The superplastic flow of aluminium bronzes

Abstract

Conditions for Newtonian viscous flow are met when the deformation stress level is low and the power series expansion for the rate of shear deformation as a function of stress may be terminated at first order. The constant of proportionality in this expression, the fluidity, is the reciprocal of the dynamic viscosity. Conventionally, the magnitude of this parameter has been used to distinguish fluids and solids. For liquids of low viscosity (~ 0.1 poise) flow is Newtonian viscous even at the highest practical shear rate (~ 105 sec−1). Hot glasses are deformed to high neck free elongations at a viscosity of ~ 107 poise. In crystalline solids a single term constitutive equation may still be proposed, relating stress to deformation rate via an exponentm, where 0.4≤m≤0.9. Here either solid state diffusion directly or grain boundary sliding accommodated by diffusion must be relied upon to produce conditions for Newtonian viscosity (m = 1). Expressions for viscosity, similarly defined as the ratio of stress to corresponding rate of deformation, may be deduced on the basis of a transgranular or intergranular diffusion path. Such calculations yield a viscosity value of ~ 105 poise. Prerequisite conditions for superplastic flow to approach the Newtonian viscous limit, resulting in large neck free elongations, are that the material grainsize be small (1–10 µm), the material be deformed at intermediate strain-rates and the deformation temperature be in excess of half the absolute melting temperature. In the aluminium bronzes employed in the present study, conditions for maximum superplastic flow occur in the two phase field above the eutectoid transformation temperature (that is 700–900 °C). Here, unlike in most systems, constitution and temperature are related variables. For a specific alloy constitution, the material is characterised by an optimum temperature for superplastic flow which decreases with increasing Al content in the composition range 8.5–12% Al. This behaviour is discussed in relation to the distribution of phases.