Final Report Mg Science & Technology Workshop Fundamental Research Issues Held May 19-20, 2011 Arlington, va by

Download 156.35 Kb.
Size156.35 Kb.
1   2   3   4   5   6   7

Coatings and Corrosion

Many view corrosion as THE issue which prevents broader application of Mg alloys. The intrinsic problem is that elemental Mg is not thermodynamically stable. In fact, thermodynamically speaking, it is the most active structural metal, and unfortunately, it does not have a good inherent kinetic barrier to corrosion. What we can do to improve its corrosion resistance is try to slow down the kinetics of the dissolution reactions by applying a barrier (coating) at the interface with the environment, adding alloying additions which promote better surface film properties, or altering the environment (where possible).

The basic corrosion mechanisms for Mg are well established, and the common Mg alloys (AZ91, AM60, AZ31, etc.) have been extensively studied. However, there still are some phenomena that cannot be explained. Further studies are urgently required, since these will affect future developments and applications of Mg alloys. In short, there is still opportunity to develop high quality, quantitative corrosion data with correlation to microstructure and processing, even in commercial alloys. Given the commercial significance of these alloys, such research is valued by industry. Similarly, there is interest in developing standardized testing schemes that focus on standard environments, relevant to the applications. For example, existing ASTM standards often do not reflect the in-service environments encountered. Thus, there is also interest in developing better qualitative (corrosion damage morphology) and quantitative (corrosion damage evolution) correlations between laboratory tests and service experience.

The detailed roles of all possible alloying/coating elements are not established and must be. Further understanding of precise roles of major and minor alloying elements of passivity and on the details of anodic dissolution must be developed. Further, the role of major and minor alloying elements, intermetallic compounds, and microstructural heterogeneity on cathodic reactions must be better understood. The role of species in complex solutions in aiding in passivation, beyond mere oxides, should be determined. Examples include the possible formation of complex mineral scales.

Given the priority of improving the corrosion resistance of Mg alloys, all possible strategies should be pursued including the development of “game changing,” passivated “stainless” alloys, alloys with more resistant surface "skins,” traditional purification approaches that have proven successful in improving the corrosion resistance of AZ91C  AZ91D  AZ91E, and coating strategies based upon metals, oxides, and polymers. Concerning metallic coatings, some feel that amorphous coatings or surface modification may have a role to play in corrosion protection of Mg.

In parallel with alloy/coating development and in support of it, the development of a predictive Mg alloy corrosion modeling capability is viewed as critical. Although thermodynamic modeling (including first-principles modeling of Pourbaix diagrams) has some potential, due to the thermodynamic limitations mentioned above, we really need a broader modeling approach including gathering kinetic data and applying it to modeling. This kind of holistic, experimentally validated modeling is expensive, requires a team of investigators, and needs a long-term financial commitment, but the workshop participants view it as important. A roadmap for this modeling work can likely be gained from the recent corrosion modeling work on Al alloys (e.g., the work of Rob Kelly and John Scully at UVa, and Farrel Marin at NRL).

Macro-galvanic corrosion due to joining of dissimilar metals is understood and viewed as an engineering design problem, not requiring deeper mechanistic understanding. Micro-galvanic coupling between phases within an alloy is a significant issue in Mg alloys that deserves significant research attention. Further, coatings should be designed to optimize protection against galvanic corrosion.

Beyond corrosion, other environmental effects, such as stress corrosion cracking and environmentally affected fatigue cracking should be investigated as these are often life-limiting in practical applications in which the uniform and localized corrosion damage accumulation rates in the absence of stress would be acceptable. There does not appear to be nearly as much quantitative data available regarding the effect of the environment on the mechanical properties of Mg alloys, relative to Al alloys, steels, etc. Along these lines, alloying and surface treatment designed to minimize hydriding and ways of suppressing/poisoning the hydrogen evolution reaction should be explored.

A final issue related to the corrosion resistance of magnesium alloys is the possible application of bioabsorbable stents and orthopedic implants. In this case, the controlled, but reasonably rapid, corrosion of the implant is actually the goal. There are many issues to be addressed, including processing strategies, drug coatings, and possible unintended health effects of locally high Mg concentrations. Significant research in this area is being done in Europe, with new programs being proposed in Canada, and limited research on-going within U.S. institutions.

  1. Mechanical Performance

Deformation Mechanisms

A great deal about the deformation mechanisms of magnesium and its alloys has been learned in the past decade. The basic roles of dislocation-based mechanisms of plasticity, including basal and non-basal slip of type dislocations, and the significance of non-basal dislocations are established. The basic role of mechanical twinning is understood as well, including {10.2} “extension” twinning, {10.1} “contraction” twinning and {10.1}-{10.2} double-twinning. EBSD has proven very effective in answering many of the previously unanswered questions. The plastic anisotropies (and asymmetries) that result from the operation of these mechanisms are basically understood and can be accounted for by existing crystal plasticity modeling approaches. The same can be said about the effect of crystallographic texture on macroscopic deformation. There are some inconsistencies in the literature, but consensus appears to be emerging on many of these aspects.

That said, there are many basic phenomena which require much more careful consideration of the various individual deformation mechanisms and of the crystallographic texture than is required for traditional cubic metals. For example, there currently is no widely accepted rule for grain size strengthening of Mg alloys, which can accurately account for all these factors, let alone the effect of grain size distribution. It is suggested by some that only crystal plasticity based approaches will suffice to accurately capture these details relevant to materials design, although this approach may not immediately solve the problems of component and process design. There are many other outstanding questions for which only vague answers exist today. There has been a great deal of confusion about the basal stacking fault energy. Experimental estimates published in the literature vary tremendously.  Some very long stacking faults have been observed, but recent experiments reveal solute segregated at such faults. (In fact, some argue these are precipitates.)  If there is no solute, then the faults tend to be very short.  The interrelation between chemistry and stacking fault energy should be clarified. Atomistic modeling may be helpful in such cases. Atomistic modeling is also demonstrating promise in the area of mechanical twinning. However, we still have a very incomplete understanding of the nucleation of twins. The effects of twinning are challenging to implement in continuum models, which are important for forming prediction. Existing continuum models are highly suspect.

It is generally observed that our understanding of the strain hardening behavior of Mg alloys is much less developed that that for fcc metals, for example. What are the interactions between various dislocation and twin types? Additionally, the effect of strain path changes on strain hardening behavior is much stronger than that in cubic metals, and must be accounted for in order to develop robust constitutive models. The effect of temperature (and strain rate) on the strength and activity of various deformation mechanisms has been heavily studied and a number of issues have become clear. The main twinning mode, {10.2} extension twinning, appears to be essentially athermal, while the {10.1} mode is thermally activated. Atomistic modeling has helped to explain distinctions like this, and the details of the twinning dislocation structure are becoming clearer.

Alloy designers, in particular, want to know the quantitative effects of alloy solute and precipitates on the individual deformation mechanisms. In this context, there is a need for control and characterization of tramp impurities.  Commercially available material is not of very high purity.  The role of Zr, for example, is not clear. Atomistic modeling is beginning to augment single crystal experiments in the area of solute effects on dislocation mobilities and stacking fault energies. There are good ideas about the relative impacts of particles: size, shape, and orientation on yield strength and various deformation mechanisms, but these are not well-tested and our ability to predict microstructure effects on strain hardening behavior are still at a relatively nascent state. Finally, there is great interest in better understanding the mechanism(s) of shear localization, which impacts plastic anisotropy, fracture during quasistatic and dynamic loading, and dynamic recrystallization during hot deformation processing. Our basic ability to predict shear banding in Mg is rudimentary.

Some researchers have proposed explanations for the poor multi-axial ductility, fracture toughness, and resistance to shear instability based upon mechanical twinning. Images of shear bands, cracking, and cavitation associated with twins have caused scientists to suggest that minimizing the occurrence of mechanical twinning would promote improved formability and improved dynamic loading response. However, hard evidence that this strategy would work is lacking. Current strategies for minimizing twinning are limited to grain size refinement. Even that notion has been challenged by some recent observations of twinning in a nano-grained Mg-Ti alloy. Our knowledge of the impact of solutes on mechanical twinning is limited. There have been some investigations of the effects of various precipitates on twinning, but this area is still considered open to investigation. The notion that twinning is detrimental to ductility must be tempered by the knowledge that pure Ti is ductile despite prolific twinning and Ti-alloys in which twinning is suppressed have much poorer ductility. Further, one study of the Mg-Cd solid solution alloys revealed that the ductility plummeted in the alloy in which the c/a ratio caused the {10.2} twinning mode to switch from an “extension” to a “contraction” twin, i.e. at that condition extension twinning did not occur.

There are various new characterization methods that have been successfully brought to bear on the problem of deformation mechanisms and these ought to be further exploited. These include, but are not limited to: neutron diffraction, 3D synchotron methods, nanoindentation, micro-pillar compression, X-ray tomography, high angle annular dark field imaging in the TEM. There is a need for mesoscale characterization, including grain size, grain orientation and grain boundary character distribution effects on mechanical response and on corrosion resistance. There also opportunities for efficient, statistical quantification of the details of the microstructure using a more rigorous framework such as the n-point statistics, since this approach has not been applied to Mg alloys. There was a discussion of FIB damage. Mg is easily damaged by Ga. This is a problem in preparing micro-compression samples, TEM specimens, and atom probe specimens.

Dynamic Loading

Mg alloys generally exhibit poor dynamic loading response. As suggested above, they have lower resistance to shear instability than many competing Al alloys and steels. Fundamental explanations and possible solutions must be sought. An explanation rooted in texture-based plastic anisotropy has been offered, however, this is countered by the fact that even randomly textured die castings suffer from this problem. There is still the possibility of a texture-softening effect during shear localization; this should be further explored. The poor resistance to shear instability may be connected with a low strain hardening rate, beyond an initial period of high strain hardening. Strategies to improve the strain hardening response should be explored. Similarly, higher strain rate hardening behavior is cited as beneficial. Finally, it would be very useful if there were models which could explain the relationships between strength/ductility and energy absorption/fragmentation.


Clarity has emerged in the area of creep deformation under service conditions, whereas there were significant inconsistencies in the recent past. For example, the role of grain boundary sliding has been disputed. Recent work has suggested that creep is dislocation accommodated in most of the alloys and stress/temperature regimes of interest for application. Emphasis has now been placed upon alloy systems which offer significant solid solution strengthening or stable precipitation strengthening, the two major avenues employed to achieve creep resistance in other alloy systems. Nevertheless, there is continued interest in understanding the role of grain boundary sliding type deformation at low temperatures and during high temperature forming. A few years ago, creep and bolt-load retention were the principal mechanical properties of interest, as researchers sought to develop new alloys for automotive powertrain applications. The successes achieved with the Mg-Al-RE and Mg-Al-Sr alloys within the engine cradle and engine block applications have demonstrated the feasibility of Mg alloys for powertrain applications.

Fatigue and Fracture

Discussion revealed that the understanding of Mg fatigue and fracture behavior is more nascent than our basic understanding of plastic anisotropy of textured alloys and creep behavior, for instance. The existing studies have been largely empirical and the models phenomenological, rather than mechanism based. Connections between microstructure and fatigue properties are not clear. One thing that is clear is that Mg alloys exhibit a very strong Bauschinger effect, at least part of which is due to the intrinsic plastic anisotropy of Mg single crystals and a phenomenon known as twinning-detwinning. The significance of the latter effect on the fatigue properties is only vaguely understood.

The state-of-the-art fatigue models are empirical and are not Mg-alloy specific, which does not provide sufficient information for modeling a broad class of cast and wrought Mg alloy developments. Cyclic plasticity modeling of pure HCP Mg and Mg alloys and understanding the interaction between the slip and twin during unloading should be viewed as a priority. Recent work on deformation and role of microstructure and texture, both experimental and modeling, needs to be extended to cyclic behavior and fatigue crack propagation. In order to achieve this goal, we would need to develop a physics-based understanding of the mechanisms of fatigue damage formation and small-crack growth. Toward this end, 3D microstructure data bases are needed for the modeling of fatigue in magnesium alloys. Such data bases have been developed for aluminum alloys of interest to the aerospace industry. Emerging physics based models for fatigue and fracture, based on 3D microstructure datasets are ready to be developed and applied for Mg-alloys. The computational, analytical, and characterization tools required to advance our understanding of Mg fatigue mechanisms and modeling are available. Applying such tools to Mg will in-turn contribute to improved Mg alloy design and optimization. In short, if we could develop sound, physics-based models for cyclic deformation, damage development and small crack propagation, more rapid alloy design and materials insertion into commercial applications would become a reality.

There is a similar need to educate the Mg research community regarding fracture mechanics and failure models/modes. While our collective knowledge of Mg fracture behavior is relatively limited, we do understand that Mg alloys do not fail by cleavage, despite the fact that many authors frequently misapply that term in their fractographic analyses. Things as basic as the impact of triaxiality, given the strong anisotropy, appear to be open questions. Some materials are more sensitive to imposed hydrostatic pressures than others. This ought to be further explored for Mg alloys. Quantitative understanding and modeling of the nucleation of cracks is needed. For example, are Gurson-type models applicable? Do they need to be modified to account for the greater tendency to localize? There have been very few systematic studies of environmental effects, such as the possibility of hydrogen embrittlement, stress corrosion cracking (which may become a bigger concern if higher strength alloys are developed), and environmentally enhanced fatigue.

  1. Download 156.35 Kb.

    Share with your friends:
1   2   3   4   5   6   7

The database is protected by copyright © 2022
send message

    Main page