Elsevier

Electrochimica Acta

Volume 56, Issue 4, 15 January 2011, Pages 1729-1736
Electrochimica Acta

Effect of grain size on corrosion of high purity aluminium

https://doi.org/10.1016/j.electacta.2010.09.023Get rights and content

Abstract

A complete understanding of how grain refinement, grain size, and processing affect the corrosion resistance of different alloys has not yet been fully developed. Determining a definitive ‘grain size–corrosion resistance’ relationship, if one exists, is inherently complex as the processing needed to achieve grain refinement also imparts other changes to the microstructure (such as texture, internal stress, and impurity segregation). This work evaluates how variation in grain size and processing impact the corrosion resistance of high purity aluminium. Aluminium samples with a range of grain sizes, from ∼100 μm to ∼2000 μm, were produced using different processing routes, including cold rolling, cryo rolling, equal channel angular pressing, and surface mechanical attrition treatment. Evaluation of all the samples studied revealed a tendency for corrosion rate to decrease as grain size decreases. This suggests that a Hall–Petch type relationship may exist for corrosion rate and grain size. This phenomenon, discussed in the context of grain refinement and processing, reveals several interesting and fundamental relationships.

Introduction

Grain boundaries have distinct properties relative to bulk material in terms of atomic coordination, reactivity, and diffusion rates. The proportion of atoms that lie at inter-crystalline regions as a total of surface area can become significant as grain size decreases [1], [2]. Consequently, it is not unreasonable to expect surfaces with relatively high grain boundary densities to exhibit different electrochemical behaviour from coarser-grained surfaces with low grain boundary densities. ‘Grain boundary engineering’ through grain refinement and various severe plastic deformation (SPD) techniques has been shown to be an effective way to improve mechanical properties such as strength and wear resistance [3], [4], [5], [6], [7], [8], [9], [10], [11]. Additionally, manipulation of grain boundary character has been used to enhance corrosion resistance of some materials. A classic example is improved intergranular corrosion resistance of Alloy 600 tubing, used in nuclear power production, through development of an increased quantity of low Σ coincident site lattice boundaries [7], [12]. Furthermore, different thermo-mechanical processing routes are known to have an important effect on the corrosion response of a material [13], [14], [15], [16], [17]. Despite these studies, which show that both mechanical and corrosion properties can simultaneously be altered through grain boundary modification and processing route, few (if any) efforts have been made to improve corrosion rates through grain size adjustment. Typically, improvements in corrosion resistance are achieved through bulk alloying or the use of various claddings and or coatings. As such, working towards materials with maximum corrosion resistance strictly by alteration of surface reactivity through grain size variation is somewhat novel. While alloying fundamentally changes the bulk material chemistry and coatings and claddings are prone to defects, grain size adjustments could be used to improve corrosion resistance without fundamentally altering the bulk composition (perhaps leading to decreased material costs).

The Hall–Petch relation reveals that yield strength is inversely proportional to grain size. However, an analogous relationship, if one exists, between grain size and corrosion rate has yet to be identified and is an area worthy of exploratory work. There have been a number of studies on different materials that involve grain size variation and corrosion performance, but as detailed in a recent review there is little consensus as to an all-encompassing effect applicable across different materials or even within similar alloy classes [17]. For instance, relatively little has been published directly or indirectly on the effect grain size variation has on the corrosion of aluminium (Al) and Al alloys [5], [6], [18], [19], [20], [21], [22]. Of studies that consider grain size and Al or Al alloys, the majority suggest that as grain size decreases corrosion resistance improves. Improved resistance generally is attributed to an ability of high grain boundary density surfaces to passivate more readily or to the physical breakdown of second phase intermetallic particles [5], [6], [20], [21]. However, fine-grained Al has also been reported to be more susceptible to corrosion than coarser-grained control specimens [18], [19]. Generally, a consensus within the literature does not exist which is not altogether surprising for a number of reasons. First, studies of grain size effects on corrosion are inherently difficult because any processing and or alloy additions used to achieve grain refinement may impart physical or chemical changes to the material in addition to the intentional grain size modification. Such secondary consequences from thermo-mechanical processing include the development of texture, internal stresses, and segregation of alloying additions to grain boundaries, each which may have an impact on electrochemical response. As a result, general interpretation of the literature is complicated as the effect of grain refinement is difficult to decouple from other processing induced effects. More generally, the relationship between grain size and corrosion resistance is likely environment specific thus studies in different electrolytes may produce opposing results. However, within more specific groupings of materials, processing routes, and environmental families, the existence of such relations seems plausible.

Different processing paths that can be used to achieve grain-refined microstructures include rolling operations and SPD techniques such as surface mechanical attrition (SMAT) and equal channel angular pressing (ECAP) [8], [9], [23]. SMAT is a multi-directional peening process which induces SPD and grain refinement in a thin surface layer [9], [10]. The ECAP process involves pressing a billet of material through a sharp bend in a die which induces SPD and grain refinement [8], [11]; a detailed description of ECAP in the context of Al can be found in [8]. The ECAP process itself has been noted to both increase [3], [4], [5], [13], [14], [22], [24], [25] and decrease [26], [27], [28] corrosion resistance. Conversely, SMAT has been reported to accelerate corrosion, posited to be from the introduction of internal stresses and defects at which attack may initiate [13], [14], [29].

The aim of this work is to investigate the role different processing routes play in corrosion performance, and to compare the relative effect from grain refinement to that of processing. More specifically, this is done explicitly for high purity Al with the long-term view to develop a fundamental understanding of the principles underpinning the relation for Al alloys.

Section snippets

Specimen preparation

Pure Al samples with a range of grain sizes were produced using different processing routes in combination with post-processing heat treatments. The parent material for all processed specimens was a single cast ingot of ultra high purity Al (99.9999, Alpha Aesar) with grains on the order of 1 cm in diameter. An Al wire (99.999, Alpha Aesar) was also evaluated for comparison. High purity Al was chosen for study to help minimize secondary processing effects (such as segregation of solute to grain

Processing to create a dispersion of grain sizes

Fig. 1 shows sample optical micrographs of grain structures from an ingot in the as-received condition and after ECAP, SMAT, cold rolling, and cryo rolling operations. All processing methods were found to induce grain refinement relative to the as-cast condition, which had grains sizes on the order of 1 cm, although the degree of refinement and shape of grains varied. After 1 ECAP pass the microstructure was observed to have undergone some refinement, particularly near the edge of the specimen,

Conclusions

  • Various processing routes were able to produce a distribution of grain sizes for specimens cut from the same starting ingot.

  • In neutral NaCl solutions, it was shown that the corrosion rate tended to decrease with decreasing grain size. This relationship has not been previously reported explicitly, and is of great significance if the relationship is generally applicable outside the realm of the system studied herein.

  • Decreasing grain size resulted in an ennoblement of Ecorr and decreased jcorr.

Acknowledgements

The Australian Research Council (Centre of Excellence for Design in Light Metals) and Victorian State Government for the establishment of the Victorian Facility for Light Metals Surface Technology are gratefully acknowledged. We would like to thank A/Prof. MingXing Zhang (University of Queensland) with the SMAT processing and A/Prof. Chris Davies (Monash University) for discussions. We also thank Silvio Mattievich, York Han, and Daniel Curtis (all of Monash University) with rolling,

References (31)

  • G. Palumbo et al.

    Scripta Metallurgica et Materialia

    (1990)
  • M.-K. Chung et al.

    Materials Science and Engineering A

    (2004)
  • R.Z. Valiev et al.

    Progress in Materials Science

    (2006)
  • H.W. Zhang et al.

    Acta Materialia

    (2003)
  • H.Q. Sun et al.

    Acta Materialia

    (2007)
  • A. Yamashita et al.

    Materials Science and Engineering A

    (2001)
  • P. Lin et al.

    Scripta Metallurgica et Materialia

    (1995)
  • W.R. Osorio et al.

    Materials Science and Engineering A

    (2005)
  • T.C. Tsai et al.

    Materials Science and Engineering A

    (1997)
  • D. Song et al.

    Transactions of Nonferrous Metals Society of China

    (2009)
  • R.Z. Valiev et al.

    Progress in Materials Science

    (2000)
  • A. Balyanov et al.

    Scripta Materialia

    (2004)
  • D. Song et al.

    Corrosion Science

    (2010)
  • Y.-W. Hao et al.

    Journal of Iron and Steel Research, International

    (2009)
  • G. Palumbo et al.

    Materials Science Forum

    (1996)
  • Cited by (326)

    View all citing articles on Scopus
    1

    ISE member.

    View full text