. The Sheil equation was incorporated to model solute redistribution in the droplet during solidification. Several cases were run to study the effects of initial droplet velocity, droplet charge, and droplet size on droplet cooling. 200 5m diameter droplets with three different initial velocities (3, 5, and 7 m/s) were used to study the initial velocity effects. The simulation results show that droplets with a higher initial velocity have a higher cooling rate; however, they appear to have a higher temperature and liquid fraction when collected at the same flight distance due a quicker flight. 200 5m diameter droplets with three different droplet charges (2.3 x 10-12 , 4.1 x 10-12 , and 5.8 x 10-12 Coulomb) were used to investigate the droplet charge effects. The simulation results show that droplet charge affects droplet cooling mainly by changing the spreading distance, i.e., the distance at which the heat transfer coefficient increases dramatically. For the droplet size effect study, 100, 200, 300 5m droplets with the same initial velocity (5 m/s) were investigated. The results show that droplet size is the most critical process parameter in controlling the droplet cooling rate. 

For the flight distance effect study, 288 5m droplets were collected using carbon steel substrates at every 0.05 m from 0.35 to 0.75 m and 181 5m droplets were collected from 0.15 to 0.55 m. Scanning electron microscopy revealed surface morphology and cross-sectional microstructures. The results showed that the 288 5m droplets solidified gradually without undercooling and the 181 5m droplets experienced about 110 K undercooling. The cross-section micrographs of the 288 5m droplet samples were image analyzed to determine the liquid fraction. These liquid fractions show good agreement with those derived from the simulation. For the droplet size effect study, 288, 245, 181, and 96 5m droplets were collected at the bottom of the chamber using an oil bath. The results show that the 288 5m droplets experienced virtually no undercooling with dendritic microstructure and surface nucleation. Three types of microstructures were observed for the 245 5m droplets. The first type is similar to the 288 5m droplet microstructure. About 28% of the 245 5m droplets collected belong to this category. The second type is characterized by a cellular structure nucleated within the droplet and dendritic structures solidified after recalescence. About 63% of the 245 5m droplets collected belong to this category. The third type is defined by a cellular structure nucleated on the droplet surface and dendritic structures solidified after recalescence. About 9% of the 245 5m droplets collected belong to this category. The 181 and 96 5m droplets were all undercooled and internally nucleated. The percentage of the cellular structure increases and the sizes of the cells and dendrites decrease when droplet size decreases. Kinetic competition between different catalysts for nucleation explains why the population of undercooled droplets, the degree of undercooling, and the tendency to nucleate internally increase as the droplet size decreases. For the oxygen effect experiment, 181 5m droplets were sprayed and collected using an oil bath with the chamber oxygen concentrations maintained at 5, 50, and 100 ppm. No significant difference in the cross-section microstructure is evident. The droplets are undercooled and internally nucleated. 

Seven experiments were performed to study the effects of the droplet thermal state and the substrate condition on the droplet microstructure by spraying droplets of 100% liquid with 85 K superheat, 100% liquid with no superheat, 70% liquid, and 40% liquid onto a substrate maintained at 426 and 446 K, respectively. Droplets with 100% liquid all produced epitaxial columnar microstructures. With 70% liquid droplets and a 446 K substrate, a fine, equiaxed, dense microstructure was produced. Porous structures resulted when 70% liquid droplets were deposited onto a 426 K substrate or when 40% liquid droplets were deposited onto a 446 K substrate. The epitaxial columnar microstructure resulted because nucleation in the molten splat was more difficult than the continuous growth of the columnar crystals into the newly added liquid layer. The equiaxed microstructure evolved mainly from randomly oriented crystals, which were originally present as dendrites in the droplets and were re-oriented upon impact. Porous structures resulted because the degree of droplet spreading was reduced either due to low liquid content in the droplets or due to high freezing rate resulting from a relatively cold substrate.

  Click Here For Full Thesis