Microhardness mapping and the hardness-yield strength relationship in high-pressure diecast magnesium alloy AZ91

https://doi.org/10.1016/j.msea.2005.04.042Get rights and content

Abstract

Microhardness maps of cross-sections of high-pressure diecast test bars of AZ91 have been determined. Specimens with rectangular cross-sections, 1, 2 and 3 mm thick, or with a circular cross-section 6.4 mm in diameter, have been studied. The hardness is generally higher near the edges in all specimens, and more so near the corners of the rectangular specimens. The hardness at the center of the castings is generally lower, due to a coarser solidification microstructure and the concentration of porosity. The evidence confirms that the surface of the castings is harder than the core, but it does not support the concept of a skin with a sharp and definable boundary. This harder layer is irregular in hardness and depth and is not equally hard on opposite sides of the casting. The mean hardness obtained by integrating the microhardness maps over the entire cross-section increased with decreasing thickness of the bars, and was found to be in good correlation with each bar's yield strength.

Introduction

The yield strength1 [1], [2], [3], [4] of high-pressure diecast (hpdc) test bars of magnesium alloy AZ91 decreases monotonically with increasing cross-sectional thickness, as Fig. 1 shows. The casting thickness is a geometrical rather than a physical parameter, but its use as a reference for the strength, as well as for other mechanical properties such as ductility [1], [2], [3], [4], [5], [6], of hpdc material, is justified, as the thickness is the single most important factor determining the casting and solidification conditions. As hpdc products of AZ91 alloys are primarily used for thin-walled castings, such as in automobile components and mobile electronics, there is a practical interest in understanding the relationship between casting thickness and strength, which, as shown in the companion paper [7], is not well established. Surprisingly, however, systematic studies of the effect of casting thickness on the strength of hpdc alloy AZ91, or similar alloys, are few.

The high solidification rate inherent to the hpdc process gives rise to a unique non-homogenous microstructure in alloy AZ91. Of particular interest is the region adjacent to the casting surface, commonly referred to as the skin [2], [3], [8], [9], [10]. Mechanically removing the outer layer in thin (1 mm) castings produces a marked decrease of the yield strength [3], while similar experiments on thick (10 mm) castings [5] have little effect, and it has been suggested that the presence of the skin accounts for the higher yield strength in the thinner sections [3], [6], [7]. Weiler et al. [10] reported that the tensile behavior of AM60 alloy diecastings tends to match that of the skin.

The central region of the cross-section, or core, is characterised by larger grains of α-Mg phase [4], [11] and tends to be softer than the skin. It is generally thought that partial solidification occurs within the shot sleeve [12], and the liquid is inserted into the die cavity with up to 20% solid fraction. These solid particles often segregate to the center of the die cavity [13], especially in simple shapes, although this is not always the case [12]. Hence, the formation of these larger grains is independent of the cooling rates in the die cavity.

Regions of concentrated porosity are also characteristic of hpdc material. These vary in size and location, depending on the casting conditions and on the shape of the casting. They often exist close to the surface in thin castings, and deeper in the center of thicker ones. They may appear as bands or closed loops, or they can be densely scattered without much pattern [7], [14].

A skin seems to be a common microstructural characteristic of hpdc material, although a universally agreed definition is lacking. Existing definitions used in a number of previous studies are mostly based on microhardness measurements, but they tend to be mutually inconsistent. For instance, Sequeira et al. [3] have defined it for AZ91 as the depth at which the hardness value across the thickness is above 82 Hv. Hence, they found that the skin relative depth decreases with increasing casting thickness. In other instances, the skin has been defined as the region outside the porosity band [4], which results in an opposite conclusion, with skin depth increasing in thicker samples. While this definition is attractive, being based on an obvious microstructural feature, in practice, it can be ambiguous since castings of equal thickness can have varying porosity patterns or none at all. Thus, the very idea of a discrete skin with a definable boundary may be misleading [4].

On closer examination, it is evident that only limited attempts have been made to map the hardness values across the cross-section of bars of different thickness. In fact, the hardness profiles that have been reported to date have been limited to several line transverses through the middle of the specimen's cross-sections, while it seems obvious that as the thickness of the casting decreases, the regions closer to the corners become increasingly important. These need to be explored in detail if an accurate description of the skin is to be produced.

The objective of this study was to determine the hardness profiles of the entire cross-section of specimens of hpdc alloy AZ91 of different thickness and shape. It was hoped that by mapping the hardness across the full cross-section, a spatially accurate definition of the skin would be produced. This, in turn, would provide some more insight into the relationships between yield strength, hardness, microstructure and casting thickness discussed in the companion paper [7].

Section snippets

Materials

Commercial AZ91D alloy was used for the experiments. The alloy was cast to shape as tensile specimens using a 250 t Toshiba cold chamber diecasting machine. The rectangular cross-section specimens had a 50 mm gauge length, 10 mm width, and 1, 2, or 5 mm thickness. Cylindrical samples were also cast with a diameter of 6.4 mm. All materials were tested in the as-cast condition.

Tensile testing

Uniaxial tensile tests were carried out on a hard-beam testing machine fitted with hydraulic grips, at a crosshead speed of 2 

Tensile testing

Examples of stress–strain curves from the castings studied are presented in Fig. 3. There is a decrease in 0.2% proof stress (see crosses in Fig. 1) with increasing thickness of the rectangular specimens.

Cylindrical specimen

The microhardness values from measurements across the diameter of the cylindrical bar (see Fig. 2d) were averaged at each distance from the specimen center and plotted in Fig. 4. It shows that the regions close to the sample surface are generally harder, while the hardness gradually drops

Hardness behavior

Two sources of scatter affecting the hardness values can be easily identified: microporosity and segregated Mg17Al12. In the 2 and 5 mm thick samples, where concentrated porosity regions can be seen in Fig. 8b and 8c, the locus of the lower Hv values correlates with that of the two distinctive pore bands. Similar observations can be made for the round specimen, where the porosity ring in Fig. 8d can be matched with the major depressions that occur in its hardness profile (Fig. 4). On the other

Conclusions

The microhardness is higher at the narrow edges and corners of the specimens cross-section. The outer edge of the cross-section is generally harder than the core, but the harder layer is irregular in hardness and depth, and is not necessarily equally hard on opposite sides of the casting. A harder skin with a sharp and definable boundary could not be identified.

The higher microhardness values are observed in regions where the intermetallic solidification structure forms smaller cells and

Acknowledgements

The CRC for Cast Metals Manufacturing was established under the Australian Government Co-operative Research Centres Scheme. We thank CSIRO for a scholarship that supported ARP during his vacation employment from the University of British Columbia and we particularly thank E. Murray of CSIRO Manufacturing and Infrastructure Technology for his help in setting up the microhardness indentor apparatus.

References (17)

  • C.H. Cáceres et al.

    Mater. Sci. Eng. A

    (2005)
  • Z. Shan et al.

    Mater. Sci. Eng.

    (2003)
  • A.K. Dahle et al.

    J. Light Metals

    (2001)
  • G. Schindelbacher, R. Rösch, in: B.L. Mordike, K.U. Kainer, (Eds.), Magnesium Alloys and Their Applications,...
  • A. Stich et al.
  • W.P. Sequeira et al.
  • A.L. Bowles et al.
  • Y. Unigovski et al.
There are more references available in the full text version of this article.

Cited by (0)

View full text