Elevated-temperature tribology of metallic materials

Abstract

The wear of metals and alloys takes place in many forms, and the type of wear that dominates in each instance is influenced by the mechanics of contact, material properties, the interfacial temperature, and the surrounding environment. The control of elevated-temperature friction and wear is important for applications like internal combustion engines, aerospace propulsion systems, and metalworking equipment. The progression of interacting, often synergistic processes produces surface deformation, subsurface damage accumulation, the formation of tribo-layers, and the creation of free particles. Reaction products, particularly oxides, play a primary role in debris formation and microstructural evolution. Chemical reactions are known to be influenced by the energetic state of the exposed surfaces, and that surface energy is in turn affected by localized deformation and fracture. At relatively low temperatures, work-hardening can occur beneath tribo-contacts, but exposure to high temperatures can modify the resultant defect density and grain structure to affect the mechanisms of re-oxidation. As research by others has shown, the rate of wear at elevated temperatures can either be enhanced or reduced, depending on contact conditions and nature of oxide layer formation. Furthermore, the thermodynamic driving force for certain chemical reactions, the kinetics of those reactions, and the microstructure can all affect the response. The role of deformation, oxidation, and tribo-corrosion in the elevated-temperature tribology of metallic alloys will be exemplified by three examples involving sliding wear, single-point abrasion, and repetitive impact plus slip.

Introduction

The elevated-temperature friction and wear behavior of metals and alloys is an important consideration in the effective performance of certain moving parts in internal combustion engines, bearings in aerospace propulsion systems, cutting tools, and metalworking processes. Sometimes the interfacial temperature derives from external sources of heat, and at other times, as in vehicle brakes, the temperature results from frictional contact. At high temperatures, changes occur in bulk mechanical properties, bulk thermo-physical properties, and surface reactivity. Since there are few viable liquid or solid lubricants that work well at temperatures upward of 500 °C, a number of elevated-temperature applications for contacting metals depend on the ability of the bearing surfaces to self-lubricate based on reactions with their environments and their ability to form protective glazes (tribo-layers) during the contact process [1], [2], [3]. This discussion therefore includes three aspects of elevated-temperature tribology: (a) changes in bulk properties, (b) changes in reactivity, and (c) changes in tribo-layer forming tendency. Furthermore, as will be shown, differences in the dominant type of wear affect the structure of surfaces that form during repetitive contact at elevated temperatures.

As the temperature rises, the properties of the load-bearing materials and any interposed substances can change. As shown in Fig. 1 from collected data, the yield strength of metals and alloys usually tends to decrease at elevated temperatures. One notable exception is an alloy based on the intermetallic compound Ni₃Al, the wear characteristics of which have been described elsewhere (e.g., [4], [5], [6]). Due to the cross-slip characteristics of the (L1₂) crystal structure of the ordered γ′ phase, some variants of nickel aluminide alloys increase in yield strength up to about 700 °C before it declines. Furthermore, as Fig. 1 shows, the processing of the alloy (e.g., cast versus powder-processed) affects its strength versus temperature behavior. In general, however, based on the decline in strength of most metals with temperature, one would expect a reduced resistance to deformation and abrasion as temperature increases.

The modulus of elasticity also changes as a function of temperature and that affects the elastic behavior of metals. For example, Fig. 2 illustrates the calculated maximum elastic contact stress (Hertz stress [7]) for a cylinder 8 mm long and 9.53 mm in diameter loaded against a flat plane as a function of temperature. Both cylinder and plane materials were assumed to be composed of a Ni-based exhaust valve alloy called Pyromet 31V™, whose temperature-dependent mechanical properties were available from the producer [8]. As the elastic modulus declines with temperature from ambient to approximately 800 °C, the calculated maximum contact stress decreases by about 14%.

Thermo-physical properties of substances, such as the coefficient of thermal expansion (CTE), also change with temperature. Therefore, the tendency of oxide scales to adhere to the metals on which they grow as temperature changes depends not only on the specific volume difference between metals and their oxides, but also upon the similarity of their CTEs to those of the substrates upon which they grow. Other factors include the presence of brittle phases that form on and within the substrate due to depletion of alloying elements by surface segregation or inward diffusion of oxygen from the surface. Therefore, the mechanical properties of metals and their wear responses are tied to oxidation, sulfidation, and other processes that alter the chemistry of the region of tribo-contact.

The change in Gibbs free energy (G) is a measure of the phase stability of a certain compound at constant pressure.

$$\mathrm{\Delta }G=G_{\mathrm{products}}-G_{\mathrm{reactants}}=-\mathit{RT}\ln K$$

Here, R=the gas constant, T=the absolute temperature, and K is the ratio of the partial pressures of the products to the partial pressures of the reactants (for more details, see Ref. [9]). The lower the ΔG for a certain reaction, the higher its thermodynamic driving force, and presumably, the more stable is the reaction product.

An Ellingham diagram is one way to compare Gibbs free energies of formation for various oxidation reactions of metals as a function of temperature [10]. A simplified example of one such diagram is shown in Fig. 3. Based on ΔG considerations, the formation of alumina (Al₂O₃) is highly stable relative to a number of other oxidation reactions. For example, chromia (Cr₂O₃) would be most likely of the possible oxides to form on Fe–Ni–Cr alloys, like stainless steels, because its driving force is greater than that for iron oxide or nickel oxide.

Ellingham diagrams are helpful in understanding oxidation tendencies, and are valuable in extractive metallurgy, notably the steelmaking process. However, they are only a part of the picture in understanding the nature of tribo-surfaces at high temperatures. Other factors include microstructural effects and the kinetics of oxide growth. For example, oxidation kinetics of exposed clean metals have been observed to follow a few basic relationships in terms of oxide thickness (x),time (t), and with various rate constants (e.g., k_x, A, t_o) [11]:

$$\mathrm{Linear}:x=k_{l}t$$

$$\mathrm{Parabolic}:x=(2k_{p}t{)}^{1/2}$$

$$\mathrm{Logarithmic}:x=k{\log }_{10}(t+t_o)+A$$

As oxidation progresses, the relationship for the oxidation kinetics can change as well. Chemical reaction rates tend to increase when temperature increases, but they are also dependent on other factors like the surface area, material microstructure, and the surrounding environment. As products form on exposed surfaces, they deplete the reactants in the near-surface zones. Oxygen must either diffuse into the surfaces or reactants must diffuse out of the scales to enable the reaction to continue. Defect structures and grain boundaries can act as short-circuit paths to facilitate the diffusion of reactants. However, such short-circuit paths become less important as temperatures rise above approximately two-thirds of the alloy's melting point. The result of these interactive phenomena in multi-component alloys is complex, multi-phase microstructures.

Consider, for example, the formation of oxidation products in certain nickel aluminide alloys. Table 1 indicates the various phases that have been observed to form in the near-surface regions during exposure of a Ni₃Al-based alloy to oxidizing conditions at different temperature ranges [12]. At comparable exposure times, they range from predominantly NiO to predominantly Al₂O₃, depending on the temperature, but microstructures reveal that the near-surface layers are dominated by reaction products of various types.

As oxides form, their adhesion to the substrate becomes important. The Pilling-Bedworth Ratio (PBR) depicts the difference between the volume per metal ion in an oxide and the volume per atom in the related metal (see Ref. [11], pp. 118–119). If there is a large volume difference, stresses can build up in the oxide scales, leading to cracks, a loss of adhesion, and increased propensity for scales to be removed by mechanical contact. When the PBR exceeds 1, there is expected to be a compressive stress in the oxide, and when it is less than 1, a tensile stress. For example, the PBR for aluminum oxide on aluminum is 1.28, and than that for Fe₃O₄ on α-Fe is 2.10. As temperatures increase or decrease, differences in thermal expansion for the scales and substrates can also cause scales to detach.

When one adds the complications of different kinds of mechanical interactions, such as repetitive impact, abrasion, or sliding, to the static evolution of oxide scales on metals, the picture becomes more complicated. At the same time as the various processes of tribochemistry are taking place, material is being displaced or removed, and new near-surface defects are being created.

In the context of this paper, a tribo-layer is any distinct material that forms in an interface as a direct result of mechanical contact. For example, it could be a highly-deformed layer at the near surface of a metal subjected to tangential shears during sliding; it could be a mechanically-mixed layer of metal and oxide that results from fretting or repetitive impact; or it could be a deposit of accumulated wear particles that become trapped within an interface. Applications like car and truck brakes depend on the formation of tribo-layers on contact surfaces to control and stabilize friction. Mechanically-mixed layers can form during sliding and, as Fu et al. [13] have shown by elegant computer modeling, the contacting solids can mix in a turbulent manner, producing stringers and dissociated islands of one material surrounded by another of different composition.

Stott, Wood, and their colleagues have investigated the interaction between sliding wear and high-temperature oxidation in superalloys [1], [2], [3]. They described the role of tribo-layers called ‘glazes’ that form on sliding surfaces during frictional contact. Glazes can temporarily protect the surfaces from further contact damage. If they happen to wear off, new glazes can be formed to take their place. This progression of formation, loss, and reformation can result in short-term friction or wear transients. The temperature experienced in sliding interfaces is the sum of the temperature of the surroundings plus that generated by frictional contact. Therefore, the response of metals and alloys to elevated temperature sliding needs to consider these two contributions, as well as the effects they may have on the tribological evolution of glazes. For example, substrates that are locally softened by the superimposed effects of environmental temperature and frictional heating, can more easily mix into glazes to create marble-cake-like structures with swirled, alternating layers of oxide and metal.

It is also worthwhile to consider how different forms of wear – sliding, abrasive, and impact for example – affect the tribo-corrosion behavior of metals at elevated temperatures.