Densification of Al powder and Al–Cu matrix composite (reinforced with 15% Saffil short fibres) during axial cold compaction

Abstract

Pure Al, Alumix 13 (Al–4.5 wt.% Cu 0.5 Mg 0.2 Si) powders and Alumix13 reinforced with 15 wt.% Saffil short fibers were compacted up to 250–386 MPa in an axial die to study their compacting behavior. The final relative densities D were higher than 0.95 for all unreinforced powders and 0.86 for the composite. Different micromechanical and phenomenological models were used to fit density–pressure relations. Arzt model describes the powder compaction with good agreement up to D ~ 0.85. Kawakita equation results as a best linear fit for all tests, but its compressibility parameter b is not in agreement with the hardening behavior of the composite. Panelli and Ambrosio equation could describe the data fairly well qualitatively for all compactions tests, however, over a limited pressure range. Finally, Konopicky relationship turned out to be very useful and fitted the densification data of all three materials quite well. Its slope from linear P vs. ln (1/(1 − D)) plots, is related to the yield stress and characterizes the work hardening developed during plastic deformation while the density was increased. Microhardness values increase with the compacting pressure and such tendency agrees with the rising values of yield stresses, obtained by Konopicky.

Introduction

Aluminum-based Metal Matrix Composites (MMC) can be produced [1], [2] by powder metallurgy (PM) mixing Al powder with hard (usually brittle) discontinuous phases like ceramics, intermetallic particles [1] or short fibres. The traditional PM involves the cold consolidation by closed-die or by isostatic compaction followed by sintering. For both PM techniques the porosity (1 − D) is eliminated from the initial density D₀ [2]. The fraction of hard phase decreases the obtained density due to the reduction of the effective pressure over soft Al particles [2]. The ductile particles need extra plastic deformation to fill the volume gap between the hard reinforcements. Most PM/MMC works are focused on particles composites and only a few are on short fibres composites. The most widely test used to study this process is the axial die compaction due to its simplicity. However, some geometric parameters like pellet height/diameter ratio must be reduced to minimize friction effects. Several parameters are used such as: powder morphology, particles size, oxide content, friction between particles and mechanical properties which are involved in the densification control. Many phenomenological and micromechanical relations have been proposed to describe the plastic powders densification process. The basic developments relate relative density D with pressure P introducing characteristic powder parameters.

Artz [3] developed a micromechanical model for a random array of monosized spheres of radii R isostatically pressed during the initial stage [4] of densification. The spheres deformed suffering of a progressive flattening, thus bringing them closer and rising the density of the packing. The effective pressure between the particles is a continuous function of contact area A = 3 (D − D₀), in units of R², and coordination number Z = Z₀ + 9.5 (D − D₀), and it is described below, where σ_YS is the yield stress

$$\text{P}={\text{3 σ}}_{\text{YS}}A\text{ZD }/(4\pi {\text{R}}^2)\approx {\text{3 σ}}_{\text{YS}}{\text{D}}^2\left(\text{D}-{\text{ D}}_0\right)/\left(\text{1 }-{\text{ D}}_0\right)$$

Kawakita [5] model expresses the compaction pressure P and the degree of the volume reduction C to define two parameters characteristic of the powder a and b

$$\text{P }/\text{ C }={\text{ P V}}_0/\left({\text{V}}_0–\text{ V}\right)=\left[\text{1 }/\left(ab\right)\right]+\text{ P }/a$$

where a is the initial porosity equal to (1 − D₀), V and V₀ are the actual and initial volumes respectively, and b represents the ability to compact porous materials, sometimes called compressibility parameter, and in certain conditions b is related to 1/σ_YS.

Panelli and Ambrosio Filho [6] have proposed a phenomenological relation of the logarithm of porosity inverse (1/(1 − D)) as the result of square root function of pressure

$$k_{\text{PAN}}{\text{P}}^{0.5}=\text{ ln }\left[\text{1 }/\left(\text{1 }-\text{ D}\right)\right]+{\text{ B}}_{\text{PAN}}$$

where B_Pan = ln (1/(1 − D₀)) and k_PAN is related to the compaction capacity of the powder which is inversely proportional to the yield stress.

Earlier Konopicky and Shapiro [7] developed a grossly similar relationship to Panelli's function, but linearly dependent on the pressure.

$${\text{K}}_{\text{MK}}\text{P }=\text{ ln }\left[\text{1 }/\left(\text{1 }-\text{ D}\right)\right]+{\text{ B}}_{\text{MK}}$$

where B_MK is a constant depending on D₀. It is interesting to remark that Konopicky's empirical model can be obtained from a micromechanical model developed by Torre [8] taking into account a hollow sphere isostatically pressed. The sphere of relative density D is made of fully plastic material. The work showed that k is proportional to 3/(2 σ_YS). Later Hewitt and his co-workers [9] extended this model to strain hardening materials.

The aim of this work is to analyze the plastic densification during die compaction in the light of previous described models, for the three different powder compositions. Theoretical and experimental limitations and difficulties found in each model are discussed. Finally, measures of microhardness were made to compare and analyze the mechanical properties obtained with those models.