PROPERTIES OF SURFACES D. H. Buckley INTRODUCTION The first chapter showed that engineering surfaces are not smooth on a microscale but rather are rough, containing hills and valleys. The next step is to examine the chemistry, physics, metallurgy, and mechanical nature of the surfaces. NATURE OF REALSURFACES Metals and alloys are, by far, the most widely used materials in practical tribological systems. If a metal is taken from the ordinary environment, placed in a vacuum, and heated mildly, the surface almost always liberates water. Various hydrocarbons are also detected if the component has been in the vicinity of operating machinery. Desorption with mild heating indicates that bonding to the surface is weak and of a physical nature. Beneath this outer layer of absorbed water and gases, all metal surfaces (except gold) have a metallic oxide layer as indicated in Figure 1. With elemental metals, the particular oxide or oxides present depend on the environment, the amount of oxygen available to the surface, and the oxidation mechanism for the particular metal. If copper is heated in air, the outer black layer is CuO while an inner rose-colored layer is Cu 2 O. Surface oxides of iron can be Fe 2 O 3 , Fe 3 O 4 , or FeO. Oxides present on an alloy surface depend upon con- centration of alloying elements, affinity of alloying elements for oxygen, ability of oxygen to diffuse into surface layers, and segregation of alloy constituents to the surface. The importance of alloy element concentration is apparent, for example, in stainless steels. A certain concentration of nickel and chromium is necessary to ensure formation of nickel and chromium oxides to passivate the surface. With the relatively small concentration of 1.5% chromium in conventional 52100 ball bearing steel, presence of chromium may not even be detected in the surface oxide. Environmental conditions will vary the composition of the oxides. When 440-C bearing steel is heated in air to 500 °C, all alloying elements are present in the surface oxide layer. At 600 °C, however, iron is missing and only chromium is detected. Chemical affinity of the metallic elements also bears on the surface oxide formed. While gold does not form a stable surface oxide in room temperature air, oxides (Cu) form on gold-copper alloys. Gold has no affinity for oxygen but copper does. Titanium reacts so readily with oxygen that it is used as an oxygen “getter” in vacuum systems. Metals such as copper (and more particularly silver) react much more slowly. Under the absorbed gases, water, and the oxide layer, a mechanically worked layer is indicated in Figure 1. A finished mechanical component may have been machined, ground, cast, forged, extruded, or prepared by some other forming process. The energy which goes into the near surface region can result in strain hardening, recrystallization, and texturing. Grinding, for example, at high speeds is more likely to produce recrystallization than at low speeds. Further, the greater the deformation during grinding, the lower may be the recrys- tallization temperature. With titanium a 60% strain decreases the recrystallization temperature from 900 to 400 °C. METHODS OF CHARACTERIZATION OF SURFACES Microscopy Microscopy is the most common technique employed for characterization of surfaces. Volume II 17 Copyright © 1983 CRC Press LLC Volume II 19 FIGURE 2. (a) Field ion micrograph; clean tungsten surface, (b) Concluded; vacancy in the (203) plane of a platinum surface. a b Copyright © 1983 CRC Press LLC The atoms making up the faces of the cube for the face-centered cubic and the body- centered cubic are referred to as the (100) surfaces. As shown schematically in Figure 3a, these planes of atoms move relative to each other when the crystal is deformed plastically and are, therefore, referred to as slip planes. Under applied stresses, the (110) set of planes is commonly observed to slip over one another in the body-centered cubic system. The (111) planes are those upon which slip and cleavage in face-centered cubic materials is most frequent. The planes in Figure 3 are only three among these which can appear at the surface. X-ray diffraction reveals a host of different crystalline planes in a polycrystalline sample. The atoms in the three planes of Figure 3 are arranged as indicated in Figure 4. The (111) planes have the closest atomic packing and the lowest surface energy in the face-centered cubic system and, therefore, are least likely to interact chemically with environmental constituents. The (110) planes in the face-centered cubic system are least densely packed, have higher surface energies than the (111) planes, and are, therefore, much more reactive. Because they are less densely packed, their elastic modulus and microhardness are also less. Atomic packing can vary with direction of movement. For the (111) plane in Figure 4, two basic directional packing variations are seen to exist. Surface energies vary in these two directions. Low Energy Electron Diffraction (LEED) LEED is widely used for characterization of the surface atomic structure on crystalline solids. Single crystals arc generally studied although large grained polycrystals can also be examined. Low energy electrons in the range of 20 to 400 eVare diffracted from the surface crystal lattice to produce a reciprocal image on a phosphorus screen. Figure 5 contains three LEED patterns from an iron (110) surface. The photograph and pattern at the left is for the iron surface with oxide removed. Upon oxide removal and heating of the iron in vacuum, the iron becomes covered with a film which produces a ring structure of diffraction spots on the surface. Auger electron spectroscopy analysis (discussed in the next section) identified the surface film to be graphitic carbon which segregates from the iron bulk. When the iron is argon ion bombarded, the carbon disappears and four diffraction spots remain in a rectangular array representative of the clean iron (110) surface. The iron diffraction spots are fuzzy and elongated from the strain in the iron surface lattice produced by the argon bombardment. Very mild heating removes the strain and produces the clean iron diffraction pattern of the right photograph. LEED is effective in identifying a clean metal surface, its structure, condition or state, and structure of films. In a manner similar to that illustrated in Figure 5, LEED will also distinguish between organic films and their structural arrangement on an iron surface (see References for more information on LEED). Auger Electron Spectroscopy (AES) AES analyzes for all elements present on a surface except hydrogen and helium. It is 20CRC Handbook of Lubrication FIGURE 3.Planes of possible slip in a cubic crystal; (a) three (100) planes, (b) six (110) planes, and (c) four (111) planes. Copyright © 1983 CRC Press LLC With XPS a monochromatic X-ray beam is used as the energy source. The beam causes ejection of electrons with kinetic energies characteristic of the surface atoms. Aspectrum of the elements present is obtained by plotting the total number of electrons ejected as a function of kinetic energy. XPS gives binding energies of the elements which enables identification of the compounds in which these elements exist. The binding energy of the electrons ejected from the surface is determined by their chemical environment and is roughly a function of the atomic charge. The binding energy measured with XPS will be altered by changing the particular elements bound to the element being examined. Elemental sulfur has a characteristic binding energy of 162.5 eV. Negatively charged S −2 has a lower binding energy. When oxygen is bound to the sulfur, the sulfur binding energy increases. Further, the SO 4 −2 structure has a greater binding energy than SO 3 −2 which can be used to distinguish between sulfur bound in these two states. Other Techniques Over 70 surface tools have been developed for analysis and chemical characterization. A few more commonly used techniques are indicated by their acronyms in Table 1. The nondestructive techniques are nuclear back scattering spectroscopy (NBS) and electron mi- croprobe (EM). Auger electron spectroscopy (AES), X-ray photoemission spectroscopy (XPS or ESCA), ion-scattering spectroscopy (ISS), and appearance potential spectroscopy (APS) are destructive only if sputter etching or depth profiling is used. Two techniques which are destructive are secondary ion mass spectroscopy (SIMS) and glow discharge mass spectroscopy (GDMS). These techniques detect the species sputtered from the surface (see book by Kane and Larrabee in the References for more details). Note from Table 1 that they both detect all elements except hydrogen and helium, provide excellent chemical identification, and have sensitivities of surface elements to as little as 0.01 mono- layer. Their disadvantage is that they must be operated in a vacuum system. Probably the most versatile tool is the scanning electron microscope (SEM). It is extremely useful in obtaining a view of features on a surface such as asperities, surface irregularities, and topography where adhesion and wear have occurred. When SEM has incorporated into it X-ray energy dispersive analysis, both topography and chemistry can be determined. The X-ray analysis is not a surface analytical tool, but it can provide considerable information where material transfer takes place in adhesion or sliding. An SEM photomicrograph of an Volume II 23 FIGURE 7. Auger electron spectra of car- bon. (a) Segregated at a Mo(110) surface during initial cleaning (labeled Mo-C); (b) CO on a clean Mo(110) surface (labeled Mo- CO); and (c) in graphite. Copyright © 1983 CRC Press LLC aluminum surface is shown in Figure 8a after sliding on an iron surface. The photomicrograph reveals surface topography while the X-ray map for iron reveals the white patch in Figure 8b where iron is detected on the aluminum wear surface. PROPERTIES OF SURFACES Metallurgy and Crystalline Structure The crystal structure of ideal surfaces has already been examined in Figure 4. All engi- neering surfaces vary from this ideal and have grain boundaries which develop during solidification as large defects which exist in the solid and extend to the surface. They do not possess a regular structure, are highly active regions, and on the surface are very energetic. Lesser defects include subboundaries, twins, dislocations, interstitials, and vacancies. Subboundaries are low-angle grain boundaries and usually occur where there is only a slight mismatch in orientation of adjacent grains on either side of the boundary. When the crystal lattices of adjacent grains are slightly tilted one toward the other, there is a tilt boundary. Where the lattices remain parallel but one is rotated about a simple crystallographic axis relative to the other with the boundary being normal to this axis, a twist boundary develops. The twin boundary occurs where there is only a degree or two of mismatch with the twins being mirror images. They are frequently seen on basal planes of hexagonal metals with deformation. Dislocations are atomic line defects in crystalline solids. They may be subsurface and terminate at the surface or they may be in the surface. Edge dislocations are entirely along a line where an extra half plane of atoms exists. Screw dislocations form along a spiral dislocation line. Small angle boundaries or subboundaries are generally composed of edge dislocations. These defects in crystalline solids cause them to deviate markedly from the theoretically achievable strengths of ideal crystals. Some of the crystalline surface defects are presented schematically in Figure 9. The vacant lattice site was seen on a real surface in the photomicrograph of Figure 2b. An interstital atom is crowded into the crystal lattice of Figure 9a. Edge and screw dislocations and a small angle boundary are also shown. Worn surfaces generally have undergone a high degree of strain and may contain large amounts of lattice distortion and defects such as dislocations. While initial dislocations cause a reduction in strength, their multiplication and interaction during deformation increase surfacial strength. Microhardness is generally higher in grain boundaries than in grains. With plastic deformation, the strain generally produces a reduction in recrystallization Volume II25 FIGURE 8. (a) Electron image of aluminum rider wear scar; (b) iron Kαmap of aluminum rider. a b Copyright © 1983 CRC Press LLC CHEMISTRYOF SURFACES Clean Surfaces Very clean surfaces are extremely active chemically. Acopper atom which lies in a (111) plane in the bulk of the solid will have a coordination number of 12: it is bonded to 12 nearest neighbors. That same copper atom at the surface will, however, have a coordination number of only 9 with only 9 nearest neighbors. The energy normally associated with bonding to three additional atoms is now available at the surface. This energy expressed over an area of many atoms is referred to as the surface energy. Surface energy is also the energy necessary to generate a new solid surface by the separation of adjacent planes. The energy required for separation is a function of the atomic packing. For example, for copper the atomic packing density is greatest in (111) planes (greatest number of nearest neighbors within the plane). As a result, bonding forces between adjacent (111) planes is least and the surface energy of new (111) surfaces generated, say by cleavage, is less than for the (110) and (100) planes. This lesser binding strength is also a function of the distance between adjacent planes, it being greater between adjacent (111) planes than between (110) and (100) planes. Because surface atoms have this unused energy, they can interact with each other, with other atoms from the bulk, and with species from the environment. Not bound as rigidly as atoms in the bulk, surface atoms can alter their lattice spacing by reconstruction, as depicted schematically in Figure 11. By use of LEED, this process has been found to occur in some crystalline solids but not in others. In solids containing more than a single element, atoms from the bulk can diffuse to the surface and segregate there. In a simple binary alloy, solute atom can diffuse from near surface regions to completely cover the surface of the solvent. This has been observed for many binary systems including aluminum in copper, tin in copper, indium in copper, alu- minum in iron, and silicon in iron. One hypothesis for the segregation mechanism is that the solute segregates on the surface because it reduces the surface energy. Asecond theory is that the solute produces a strain in the crystal lattice of the solvent, and this unnatural lattice state ejects solute atoms from the bulk. Chemisorption In addition to the solid interacting with itself at the surface, the surface can interact with Volume II27 FIGURE 10.The principle types of crystalline binding forces. Copyright © 1983 CRC Press LLC Figure 11. An example is the oxidation of ironin moist air which continues to consume iron. In contrast, oxidation of aluminum to form aluminum oxide results in a thin, dense oxide of 120 Å which retards diffusion and film growth. Environmental Effects Chemical, physical, and metallurgical properties of atomically clean metal surfaces are markedly altered by foreign substances. This is extremely important because most real surfaces are not atomically clean but have film(s) present on their surface (Figure 1). The wide variations found in the literature for surface properties of materials can be attributed to the effect of these films. Presence of oxides on metal surfaces has been observed to produce a surface hardening effect. One explanation for this hardening is that the oxygen pins dislocations which emerge at the surface, impeding their mobility. Other surface films increase ductility. For example, water on alkali halide crystals will allow an otherwise brittle solid to deform plastically. This effect is also observed with ceramics. Magnesium oxide (MgO) is normally very brittle with a surface hardness in the clean state of about 750 kg/mm 2 . Figure 12 presents the hardness of MgO as a function of indentation time in dry toluene and moist air. The increased surface ductility in the presence of water is striking, and the difference increases with increasing indentation time. This change with time makes the film effect a true surface property and not simply a lubricating effect produced by the water. In the 1920s Rehbinder found that certain organic molecules on the surface of solids produced a softening. Such substances as oleic acid in vaseline oil were examined. This surface softening by lubricating substances can be very beneficial in certain instances such as in arresting the formation of fatigue cracks in bearing surfaces. REFERENCES Introduction 1. ASTM,Symposium on the properties of surfaces, ASTM Mater. Sci. Ser. 4, 1963. 2. SCI, Surface Phenomena of Metals, Monograph No. 28, Society of Chemical Industry, London, 1968. 3. Anon., Conference on clean surfaces, Ann. N.Y. Acad. Sci., 101, 583, 1963. 4. Adamson, A. W., Physical Chemistry of Surfaces, 2nd ed., Interscience, New York, 1967. 5. Gatos, H. C., Ed.,The Surface Chemistry of Metals and Semiconductors, John Wiley &Sons, New York, 1960. 6. Blakely, J. M., Ed., Surface Physics of Materials. Vols. 1 and 2, Academic Press, New York, 1975. Volume II29 FIGURE 12. Illustration of the effect of time on microhardness of MgO in tol- uene and in moist air (after Westbrook). 21 Copyright © 1983 CRC Press LLC Method of Characterization of Surfaces 7. Kane, P. F. and Larrabee, G. B., Eds., Characterization of Solid Surfaces, Plenum Press, New York, 1974. 8. Bunshah, R. F., Ed., Technique of Metals Research, Vol. 2, Techniques for the Direct Observation of Structure and Imperfections, Part 2. Interscience, New York, 1969. 9. Blakely, J. M., Ed., Surface Physics of Materials, Materials Science Series, Vols. 1 and 2, Academic Press, 1975. 10. Somoraji, G. A., Principles of Surface Chemistry, Prentice-Hall, Englewood Cliffs, N.J., 1972. 11. Proc. 2nd Int. Conf. on Solid Surfaces. II, Jpn. J. Appl. Phys., Suppl. 2, 1974. 12. Muller, E. W. and Tsong, T. T., Field Ion Microscopy, American Elsevier, New York, 1969. Properties of Surfaces 13. Ehrlich, G., Atomistics of metal surfaces, Surface Phenomena of Metals, Monograph No. 28, Society of Chemical Industry, London, 1968, 13. 14. Hayward, D. O. and Trapnell, B. M. W., Chemisorption, 2nd ed., Butterworths, Washington, D.C., 1964. 15. Ferrante, J. and Buckley, D. H., A review of surface segregation, adhesion and friction studies performed on copper-aluminum, copper-tin, and iron-aluminum alloys, ASLE Trans., 15(l), 18, January 1972. 16. Burke, J. J., Reed, N. L., and Weiss, V., Eds., Surfaces and Interfaces 11, Physical and Mechanical Properties. Syracuse University Press, New York, 1968. 17. Westwood, A. R. C. and Stolaff, N. S., Eds., Environment-Sensitive Mechanical Behavior, Metallurgical Society Conference, Vol. 35, Gordan and Bovach Science Publishers, New York, 1966. 18. Jenkins, A. D., Ed., Polymer Science, A Materials Science Handbook, Vols. 1 and 2, North-Holland, Amsterdam. 1972. 19. Buckley, D. H., Definition and Effect of Chemical Properties of Surfaces in Friction, Wear, and Lubrication, NASA TM-73806, National Aeronautics and Space Administration, Washington, D.C., 1978. 20. Likhtman, V. I., Rehbinder, P. A., and Karpenko, G. V., Effect of Surface-Active Media on the Deformation of Metals, Chemical Publishing Company, New York, 1960. 21. Westbrook, J. H., Ed., Mechanical Properties of Intermetallic Compounds, John Wiley & Sons, New York, 1960. 30 CRC Handbook of Lubrication Copyright © 1983 CRC Press LLC FRICTION K. C. Ludema DEFINITION OF FRICTION The usual engineering definition of friction is resistance to relative motion of contacting bodies. Commonly encountered types of friction include dry, lubricated, sliding, rolling, dynamic or kinetic, static or starting or limiting, internal or hysteretic, external and viscous. Magnitude of friction is usually expressed as a coefficient of friction µ, which is the ratio of the force F required to initiate or sustain relative tangential motion to the normal force (or weight) N which presses the two surfaces together. Thus, µ = F/N. In the early years of technology, the value of F/N was found to be reasonably constant for each class of materials. In modern technology, µ is regarded to be widely variable, depending on oper- ational variables, lubricants, properties of the substrate, and surface films. 1-5 CLASSIFICATION OF FRICTIONAL CONTACTS Friction is a phenomenon associated with mechanical components. Some are expected to slide and others are not. Four categories within which high or low friction may be desirable are given below. 1. Force transmitting components that are expected to operate without displacement. Examples fall in the following two classes: a. Drive surfaces or traction surfaces such as power belts, shoes on the floor, hose clamps, and tires and wheels on roads or rails. Some provision is made for sliding, but excessive sliding compromises the function of the surfaces. Normal operation involves little or no macroscopic slip. Static friction is often higher than the dynamic friction. b. Clamped surfaces such as press-fitted pulleys on shafts, wedge-clamped pulleys on shafts, bolted joining surfaces in machines, automobiles, household appliances, etc. To prevent movement, high normal forces must be used and the system is designed to impose a high but safe, normal (clamping) force. In some instances, pins, keys, surface steps, and other means are used to guarantee minimal motion. In the above examples, the application of a (friction) force frequently produces microscopic slip. Since contacting asperities are of varying heights on the original surfaces, contact pressures within clamped regions may vary. Thus, the local re- sistance to sliding varies and some asperities will slip when low values of friction force are applied. Slip may be referred to as microsliding as distinguished from macrosliding, where all asperities are sliding at once. The result of oscillatory sliding of asperities is a wearing mechanism, some cases of which are known as fretting. 2. Energy absorption-controlling components such as in braking and clutching. Efficient design usually requires rejecting materials with low coefficient of friction because such materials require large values of normal force. Large coefficients of friction would be desirable except that suitably durable materials with high friction have not been found. Thus, many braking and clutching materials have intermediate values of coef- ficient of friction in the range between 0.3 and 0.6. An important requirement of braking materials is constant friction, in order to prevent brake “pulling” and unex- pected wheel lockup in vehicles. A secondary goal is to minimize the difference Volume II 31 Copyright © 1983 CRC Press LLC [...]... of very rough surfaces where some of the roughness may be due to carbide particles, there may be a second component of friction due to asperity collision Laws of Friction The earliest law of friction is due to Leonardo DeVinci (14 52 to 15 19).8 He observed that F is proportional to N, where F is the force to initiate sliding and N is the normal force holding the surfaces together Amontons (16 63 to 17 05),... pressure of contact, N = ArPf, where Pf is the average pressure of contact on the asperities Altogether, the coefficient of friction is taken as Copyright © 19 83 CRC Press LLC 36 CRC Handbook of Lubrication Ss is usually approximately Y /2 where Y is the yield strength of the material in tension Pf is usually no more than 3Y Thus, the ratio Ss/Pf is about 1/ 6, which is not far from 0 .2, a value often found... however, the closest estimate of friction is only 1/ 10 of the measured values Estimation of the real area of contact is generally considered the most difficult problem in this model From 19 38 when the above model was proposed, there have been many developments in technology, particularly in the use of vacuum equipment In vacuum, the coefficient of friction is often seen to exceed 0 .2 by a large margin and... roughness of 10 µin Neither of the Tabor models or the interlocking theory explain the influence of close lateral proximity of asperities which imposes a limit on the high value of µ This is the case in metal working where there is high-contact pressure Copyright © 19 83 CRC Press LLC Volume II FIGURE 1 37 String-pulley-weight measurement of coefficient of friction FIGURE 2 Tilting plane measurement of coefficient... designs are a few of many in use Frequently, it is more convenient to use two flat surfaces, a shaft in a bearing, or three pins instead of one Copyright © 19 83 CRC Press LLC 40 CRC Handbook of Lubrication FIGURE 5 General effect of sliding speed on coefficient of friction for metals and other crystalline solids (e.g., ice) FIGURE 6 Influence of sliding speed on coefficient of friction of a steel sphere... found at which sustained sliding of uniform velocity occurs, tan θ is the kinetic coefficient of friction As technology developed, it became possible to measure the coefficient of friction to a Copyright © 19 83 CRC Press LLC 38 CRC Handbook of Lubrication FIGURE 3 One-piece device for measuring pin-on-flat coefficient of friction Strain gages on flexible sections 1 and 2 measure normal force; strain... coefficient of friction COEFFICIENT OF FRICTION Measurement of Friction Measurement of the coefficient of friction involves two quantities, namely F, the force required to initiate and/or sustain sliding, and N, the normal force holding two surfaces together Some of the earliest measurements of the coefficient of friction were done by an arrangement of pulleys and weights as shown in Figure 1 Weight Ps... adhesion theory of friction are Bowden and Tabor .1 An early model from this school began with the idea that the force of friction is the product of Ar, the summation of the microscopic areas of contact, and the shear strength, Ss, of the bond in that region; i.e., F = ArSs To complete the model, the load, N, was thought to be borne by the tips of asperities, altogether comprising a total area of contact,... makes a total denial of adhesion untenable The influence of a cycle of loading and unloading of a sphere on a flat plate with and without adhesion may be seen in the illustration of a rubber ball pressed against a rigid flat surface As each increment of load is added, a ring of larger diameter of contact forms between the ball and flat plate The reverse occurs upon progressive removal of the load If the... applied load ReCopyright © 19 83 CRC Press LLC 34 CRC Handbook of Lubrication Table 1 COEFFICIENT OF ADHESION FOR VARIOUS METALS search suggests that real contact area between nominally flat surfaces increases more neary as the 0.8 power of applied load.7 Adhesion and Peeling In the above model of the elastic sphere pressing against an elastic flat plate, the radius and area of contact increase as the . helium. It is 20 CRC Handbook of Lubrication FIGURE 3.Planes of possible slip in a cubic crystal; (a) three (10 0) planes, (b) six (11 0) planes, and (c) four (11 1) planes. Copyright © 19 83 CRC Press. adjacent (11 1) planes is least and the surface energy of new (11 1) surfaces generated, say by cleavage, is less than for the (11 0) and (10 0) planes. This lesser binding strength is also a function of. York, 19 75. Volume II29 FIGURE 12 . Illustration of the effect of time on microhardness of MgO in tol- uene and in moist air (after Westbrook). 21 Copyright © 19 83 CRC Press LLC Method of Characterization