Early stage consolidation mechanisms during hot isostatic pressing of Ti–6Al–4V powder compacts

Abstract

Mechanisms contributing to early stage compaction of metal powder compacts were identified in a series of Ti–6Al–4V powder compacts hot isostatically pressed to relative densities ranging from 71% to 100%. The partially dense compacts, consolidated from loose powder in thin-walled containers, were examined using optical microscopy of polished sections and stereo pair scanning electron microscopy of fracture surfaces. Relative particle motion, characterized by small relative movements of particles and clusters of particles, was found to contribute significantly to compaction over a broad range of relative densities (from 63% to 90%). A mechanism was identified by which the preferential deformation of small particles at large–small particle contacts enables rigid body motion of larger particles which, in turn, increases the relative density of the compact. Tensile fractures within the compact occurred between 82% and 90% relative density providing direct evidence of cooperative movement among clusters of particles. In comparisons with mechanistic powder compaction models, measurements of particle deformation, contact areas, and coordination numbers were found to substantiate the importance of the experimentally identified mechanisms.

Introduction

Metal powder consolidation comprises an initial process of random packing and subsequent processes of powder compaction and pore closure. Each of these processes involves various material transport mechanisms that eliminate porosity at various rates driven by thermal and mechanical processing conditions. The ability to predict consolidation behavior, net shape, and subsequently improve outcomes requires an accurate understanding of the important mechanisms, their interactions, and their absolute and relative rates. As in all material transport processes, the active mechanisms and their rates depend on structural geometry, material properties, and processing conditions.

Numerous efforts have been reported which attempt to predict the macroscopic response of porous materials subjected to externally applied mechanical stresses and thermal conditions1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19. Models based on rate-independent1, 2, 3, 4 and rate-dependent[5] continuum plasticity relate macroscopically applied stresses to the macroscopic deformation response of a porous body. Such approaches inherently require assumptions that the body responds as a continuum having specific average properties determined through mechanical testing of the specific system in specific loading conditions. For example, models based on the approach of Kuhn and Downey[2] require measurements obtained in compression experiments to fit yield surfaces which are assumed to have elliptical shapes that depend on relative density. Variations in local geometry within the compact20, 21 and interactions between individual locally active mechanisms, possibly including granular behavior, are not directly addressed. Therefore, the predictive ability of continuum models can depend on a close match between the calibration processing conditions (e.g. applied stress state, interparticle bond strength, etc.) and the prediction processing conditions.

Models based on a mechanistic approach6, 7, 8, 9, 10, 11, 12, 13 represent individual local mechanisms through unit models of deformation, and scale the unit model behavior to predict the overall macroscopic deformation response of the porous body as proposed by Arzt and coworkers6, 7, 8 and Helle et al.[9]. In these formulations, the unit models are simple geometric relationships between two particles and/or the material surrounding pores. The scaling computation which extrapolates macroscopic behavior from the unit models requires considerable simplifying assumptions with regard to the geometric structural complexity of a porous body. The resulting geometric representation of the compact is combined with detailed material property values such as diffusion and creep parameters to predict macroscopic behavior. Values for these detailed material properties can be difficult to measure or estimate, and can change as microstructure evolves during consolidation[19]. As was the case for continuum models, predictive success relies on extensive calibration experiments limiting the accuracy of extrapolations, particularly in the case of large powder compacts in which dramatically different boundary conditions (i.e. relative container dimensions and strength, etc.) and heat transfer effects (i.e. thermal conductivity through a porous body) can have a large influence.

Powder consolidation is controlled by microscopic deformation and material transport mechanisms, and both modeling approaches, continuum and mechanistic, require accurate macroscopic representations of those mechanisms. Although several variations and enhancements of consolidation models have been proposed14, 15, 16, 17, 18, 19, few experimental studies20, 21, 22, 23 have examined and attempted to characterize the actual mechanisms, particularly during the early stages of powder consolidation. The present study examines the mechanisms active during hot isostatic pressing and relates important newly identified mechanisms to analytical models.