The genesis of solid friction

The laws of friction

Friction is the resistance encountered when one body moves tangentially over another with which it is in contact. The work expended against friction is often redundant, that is, it makes no useful contribution to the overall operation of the device of which the bodies are part, and ultimately must be dissipated as waste heat. Consequently, in most tribological designs our aim is to keep these frictional forces as small as possible. Of course there are exceptions to this general rule, occasions when sufficient friction is essential to continued progress and there are many practical devices which rely on the frictional transmission of power: automobile tyres on a roadway, vehicle brakes and clutches, as well as several of the variable-speed transmission systems now finding wider application. When two objects are to be held together, the only alternative to methods which rely on friction is the formation of some sort of chemical or metallurgical bond between them. The development of this sort of technique–adhesives and ‘superglues’, and even welding and brazing–are relatively recent; ‘traditional’ forms of fixing rely almost exclusively on friction. A nail hammered into a piece of wood is held in place by frictional effects along its length; if the frictional interaction were substantially reduced, the nail would be squeezed out. Similarly, the grip between a nut and a bolt depends on adequate friction between them.

Introduction

With a few important exceptions, engineering devices which involve the contact of loaded, sliding surfaces will only operate satisfactorily, that is, without giving rise to unacceptable amounts of surface damage or wear, when they are provided with adequate lubrication. The lubricant can act in two distinct, but not necessarily mutually exclusive, ways. The first of its functions may be to physically separate the surfaces by interposing between them a coherent, viscous film which is relatively thick (i.e. significantly larger than the size of likely surface asperities). In hydrostatic bearings this film is provided by an external pump and so its presence depends on the continuous operation of an external source of energy. In hydrodynamic bearings its generation relies only on the geometry and motion of the surfaces (hence the term dynamic) together with the viscous nature of the fluid. The second role of the lubricant may be to generate an additional thin, protective coating on one or both of the solid surfaces, preventing, or at least limiting, the formation of strong, adhesive and so potentially damaging friction junctions between the underlying solids at locations of particularly acute loading. If this protective coating has a comparatively low shear strength then the ultimate tangential force of friction can be much reduced: this mechanism of friction limitation is generally known as boundary lubrication. Such boundary films are generally very thin, perhaps only a few (albeit very large) molecules thick, and their formation and survival depends very much on the physical and chemical interactions between components of the lubricant and the solid surfaces.

Introduction

All real materials possess some microstructure which becomes apparent when a representative specimen is viewed with a sufficient degree of magnification. However, engineers usually work at a comparatively large scale and so use macroscopic models of material behaviour: these are likely to involve a very large number of the individual microstructural units which contribute to the overall material response. Engineering properties of the bulk represent average or integrated values, and when dealing with either solids or liquids the material is often treated as a homogeneous continuum. In the case of the linear elastic theory of solids what is required is Hooke's law and the observation that most metals undergo only a very small strain before yielding. Similarly, a Newtonian compressible fluid is described by the linear relationships between the applied shear stress and the resultant shear strain rate, and between density and the applied hydrostatic pressure. Many liquid lubricants, including hydrocarbon mineral oils, are effectively incompressible at moderate pressures so that their density remains constant, and this simplifies the analysis still further.

There is no fundamental reason why the lubricant in a hydrodynamic bearing should not be a vapour or a gas (rather than a liquid such as oil or water); indeed considered as a potential lubricant, air has a number of advantages–such as cleanliness and ease of supply–and it has been used in aerostatic and aerodynamic bearings for several decades.

The nature of engineering surfaces

No real engineering surface, no matter how carefully, or indeed expensively, prepared can possess perfect geometry. As well as errors in the form or shape of the component there will always be a roughness on the surface which is apparent when this is examined at a sufficiently high magnification. When two such surfaces are loaded together it is the tips of the surface roughnesses or asperities that must first carry the applied load: the geometry of individual contact spots and the way in which these islands of real contact are distributed throughout the nominal or apparent contact area is clearly of interest to tribologists in attempting to predict the overall performance, or likely life history, of the contact.

The geometric texture of an engineering surface reflects both its production route and the nature of the underlying material. It is possible to produce a truly smooth surface (for example, cleaving specimens of mica can produce a surface with roughness only on the atomic scale) and if two such surfaces are loaded together real and apparent areas are very nearly equal. The asperities on the surface of a very compliant surface, such as a soft rubber, may, if sufficiently small, be squeezed flat by quite modest contact loads, and in this way there can again be equality between real and apparent areas of contact. However, these are special cases; in general, useful metal surfaces exhibit a range of surface fluctuations which, although large compared to molecular dimensions, are small compared to the dimensions of most engineering components.

The term tribology is scarcely twenty-five years old and yet there can be few university or college courses in mechanical engineering which do not now include material under this heading. Of course, the problem of producing bearings, slides, seals, and other tribological systems to give smooth machine running and long component lives is one which has faced practitioners for generations, and consideration of their design has always played a part in the education of mechanical engineers. What has become increasingly obvious in recent years is the inherently interdisciplinary nature of the tribologist's task; as well as involving practising and academic engineers, advances in the subject have drawn upon the ingenuity and expertise of physicists, chemists, metallurgists, and material scientists. Consequently, although envisaged principally for use by final year undergraduates and post-graduate students in mechanical engineering, I hope that this volume may be of interest to students and specialists in these other related areas. Tribology is still very much an area of active research and the published literature in the fields of lubrication, friction and wear–already dauntingly voluminous–continues to grow at an alarming rate. I have made no attempt to produce a research monograph but rather to provide a framework of fundamental analytical tools which can be used in a wide variety of different physical situations. Each chapter is concluded with a short list of suggestions for further reading which provide access to the more specialised literature.

Introduction

The satisfactory operation of both hydrostatic and hydrodynamic bearings requires that the solid surfaces which constitute the bearing faces are completely separated by the intervening fluid film. Since the bearing surfaces are then not physically touching, the resistance to their tangential motion, that is, the force of friction, is directly attributable to viscous losses in the lubricant. If the lubricant (whether liquid or gas) exhibits Newtonian rheological behaviour with constant viscosity then the value of this frictional force, and the associated coefficient of friction, will increase with the value of the tangential sliding velocity. We have seen (eqn (7.27)) that the coefficient of friction within a hydrodynamically lubricated bearing is generally dependent on the square root of the group ULη/W where U is the relative sliding speed of the surfaces, W/L the normal load supported per unit length, and η the Newtonian viscosity. A reduction in speed, or an increase in the specific load on the bearing, leads to a fall in the friction coefficient. However, there is a limit to this process: when the specific load is very high, or the relative sliding speed small, it is difficult to build up a sufficiently thick film to entirely separate the bearing faces, and so there will be some mechanical interaction between opposing surface asperities. This is inevitable, even allowing for the large increase in effective lubricant viscosity and the elastic flattening of the surface profiles that can occur in the elasto-hydrodynamic regime.

Introduction

The definition and measurement of wear

Wear is the progressive damage, involving material loss, which occurs on the surface of a component as a result of its motion relative to the adjacent working parts; it is the almost inevitable companion of friction. Most tribological pairs are supplied with a lubricant as much to avoid the excessive wear and damage which would be present if the two surfaces were allowed to rub together dry as it is to reduce their frictional resistance to motion. The economic consequences of wear are widespread and pervasive; they involve not only the costs of replacement parts, but also the expenses involved in machine downtime, lost production, and the consequent loss of business opportunites. A further significant factor can be the decreased efficiency of worn plant and equipment which can lead to both inferior performance and increased energy consumption.

The wear rate w of a rolling or sliding contact is conventionally defined as the volume lost from the wearing surface per unit sliding distance; its dimensions are thus those of [length]. For a particular dry or unlubricated sliding situation the wear rate depends on the normal load, the relative sliding speed, the initial temperature, and the thermal, mechanical, and chemical properties of the materials in contact. There are many physical mechanisms that can contribute to wear and certainly no simple and universal model is applicable to all situations.

Tribology

The word tribology was coined only just over twenty years ago and appears in only the most up to date of dictionaries; however, the topics with which tribologists are concerned have been of vital interest to scientists, engineers, and those who design or operate machinery, for as long as mechanical devices have existed. Formally, tribology is defined as the science and technology of interacting surfaces in relative motion and of related subjects and practices; it deals with every aspect of friction, lubrication, and wear. The word is derived from the Greek τριβοσ (TRIBOS) meaning rubbing, although the subject embraces a great deal more than just the study of rubbing surfaces.

Perhaps as much as one third of our global energy consumption is consumed wastefully in friction: at a time when energy resources are at a premium the contribution that can be made to their efficient utilization, as well as to the reduction of pollution, by making use of the best tribological practices is obvious. In addition to this primary saving of energy there are very significant additional economies to be made by reductions in the cost involved in the manufacture and replacement of prematurely worn components. An important landmark in the development of the subject was the publication in 1966 in Great Britain of the report of the government committee, chaired by Mr Peter Jost, which had been formed to report on the position of industrial lubrication in the United Kingdom: it was asked specifically to identify those areas of industrial practice where significant improvements could be made.