Critical Assessment 9: Wrought magnesium alloys

Introduction

As has been mooted for the past two decades or more, magnesium (as the lightest structural metal) has potential to lead to a step change mass reduction for structural applications in the automotive and aerospace industries. ¹ However, magnesium alloy usage is currently limited predominantly to cast applications (mainly die castings). Cars and aircraft require mainly wrought products in the form of sheet, plate, extrusions and forgings. Despite the theoretical weight saving benefits, wrought magnesium alloys have yet to find widespread application, although the use of cast magnesium has grown strongly. ²

Magnesium competes in transportation applications with aluminium, which has higher density but lower manufacturing cost, and carbon fibre reinforced polymer, which offers even greater weight savings albeit at much higher cost. ¹ Wrought magnesium alloys have two major barriers to wider application; poor low temperature formability and poor corrosion resistance. The corrosion behaviour of magnesium and the effects of alloying are discussed comprehensively in a recent review. ³ The present paper therefore focusses on formability, joining, and strengthening of magnesium alloys, with only a brief discussion of the most critical corrosion issues. Recent developments in these areas are summarised and remaining challenges are identified. Based on the current state of the art, a prognosis about future developments is made and topics requiring further research are identified.

Formability

A key limitation of wrought magnesium alloys is poor room temperature formability. ⁴ Most automotive body panels are currently cold stamped from steel but this method cannot be used to form magnesium. The formability issue can fundamentally be attributed to the hexagonal close packed structure of magnesium. However, behind this simple statement there is a complex interplay of alloying additions, deformation modes and texture development that determine the ultimate formability of the final product. The past decade has seen major progress towards understanding these interactions through both experimental studies and computer simulation.

Computer simulations, particularly crystal plasticity models, have helped to understand formability in magnesium alloys. The very high inherent anisotropy of magnesium crystals makes the behaviour of polycrystalline assemblies highly complex since typically there will be a mixture of grains in harder and softer orientations. Classical approximations to polycrystalline behaviour from single crystal data do not accurately reproduce the behaviour seen in magnesium. Twinning results in a rapid change of local crystal orientation so that a grain in a soft orientation can suddenly flip into a hard orientation (for example). The development of crystal plasticity models that can include twinning effects has provided a major advance in understanding how deformation is accommodated in polycrystalline magnesium.^5,6 Advances in experimental methods such as neutron diffraction techniques have enabled in situ probing of deformation activity in bulk magnesium specimens undergoing deformation, providing validation of models and new insights into load sharing during polycrystalline deformation.^7,8

Computer simulation has helped in understanding how strain localisation and failure initiation occurs during forming. For example, the importance of cascades of twins produced in strongly textured material has been demonstrated using crystal plasticity finite element methods (CPFEM). ⁹ Such twin cascades can lead to strain localisation across sheet material, producing early failure. By manipulating texture, a microstructure that is resistant to such cascades can be developed. ⁹

Although the ideal scenario of directly substituting magnesium for aluminium or steel in a cold stamping line is unlikely to ever become reality, fundamental research and modelling such as that described above have helped lead to sheet products that can be formed reliably at temperatures below 250°C. ¹⁰ This research has focussed on understanding the relationships between active deformation systems, texture, anisotropy, and forming performance as summarised in the following section.

Anisotropy and asymmetry

The room temperature yield behaviour of magnesium crystals is highly anisotropic. ¹¹ The critical resolved shear stress for basal slip is approximately 40 times less than that of prismatic slip, the next easily activated slip mode. The easiest slip mode that can accommodate c-axis deformation (<c+a> slip) is 50 times harder to activate than basal slip. ¹¹ The very high critical resolved shear stress required for <c+a> slip results in twinning in response to stress applied along the c-axis. Twinning is inherently asymmetric, and the twin that produces c-axis extension in magnesium ( c-axis tension twin) is activated at a lower stress than the c-axis compression twin. This leads to mechanical asymmetry, where the strength in compression is markedly different from that in tension along the same axis. The high anisotropy and asymmetry of magnesium are key contributing factors to the poor formability, especially in biaxial or complex loading conditions, where crystals are required to accommodate deformation in many directions simultaneously.^12,13 When magnesium is heated, non-basal slip modes become more easily activated, and good ductility is achieved above around 250°C. ¹⁴ However, forming at such elevated temperatures leads to an uneconomic increase in manufacturing cost for many applications.

Apart from heating, other less conventional methods have been considered to directly improve the ductility of magnesium alloys. One possibility is to exploit the electroplasticity effect that has been observed in magnesium (and other materials). Electroplasticity refers to the enhanced plasticity that can be obtained by passing a high current through the magnesium alloy.^15,16 Although electroplasticity has been reported as being useful in magnesium, current evidence suggests that most of the claimed benefits arise from the heating induced by the current rather than special effects that can be attributed to the electron wind. ¹⁶

There are two strategies available in polycrystalline magnesium alloys to reduce room temperature mechanical anisotropy and asymmetry. One method involves decreasing the inherent difference in critical resolved shear stress (CRSS) between deformation modes by alloying. It is important to note that any strengthening mechanism that increases the CRSS of all deformation systems (e.g. Hall–Petch strengthening) will lead to a convergence of relative CRSS values, as explained by Hutchinson and Barnett. ¹¹

The second method is to manipulate the texture via recrystallisation so that a balance of soft and hard modes is achieved regardless of loading condition. These strategies are not independent, since the recrystallised texture develops from the deformation texture, which itself depends on slip and twinning system activity.

Texture modification

In polycrystals, an additional and highly effective method to control deformation behaviour is through manipulation of texture. A random or weak texture will mean that some grains will be favourably oriented for basal slip whatever the loading direction. Unfortunately, most magnesium alloys produce strong textures after thermomechanical processing. ³³ In this case, the polycrystalline alloy behaves more like the single crystal case, although it is important to note that even a small spread in orientations from a ‘perfect’ texture is very useful in reducing anisotropy. ³⁴ Most magnesium alloys form strong basal textures after deformation due to the dominant role of basal slip and twinning. This texture is often preserved after recrystallisation, although its strength may change. ³⁵ Certain alloying additions can lead to a marked change in texture and this can be accompanied by a large improvement in formability.^36–38 Understanding how to use alloying to manipulate texture has become a topic of major research activity over the past decade.

Rare earth (RE) alloying additions are most effective in producing a texture change effect. The combination of RE and another addition (particularly Zn) can lead to a more potent effect than RE alone. ³⁹ As an example, Fig. 1a and b shows the textures generated in an Mg–Zn binary alloy sheet and the same sheet with a small RE addition (0·052 at.-Ce). ⁴⁰ The alloy without RE exhibits a strong basal texture with a slight split towards the rolling direction (RD). The RE containing alloy shows a weaker texture, with a pronounced split in the basal poles towards the transverse direction (TD). This texture difference leads to a better forming performance for the RE alloy. Figure 1c shows a room temperature forming limit diagram (FLD) comparing AZ31 (no RE), ZEK100 (Mg–Zn–RE–Zr) and AA6016, a widely used age hardenable aluminium alloy.^41,42 It can be seen that the forming limit strains of ZEK100 can be more than double that of AZ31 and approaches or exceeds that of the aluminium alloy.

OPEN IN VIEWER

1

Basal pole figures for a Mg–Zn binary (non–rare earth (RE) alloy) and b Mg–Zn–Ce (RE containing alloy) showing difference in texture, ⁴⁰ c forming limit diagram for AZ31 (non-RE magnesium alloy), ZEK100 (Mg–Zn–RE alloy ⁴¹ ), 6061 aluminium alloy ⁴² (RD, rolling direction; TD, transverse direction)

RE additions are expensive and lead to an undesirable increase in alloy cost. Fortunately, it has been shown that only a very small concentration of RE (as little as 0·01 at.-) is needed to produce the desired texture change, with the critical amount scaling with the RE solubility. ³⁸ It has now been convincingly demonstrated that one of the main reasons RE is so effective at low concentrations is that RE solute atoms, which are large in size, segregate strongly to dislocations and grain boundaries.^43–46 This means that the local concentration at these sites is far higher than the global addition. Recent advances in high resolution energy dispersive X-ray (EDX) analysis allows this segregation to be directly observed, for example Y segregation at a grain boundary as shown in Fig. 2a. Other recent work has shown that Gd behaves in the same way, and the texture change effect can be correlated with grain boundary segregation. ⁴⁶ A strong tendency for REs to segregate is also predicted by a simple model based on size mismatch alone (Fig. 2b). ⁴⁵ Other large atoms such as calcium would be expected to show the same behaviour, although this remains to be confirmed experimentally.

OPEN IN VIEWER

2

a segregation of rare earth (Y) on grain boundary measured using TITAN ChemiSTEM and b predicted grain boundary concentration of various RE additions for mean alloy composition of 0·22 at.- (dashed line) using simple classical size based model ⁴⁵

The mechanism by which RE additions lead to texture modification and formability improvement is not simple, and has been the subject of intense study. A recent review summarises this work. ⁴⁷ It has been widely demonstrated that the RE texture emerges during recrystallisation, whether static or dynamic.^40,48,49 RE additions have been shown to retard recovery and dynamic recrystallisation in magnesium, leading to a higher stored energy in the deformed microstructure with a greater spread of orientations from which recrystallised grains form.^49,50 The change in texture during recrystallisation therefore has its origins in a change in the deformation structure when RE elements are present.

The role of RE in changing the nature and distribution of shear banding in the deformed state is important.^25,44,51,52 Recrystallisation inside shear bands leads to off-basal grain orientations, and the presence of RE has been shown to lead to more extensive and uniformly distributed shear banding. This change in behaviour itself has been attributed to an increased activity of c-axis compression ( ) twinning and pyramidal slip.^25,53 This, in turn, can be explained by a reduction in the I₁ stacking fault energy in magnesium when RE is added,^22,54 which is predicted to be even more pronounced when zinc is also present. ⁵⁴ The changes in SFE combined with the tendency for RE to segregate to dislocations and boundaries are therefore the fundamental underlying factors that are necessary to explain all the other phenomena that emerge when RE additions are present.

RE additions do appear to have a unique effect on the texture and formability of magnesium that cannot be directly replicated by any other lower cost alloying addition. Although Ca additions have shown promise in changing the texture of magnesium, ⁵⁵ Ca also precipitates as a brittle intermetallic (Mg₂Ca) phase that can undo the benefits of the texture effect on formability. Nevertheless, Mg–Zn–Ca alloys have attracted considerable interest for their promising mechanical properties and biocompatibility.^56–58

Alternative processing routes

The traditional method of producing sheet by rolling from a thick as cast billet is unlikely to be the most efficient and cost effective way to process magnesium. One attractive possibility is to use twin roll casting to directly cast material close to the final required sheet thickness.^59–61 By only imparting a small amount of deformation to the material following casting, there is a possibility to avoid the generation of the detrimental strong basal texture. However, this process presents its own challenges. These include the difficulties in producing a uniform, fine grained microstructure and avoiding composition variations through the sheet thickness. One promising route to avoid these problems is to use melt conditioning prior to casting to produce a refined and more uniform grain structure. ⁶²

Other methods to modify the texture by processing include using equal channel angular extrusion (ECAE), ⁶³ cross rolling, and shear (differential speed) rolling. ⁶⁴ ECAE leads to shear textures and such material can show marked improvement in ductility for axial loading. ⁶³ However, ECAE is difficult to scale up to the industrial level. Cross rolling and shear rolling tend to produce only a modest change in texture and have some practical problems in industrial implementation for sheet production.

An alternative method to overcome the poor low temperature formability of magnesium is to use a high temperature forming process. Superplastic forming is well suited to magnesium; unlike in aluminium, special alloys are not required. ⁶⁵ This is because many magnesium alloys can undergo dynamic recrystallisation during superplastic forming, which allows the retention of the necessary fine grain size. Although relatively expensive, the process presents opportunities for cost reduction from lower manufacturing costs if a complex assembly can be replaced by a single superplastically formed component.

Corrosion and joining

Other key obstacles to the wider use of wrought magnesium are corrosion and joining. In dealing with these challenges it is important to recognise that a completely magnesium structure is never envisaged, so magnesium will always be used in combination with other materials according to the ‘right material, right place’ philosophy. ⁶⁶ The greatest challenges to solve are therefore dissimilar material joining of magnesium to other candidate materials such as aluminium alloy or steel, and protection against corrosion of such multi-material structures.

Magnesium is high in the electrochemical series, but forms an oxide that reduces its corrosion rate to be comparable with that of aluminium in most environments. ¹⁴ Corrosion resistance in magnesium is highly sensitive to impurities (e.g. iron) and it is essential to keep these to below critical levels to avoid very rapid degradation. ¹⁴

The high activity of magnesium does become a particular problem when coupled to other materials due to galvanic corrosion. Magnesium will serve as a sacrificial anode in such situations and will suffer rapid attack. Mechanical fastening (e.g. bolting, riveting) is currently the most common method used to join magnesium parts to those made from other metals. To protect such joints from galvanic corrosion, careful assembly practices are employed including the use of a shim or sealant between the dissimilar metals and the use of bolts and washers plated with metal that leads to a low galvanic activity with magnesium. ⁶⁷ However, this all adds to the assembly cost, and concerns about galvanic corrosion, particularly for automotive parts exposed to road salt, remain a key barrier to the wider use of magnesium parts. Other protection methods are also widely used including electroless conversion coatings, anodising, and physical coating systems. ⁶⁸

Welding offers a number of advantages over mechanical fastening including high speed and reduced cost. Welds between similar magnesium alloys can be made using a wide range of both fusion and solid state processes, similar to those employed for aluminium alloys. ¹⁴ The acute challenge is using welding to made dissimilar joints between magnesium alloy and other metals. Welding of magnesium alloys to aluminium alloys is of particular interest since such a combination is likely to be required in any future magnesium intensive multi-material vehicle. The difficulty in this case is to control the rapid reaction at the weld interface that occurs between the aluminium and magnesium.

Solid state welding processes, such as friction stir or ultrasonic welding are most likely to be successful since the reaction in the solid state is much slower than between liquid and solid. ⁶⁹ Nevertheless, even using a solid state process, a thick reaction layer forms rapidly on joining magnesium to aluminium (e.g. Fig. 3). This reaction layer typically consists of two intermetallic phases, both of which are brittle, leading to a steep reduction in the fracture energy of the weld and an interfacial failure mode.^70–72 Methods to control the reaction layer growth that have been explored include the use of a barrier layer surface coating (e.g. Mn) ⁷² or pre-coating of magnesium with aluminium. ⁷¹ At best, these approaches have so far been only partially successful in improving weld properties.

OPEN IN VIEWER

3

Evolution of reaction layer during ultrasonic welding of magnesium to aluminium alloy, after a 0·24 s, b 0·4 s and c 0·7 s; d electron backscattered diffraction phase identification map showing two intermetallic phases in reaction layer (0·4 s welding time) ⁷⁰

Strengthening

In many applications, strength is a key design parameter. The high strength aluminium alloys used in the aerospace and automotive industries obtain their properties by age hardening (precipitation). Age hardened aluminium alloys have reached a high level of sophistication and can reach strengths that are 100 times that of the base metal. Meanwhile, the best age hardened magnesium alloys are only around 10 times the strength of pure magnesium. ⁷³ Therefore, in strength critical applications, there is a strong motivation to develop magnesium alloys with an improved age hardening response. A desirable wrought magnesium alloy would combine the ability to be easily formed in a soft condition with a large hardening response after forming to provide both good formability and high strength.

The poor age hardening response of Mg–Al–Zn alloys is initially surprising, because the volume fraction of precipitates that can be formed in such alloys is high (e.g. 15 in AZ91 ⁷³ ). There are three main reasons for this poor response. First, there are no metastable precursors to assist nucleation, so the nucleation rate is low and relatively few widely spaced precipitates form rather than the desired fine dispersion. ⁷⁴ Second, the precipitates that do form are mainly plate shaped and lie on the basal plane where they are in a poor orientation to block basal dislocations.^32,75 More effective are the prismatic plate shaped precipitates formed in alloys such as WE43 (Mg–Y–RE) since these are well oriented to block basal slip. ⁷⁵ Finally, discontinuous precipitation (DP) can occur, which leads to a relatively coarse lamella structure and reduces the solute available for strengthening continuous precipitation. ⁷⁶ Although DP is usually considered undesirable (e.g. for the creep resistance of the alloy) if controlled and exploited correctly it has also been proposed as a route to producing in situ magnesium composites.

Improving the age hardening response of magnesium alloys requires promotion of nucleation to form a fine, homogeneous dispersion of precipitates with a shape and habit that provide effective barriers to dislocation motion. In aluminium alloys, micro-alloying (e.g. with Ag) has been used to this end and there is potential to use this approach in magnesium.^77,78 Significant research activity is now underway to explore this concept, using a combination of modelling and experiments to identify promising candidate micro-alloying additions.^79,80

An alternative strengthening approach is suggested by the discovery that certain magnesium alloys can form long period stacking ordered (LPSO) phases that can impart high strength while retaining useful ductility.^81,82 Typical LPSO forming magnesium alloys are based on the Mg–Zn–RE system (high solubility REs only, e.g. Y, Gd, Dy, Er, Ho) and can form a volume fraction of LPSO phase around 25. ⁸³ The 0·2 proof stress of such alloys can exceed 600 MPa, ⁸¹ approximately twice that of conventional age hardened magnesium alloys. Various different stacking sequences are possible, producing different LPSO phases. ³⁰ The strengthening provided by these phases is highly anisotropic, meaning that their orientation and spatial distribution is also important in determining the overall strengthening effect.^83,84 Understanding how to best optimise the chemistry and processing of this class of alloys remains a topic of active research.

Future outlook

The current state of the art for cost effective formable magnesium sheet is strip cast (TRC) material with a fine grain size based on the Mg–Zn–RE system. For magnesium alloys with improved strength, micro-alloying guided by atomistic simulation offers the prospect of greatly enhancing age hardening response. Alternatively, alloys based on strengthening by long–period stacking ordered (LPSO) phases also show considerable promise. Solid state welding processes (e.g. friction stir welding), possibly using a filler or barrier material to control interfacial reaction, show the best potential for low cost dissimilar metal joining of magnesium parts to other metals. New low cost protection systems are required to protect such dissimilar joints against galvanic corrosion.

Technologically, there is a limited window of opportunity for wrought magnesium alloys to find wide acceptance as structural materials for automotive and aerospace applications. Mass produced, aluminium intensive vehicles are now a reality. Carbon fibre reinforced polymer is also beginning to find limited application in mass produced vehicles. The historical precedent of the aerospace industry suggests that once the transfer to composite construction has been made, a shift back to metal is unlikely, even if better magnesium alloy solutions were to be subsequently developed.

The intense research activity on wrought magnesium over the past decade has led to a greatly improved understanding of these alloys, but there remain a number of key scientific challenges to be solved and opportunities where the full potential of magnesium based alloys has yet to be realised. Six of the most pressing, in the opinion of the author, are highlighted below.