Instability Phenomena in Materials Science and Manufacturing

Louis G. Hector, Alcoa Technical Center, Alcoa

Most metal manufacturing processes are inherently unstable due to interactions between processing conditions and material properties. These instabilities often manifest themselves as a loss of control over a manufacturing process (e.g. chatter in rolling), or the development of undesired geometrical features and material properties (e.g. ingot cracking). In some instances, the instabilities die out (we refer to these as practical instabilities), and in others, the instabilites force process termination due to total loss of control over the process (we refer to these as mathematical instabilities). Industrial America has invested an enormous amount of money on novel control methodologies for these and a multitude of other problems. In many instances, however, materials science-related instabilities in manufacturing processes continue to rear their ugly heads despite demonstrated control methodologies. At the present time, there is a great need for mathematical models of the complex physical phenomena that lead to instabilities in manufacturing processes. These models must necessarily weave together key principles from a variety of subjects, which include (but are not limited to), heat transfer, solid mechanics, fluid mechanics, metal and alloy physics, and materials science.

This seminar will focus on two practical instabilities in metal manufacturing. The first instability occurs in metal rolling and pertains to the onset of adhesive metal transfer as reduction ratios are increased. A methodology for controlling this problem which is centered around precision texturing of the roll surfaces will be discussed from an experimental standpoint. The second instability occurs in metal casting and pertains to a mechanism for the onset of undulatory growth of a solidifying shell. This problem will be discussed with a mathematical model of the phenomena which accounts for irregular contraction of the solidfying shell and the location gap nucleation along the mold-shell interface. Planned extensions of the present model formulation will be discussed.