Review and propositions for the sliding/impact wear behavior in a contact interface

Yunxia CHEN, Wenjun GONG, Rui KANG

School of Reliability and Systems Engineering, Beihang University, Beijing 100083, China

KEYWORDS Contact interface;Coupling mechanism;Life prediction;Sliding/impact wear behavior;Wear equation

Abstract A thermal-solid-liquid complex operational environment induces structural interface developing a typical coupling sliding/impact wear behavior. It results in contact damage until systems fail, which may cause significant economic losses and catastrophic consequences. The key point of solving this problem is to reveal the coupling damage mechanism of the sliding/impact behavior in typical systems and life characterization under a complicate evolving environment.This has been a hot topic in the area of mechanical reliability.The main work in this paper can be concluded as follows. Firstly, the main industries in which the ‘‘sliding/impact behavior” takes place have been introduced. Then, existing studies on the wear mechanism and degree analysis are presented,which includes surface morphology analysis,wear debris analysis,and wear degree measurement.Meanwhile,existing problems in theoretical modeling and experiments in current research are summarized, so as to point out a bright direction for future research on wear prediction. They include interface contact modeling, mathematic coupling mechanism modeling, wear equation establishment, and wear life characterization, which can provide some new ideas for improving the existing studies on the sliding/impact wear behavior.

1. Introduction

Wear is a typical failure behavior in a mechanical system,causing replacements of numerous components in many fields. It can induce a direct or indirect economic loss up to several billion dollars. Based on the definition of wear, an obvious occurrence place of wear is the contact interface of structural motion systems. It basically refers to a combination of two or more components through an interface,such as a hydraulic spool,a mechanical transmission system,etc.Under a complex load, a typical dynamic behavior with different forms of motion like sliding and shock can occur spontaneously in the contact area of structural motion parts,which leads to surface damage.The load’s complexity is concluded into two terms:(1)multi-phase coupling conditions such as heat,solid,and liquid given in a design environment;(2)high-speed operation or various external environments in an actual operation process leading to the uncertainty of the contact interface’s damage mechanism. As a typical representative of mechanical equipment,structural motion systems have been widely used in fields including aviation &aerospace, nuclear plants, transportation industry, precision machine tools, deep drilling mining, and other areas. The frequent occurrence of oil leakage, getting stuck,and other typical failures caused by sliding/impact wear can generate some more serious accidents like air parking,fire,and explosion. It urges us to study the sliding/impact wear behavior timely, systematically propose solutions and methods, and finally form a complete theory and protective technology.

Fig. 1 Existing studies and our propositions of sliding/impact wear behavior in the contact interface.

The purpose of studying the sliding/impact wear behavior is to grasp the wear characteristics of materials, and optimize and select the particular application of a material for controlling the wear behavior during an operational period.Thus,the core scope of studying the sliding/impact wear behavior includes two terms: (1) analysis and measurement for investigating the wear mechanism and degree; (2) quantitative wear prediction modeling.At present,the main research of the sliding/impact wear behavior is mostly focused on the first part for investigating the specific wear mechanism and degree.Qualitative methods consist of surface morphology analysis and debris characteristic analysis.Meanwhile,corresponding wear degree measurements are recorded to quantitatively reflect the wear degree, such as the wear depth, wear volume, mass loss, etc.These studies just provide a good understanding of the sliding/impact wear behavior. However, it is not enough for controlling sliding/impact wear in a reasonable range. Modeling for a quantitative wear prediction is the next content we need to consider.

Thus, in this paper, we outline some of the existing results and ideas behind current research.Then,we point out deficiencies and try to give some propositions about how to make a reasonable wear prediction model from the experience of progresses in other areas that may be helpful for wear control as Fig. 1 shows. The interface wear is mainly due to the interaction of the micro-contact of a gear meshing interface. Therefore, the corresponding interface contact theory is an important basis for exploring the wear mechanism. Based on the contact theory, long-term engineering experimental research is needed to deepen the understanding of the essence of the wear mechanism and reasonably describe the coupling mechanism of sliding-impact wear in a mathematical form.Then, a possible way can be found to provide an effective and accurate wear formula for quantitative prediction. In addition,for a specific system,a reasonable wear life characterization model should be established to understand the relationship between the wear degree and the entire system function or performance degradation. Thus, from our perspective, a reasonable wear prediction model for the sliding/impact wear behavior should include four terms:(1)interface contact modeling; (2) mathematic coupling mechanism modeling; (3) wear equation establishment; (4) wear life characterization. These above four terms can provide designers effective tools to calculate the sliding/impact wear degree of a system under different operational conditions by choosing different materials and changing the structure size and process parameters. It can guide designers to control the sliding/impact wear into a reasonable rang under an operational condition to satisfy the design requirements, which can be regarded as the optimization of structural design parameters.The final section will discuss these four terms, which helps further clarify the direction and future development trend of the follow-up research work.

2. Main fields & forms

The sliding/impact behavior is that sliding in a range of distance is accompanied by a normal impact in the contact interface. The main failure mode of the coupling behavior is wear,which is defined by the American Society for Testing and Materials (ASTM) as ‘damage to a solid surface, generally involving the progressive loss of material, due to relative motion between the surface and a contacting substance or substances.’ Commonly, the sliding/impact wear is a kind of typical wear process under the condition that the tangential displacement under the normal impact is greater than 100 microns.1Through a summary of the object of the wear behavior,we can understand that common characteristics of performance include high speed, high frequency, high-energy sliding shock, and instantaneous energy transfer and wear damage,which represents a special form of wear failures.As the abovementioned wear behavior, it mainly occurs in the aviation,aerospace, defense, and other areas of high-end products and civilian products. Some details are presented in the following content (Table 1).

Table 1 Fields and objects of the sliding/impact wear behavior.

In the aviation&aerospace fields,high speed,temperature,and heavy loading with high-frequency vibration make a friction pair generate some serious wear damage. The sliding/impact wear is one of the typical wear patterns (see Fig. 2).For instance, a high-speed bearing for supporting a small turbojet engine would have an obvious wear on the contact surface. This kind of severe abnormal wear between the bearing’s guide face and holder is mainly caused by holder imbalance, engine spindle vibration, and fuel injection reaction. It finally makes the bearing fail in a very short time.1The main wear areas are the inner guiding land and the circular raceway of the outer ring.During the development of space shuttles, researchers at both NASA2and ESA3found that a situation has evolved in which all bearings placed into service failed within a few hundred seconds of operation for no apparent reason. Development of space shuttles was delayed, and significant cost increases were incurred because of the inability to resolve the abnormal wear phenomena. Based on the fault tree analysis, different parameters are analyzed to guarantee that the wear volume can be decreased. As potential failure components, engine sealing components have the same wear problem to cause fuel leakage, which may bring catastrophic accidents like in-flight shutdown and fire breaking out.1In addition, surface coatings for increasing hardness as well as corrosion, thermal, and wear resistances are used for many applications in planes and spaceships.4As a protective coating to improve superficial quality of a component, the fatigue cracking-induced sliding/impact wear under localized contact stresses during a milling operation makes it not work very well.5Therefore, there is a need for a testing method that can study wear caused by repetitive sliding/impact motions. Some tests are conducted to investigate different wear characteristics of surface coatings with different materials.

Fig. 2 Components with sliding/impact wear in aviation & aerospace.

Fig. 3 Sliding/impact wear of rod cluster control assemblies in nuclear plants.

In nuclear plants, a specific wear of rod cluster control assemblies of pressurized water reactors as shown in Fig. 3 can usually happen,which is mainly caused by relative motions due to operation processes or flow-induced vibration. It can also be identified as sliding impacts combined with a typical environment, high temperature, high pressure, and solution chemistry lead. The origin of this specific wear is not really understood today. Many experiments on these components have mainly focused on wear damage. Physical parameters of wear such as contacting bodies,environment,and dynamics,in varying test duration, are discussed in some literature.6-9For instance, Van Herpen et al.9studied the sliding/impact wear behavior caused by the vibration of the nuclear power plant piping system, and the wear test was carried out with 304L steel as the test material. It was considered that the sliding/impact wear behavior was very special, not a simple stack of both actions. The severe damage to contact zones has been evaluated by profilometry and weighting. Significant changes of the material surface micromechanical states have been measured, but they cannot be directly related to the test duration time.

In the transportation industry,commercial vehicles,trucks,and trains are the most frequent equipment for conveying goods. However, their typical operational environment makes sliding/impact wear often happen, which affects the normal operation of equipment.For instance,the exhaust valve system of a truck engine experiences a very complex contact situation including frequent impact involving micro sliding, high and varying temperatures, complex exhaust gas chemistry, and possible particulates.10-13While the engine is in operation,the valve’s seating face makes a direct contact with the seat insert’s seating face (see Fig. 4(a)). The contact between the two seating faces causes wear, and the wear of both faces reduces the sealing ability between the valve and the seat insert. Chun et al.11investigated the wear of the seating faces depending on the cycle number and operational frequency.While the cycle number increased, the wear volume represented by the average roughness increased linearly. Besides,the valve’s closing velocity had a greater influence on the average roughness than that of the cycle number (mileage). Moreover, the effect of high temperature was also investigated by using a new impact-sliding tribometer.12Another case is the wear caused by the contact of wheel/rail in a rail way system(see Fig.4(b)).As we all know,rolling of the wheel is the main operation pattern, but in traction, braking, and turning, the sliding wear corresponding to the impact between the wheel and the rail plays a significant role on the failure process of the wheel/rail system.14With an increase of the sliding speed,the wear rate of wheel steel decreases and then increases.15At present, scholars are still in the early stage of this research,without any significant results.

The wear pattern of some components of precision machine tools is a combination of sliding and impact wear,1such as rails of machine tools (see Fig. 5), etc. Main rail wear factors include the hardness of the rail material, carrying weight,sliding speed,working conditions,and other aspects.16Therefore,a study of sliding/impact wear should be taken thoroughly to support the advancement of manufacture technology.In addition,there are a few components like clutch and conjugate cam in textile machinery that would also undergo the sliding/impact wear process.17The gap between friction pairs caused by the wear makes the accuracy decreased, which results in an additional dynamic load and vibration and noise of the loom, affecting the normal work of the loom.

Fig. 4 Sliding/impact wear cases in transportation systems.14

Fig. 5 High-frequency reciprocating servo-guide of a precision machine tool.

Fig. 6 Drilling systems of deep & ultra-deep wells.

During drilling of deep & ultra-deep wells, casing wear problems cannot be neglected.18-22Considering the operational condition, the sliding/impact wear behavior is usually found in actual failure cases as Fig. 6 presents. It can affect a drilling work,which brings a tremendous cost loss. The current casing wear theory is based on the operational condition of static loading and sliding, and it cannot effectively explain the sliding/impact wear behavior. Therefore, it is necessary to focus on the study of sliding/impact wear to solve existing problems. In addition, coal mine work also faces the same problem. Coal mine machines should run in a long time without lubrication under harsh environment conditions with different geological rock layers. As a result, wear failure of coal mining machinery is extremely serious,and various wear forms are common in down-hole equipment for mining,cutting,shipping,anchoring,crushing,and coal processing.A lot of scholars have done many research on the impact wear. However,most of the mechanical equipment undergoes impact wear with sliding friction,causing the surface’s complex damage,and the wear mechanism is still not very clear.23

Fig. 7 Sliding/impact wear process of dental material.

In some other areas, there are also a few reports about the sliding/impact wear phenomenon. Gold-plated contact terminals are applied into a lot of micro electrics like connectors and relays. Because of the existence of a super stroke motion,a corresponding sliding wear process with an impact from an outside mechanical force would happen during an operational period. Chen et al.24specifically investigated the effect of the sliding and impact wear on the contact of a relay.Their experiments showed that the coupling sliding/impact wear happened in normal opened contact areas.Owing to the surface material being removed, the base material may be exposed outside,making some other mechanisms like oxidization and sulphidization.All of these behaviors cause the contact resistance to change and then a contact failure.In the medical field,dental restoration is necessary to guarantee teeth’s normal working (see Fig. 7). Based on actual observation, the wear on the occlusal surface of teeth results from the impact contact and sliding friction of teeth.The wear volume of tooth samples increased nonlinearly with the number of impacts. Therefore,it is necessary to study the teeth sliding/impact wear behavior to comprehensively understand the wear properties of teeth and dental materials. It can also provide theoretical guidance and reference for clinical practices.25

3. Wear mechanism & degree analysis

Tests for studying the wear mechanism & degree are mainly related to the type of material specimen applied in different engineering scenarios with distinguished loading and environmental conditions. To investigate their wear properties under some specific loading conditions, test devices of simulating the sliding/impact wear behavior are emerging with the development of mechanics and measurement techniques. From these test devices,the intentions of tests can be easily classified into two terms: (1) try to investigate the wear property of the material; (2) simulate the possible environment condition to understand the wear mechanism and degree. Response variables and control variables in literatures are summarized in Table 2. They include qualitative and quantitative measurements. The qualitative measurement includes the surface morphology and debris composition analysis. The quantitative measurement includes the surface roughness & wear volume measurement.

3.1. Surface morphology analysis

Surface morphology as an instant evidence can reflect different variations which can distinguish different wear mechanisms like fatigue,adhesion,abrasion,and corrosion.With the deepening of research on wear mechanisms,many experiences help us to conclude the existing mapping relationship between worn surface features and wear mechanisms.For example,the abrasive wear mechanism lies in the plastic deformation and themicro-cutting mechanism caused by an abrasive on the surface,and the worn surface will have typical furrows and micro-cutting scratches. A common feature of adhesive wear is the material migration.Along the sliding direction,different smear-like adhesive layers will be formed.With increases of the load and sliding distance,the smear layer changes from thin to thick; then an occurrence of more serious phenomena would happen, such as glued shearing off. Fatigue wear is due to the surface or subsurface crack growth with an angle of 10-30 centigrade,and pitting and spalling are two kinds of typical features of fatigue worn surface. Pitting cracks generally start from the surface, expand inward, and return to the surface,causing fan-shaped pits; fatigue spalling cracks start from the subsurface, expand along a direction parallel to the surface,and finally form flakes. Thus, many scholars have utilized the observation and analysis of the worn surface morphology to make a preliminary judgment of different wear mechanisms in the study of the sliding/impact wear behavior.

Table 2 Sliding/impact wear test categories.

Fig. 8 SEM observations of wear damage with impact times.26

In 1988,Shi et al.26conducted a sliding/impact wear test of a ring-shaped mold material at a high temperature level. The worn surface morphology was observed by the scanning electron microscope at different wear stages. Actual damage patterns at different impact times are shown in Fig. 8. In the initial stage (cycle number=1, 10), furrow and microcutting are mainly observed,with a small proportion of a thin adhesion layer and spot fatigue spalls; with impact times increasing (cycle number=100, 500), they gradually transfer to point or flake fatigue spalls; besides, adhesive wear corresponds to other wear mechanisms in the whole wear process;in the final stage (cycle number=1000, 3000), a large block of fatigue and thick adhesive layers becomes more and more obvious. Therefore, the wear mechanism changes from abrasive to fatigue wear, along with accumulated adhesive wear when the impact times increase. It can be seen that the dominant sliding/impact wear mechanism at high temperature changes with impact times.

In 1991, Ding27firstly investigated the wear mechanism of No. 45 steel applied in textile machinery. He found that the main wear mechanism observed by SEM could be concluded as furrow caused by adhesion and spall caused by fatigue,and different materials could affect the duration time of each wear mechanism.

Fig. 9 SEM observations of wear damage of No. 45 and T10 steels.17

In 2000, Yang and Peng17investigated the wear mechanisms of T10 and No. 45 steels caused by sliding/impact loading (Fig. 9). The wear of the materials was carried out simultaneously on the surface and the sub-surface layer, and the wear of the sub-surface layer was mainly caused by the impact, which depended on the shear strength of the metal on the surface layer.The surface layer wear was mainly caused by sliding. The pulsating impact wear mechanism has both plastic deformation caused by fatigue and chip furrow caused by adhesion.Under an impact load of 100 N with dry friction,the main factors of wear are mainly adhesive and abrasive wears.Under a high load up to 1000 N and a dry friction condition, the wear mechanism is mainly fatigue flaking and a small amount of abrasive wear.

Then in 2001,Van Herpen et al.9investigated the wear tests of control rods and guide cards in a nuclear power plant which were made from 304L stainless steel. Several techniques were used that included a descriptive analysis based on detailed examinations of worn surfaces (using SEM and twodimensional profilometry) and X-ray diffraction (XRD) analysis of the microstructure and the micromechanical state of the surface layers of the material. The brown color of wear tracks can be related to the existence of an oxide layer (see Fig. 10). This oxide layer was then subjected to further wear cycles and fragmented. Meanwhile, the shapes of wear tracks were extracted, and all unidimensional profiles were plotted for the worn positions of the tube, namely the areas where the maximum number of impacts occurred have been scanned.These curves attest to a real wear process increasing with test duration.After each wear test,a non-destructive XRD analysis was made on different parts of all oxidized areas. An alphamartensite was induced from austenite by plastic deformation.All these results indicated that the most likely wear mechanism of the worn areas seemed to be oxidation of the material.It can be described as a continuous sequence of formation, detachment, and ejection of oxides on the specimen surface. Then,some other similar tests6,7were finished to verify the conclusion.

Another typical example is the valve/valve-seat under a sliding/impact motion which has been studied by many scholars.In 2001,Zhao et al.28,29studied the wear characteristic of a valve/valve-seat by an observation of the worn surface with a scanning electron microscope. It consists of fatigue spall caused by repeated elastic-plastic deformation, creep sliding of the sub-surface metal, and oxidation, which has been verified by another test result.11Forsberg et al.13further analyzed the wear mechanisms variation of an exhaust valve system. It was found that the wear mechanisms were a combination of oxidation and adhesive wear, which generated a tribo film to protect the underlying surface from wear (see Fig. 11).

In 2005, Wang et al.30-32investigated the coupling sliding/impact wear behavior existing in GCr15/GCr15, GCr15/45,and Q235 friction pairs applied in vibratory bearings. Their worn surfaces were separately observed using an optical microscope. With increasing running cycles, the surface generates oxidation and annealing, which makes the oxidation layer spalling and pitting. The peeling material is adhered on the surface to be a protect film by the impact, which results in low roughness with a large amount of wear.

Fig. 10 SEM observation of wear damage of 304L steel.9

Fig. 11 Wear process of a valve/valve-seat under sliding/impact motion13

Fig. 12 Typical wear scars of five friction pairs by SEM.33

In 2010, Wei33separately studied the sliding/impact wear process of three friction pairs including 20CrNiMo/GCr15,QAl10-4-4/TC4,and 2Al1/GCr15 that were applicable for support bearings and two pairs for mechanical sealing that were QAl10-4-4/9Cr18 and M205D/GCr15 (see Fig. 12). Scanning electron microscopy, two- and three-dimensional surface profile tests, energy spectrum analysis, and X diffraction analysis were used to analyze the microstructure and composition of the worn surface to determine the wear mechanism. When the yield limit of a specimen is high, the material is mainly peeled off under sliding/impact action, and an increasing impact frequency aggravates the spall degree. As the friction coefficient increases, a crack tends to happen on the surface.When the yield limit of a specimen is low, the material mainly undergoes micro-cutting, abrasive wear, and adhesive peeling.When an impurity substance like quartz sand is plunged into the friction of the contact interface, particles are pressed into the metal under impact that makes strong plastic deformation.Then with the sliding process,the protruding part is peeled off and washed away. It results in a hardening layer that cannot form a certain thickness as a protective film for the metal.Thus, the wear mechanisms are a combination of abrasive and adhesive wears.

3.2. Wear debris analysis

Fig.13 Microstructure of wear debris of a casing under different load conditions.18

As an important information carrier in a wear process, wear debris can also reflect different types of wear mechanisms and wear severity. The debris analysis technology has been developing for 78 years since the first application in 1940s.At present,the two-dimensional analysis of debris characteristics has been relatively mature, and have entered the practical stage. According to two-dimensional images of debris using the SEM, wear mechanisms can be clearly identified. For instance, fatigue wear can generate spherical debris; cutting particles are produced by abrasive wear. Besides, an element analysis of debris by the EDS is another way to analyze the wear process and the severity of wear. Therefore, some scholars considered using these two methods to study the debris characteristic of the sliding/impact wear behavior.

Yu18studied the sliding/impact wear behavior in an S135/P140 friction pair (see Fig. 13). Optical microscope and scanning electron microscope were used to observe and analyze the microstructure of the wear debris of the casing under different load conditions.Typical spherical particles were seen on the worn surface.It showed that the wear of the casing has changed from abrasive wear to fatigue wear at this time.In addition, with the impact frequency f increasing under a fixed impact force F, the fatigue wear debris increases strongly.

Ramalho et al.12studied the effect of temperatures up to 400 centigrade on the impact-sliding of valve-seat contacts.The wear surfaces of a cylinder after a wear test revealed adherent material, especially when the test was carried out at high temperature. Analyses by EDS of the debris in selected areas of these adherent regions revealed a material rich in Cu,Mo,and W,which concluded that this material came from the plates, probably by a retransfer process. Thus, the wear was governed by adhesion and oxidation.

Fig. 14 Microstructure of wear debris of a casing under different load conditions.6

Reynier et al.6found that randomly larger crystals were scattered on the top of a wear surface when studying the impact/sliding wear damage of control rods and guide cards in nuclear power plants (area 1 of Fig. 14(a)). These crystals are mainly octahedral magnetite Fe3O4(or maybe with a non-stoichiometric structure NixFe3O4) whose chemical composition has been investigated by EDS analysis (Fig. 14(b)).The oxygen content of this crystal is around 40 at.%, and the iron content is about 46 at.%. This oxidation occurs also in the unworn areas of test specimens.

3.3. Wear degree measurement

Wear degree measurement is common in an experimental study of the sliding/impact wear process. It can easily reflect the wear severity on a contact surface under different test conditions. Indications include the wear depth, the wear volume,the mass loss, and so on. For example, in studies of nuclear power plants, A Van Herpen et al.9investigated the effect of test duration on the sliding/impact wear damage of 304L stainless steel at room temperature.Severe damage to contact zones have been evaluated by profilometry and by weighing after wear tests. Significant changes of the material surface micromechanical state have been measured, but they cannot directly be related to the test duration. Wear tests that were conducted by Reynier et al.6included a long test with a 0.2-Hz frequency and a short test with an 18-Hz frequency. The short test showed more severe than the long test.One observes that the wear scar maximum depth of the short test is about 25 μm while that of the long test is around 15 μm. These two records attest to a real wear damage. Moreover, the variation of the mass loss of the slowest specimen is much higher than that of the fast one. It shows that test duration has no effect on the behavior of a material even if wear damage continues to progress. In a test of valve-seat contacts,12the amount of material removed by wear was evaluated by optical profilometry, and the values of the wear volume rise significantly by increasing the test temperature. When the cycle number was increased, the average roughness of the valve and seat insert seating faces increased linearly.The test frequency(valve closing velocity) had a greater influence on the average roughness than the mileage. Wang et al.32investigated different factors that affect the wear behavior of GCr15,which include the sliding speed, impact force, impact frequency, and vibration range.The sliding speed increases the wear volume;the impact force firstly increases the wear volume and then decreases the wear volume; the impact frequency makes the wear volume increase fast; an increase of the vibration range decreases the wear volume. Besides, lubrication can also significantly affect the sliding/impact wear behavior.Lina et al.7studied the influence of water flow on the impact/sliding wear of pressurized water reactors. Two values of the water flow rate (0 and 0.4 kg/s) were compared by investigating the wear of specimens.In the case without flow, mass variations were very low. On the opposite, a significant mass decrease was noticed when experiments were conducted with a water flow. Continuous test results with different flow rates(0.2,0.4,and 0.6 kg/s)further showed that the wear volumes of specimens are slightly increased by the water flow. However, in some other areas,the conclusion would be totally different. For example, Yang et al. investigated the sliding/impact wear of T10 steel under the conditions of dry friction and 20# oil lubrication. Results reflected that the wear volume under dry friction was much greater than the case under oil lubrication.

From these above studies, we can see that an experimental study of the sliding/impact wear behavior can only qualitatively analyze the existence of the wear mechanism, or give a separate relationship like negative or positive between some characteristic parameters and the wear degree, but it cannot reflect the coupling process of various factors.The corresponding numerical simulation work mainly focuses on the analysis of the contact behavior,without the actual working conditions of a friction pair and the latest development of the interface contact theory.34From systematic viewpoints, we need to further discuss the coupling mechanism of the sliding/impact wear behavior, the coupling process of structural parameters and the dynamic behavior, the quantitative prediction of the wear degree, and the characterization of the wear life. Therefore,some future work has to be done in these areas.

4. Propositions in future work

The study on the coupling sliding/impact wear behavior of structural systems is very important. However, how to further study is a still tough work that needs more discussion. Combining the current research, we can make some propositions for future work.It needs to be carried out in the following four aspects: (1) a sliding/impact contact model; (2) a mathematic form of the sliding/impact coupling mechanism; (3) a sliding/impact wear equation; (4) a wear life characterization model.These above four terms can provide designers effective tools to calculate the sliding/impact wear degree of systems under different operational conditions by choosing different materials and changing the structure size and process parameters.It can guide designers to control sliding/impact wear into a reasonable rang under an operational condition to satisfy design requirements, which can be regarded as the optimization of design parameters, as Fig. 15 shows.

Fig. 15 Design parameter optimization framework for the sliding/impact wear behavior.

4.1. Contact theory for sliding/impact wear

Since sliding/impact wear is mainly caused by the interaction of microscopic contact, the contact interaction of the wear surface is an important part of the wear process. Therefore,rough surface characteristics and contact mechanics are the first content to understand the sliding/impact wear phenomenon,and thus become a specialized research field.In earlier studies, different assumptions, such as spherical,sinusoidal, or triangular asperity, were used to characterize rough surfaces.Because of its simplicity of analyzing the influence of asperity,this method is still used in some studies.Then,people gradually realize that a rough surface is a complex system. Different interface contact theory assumptions cause different applications, and as a basis, generate different contact models applicable for various contact surfaces. These theories can help us to find a reasonable method for the establishment of a sliding/impact wear model.

The classical Hertz contact theory establishes the theoretical basis of contact mechanics,which is widely used to describe the normal contact force-deformation relation, contact area,and pressure distribution characteristics of the contact interface of a structural system.35However, due to its neglect of the physical characteristics of material deformation and some unreasonable assumptions, it results in a lack of credibility in describing the surface of an actual structure.36-38In an actual working environment, the working surface of a friction pair has different degrees of roughness due to mechanical processing, working conditions, and other factors.39,40On rough surfaces, the actual contact area is much smaller than the entire roughened surface area, and the contact surface is only a few points or a small face.41

The statistical contact theory is a developed interface contact theory in further consideration of contact macroscopic information. Under the assumption of a rough interface,Green and Williamson42proposed the earliest statistical rough surface contact model (GW model). The GW model is based on a number of conditional assumptions, including that the height of the surface roughness is consistent with a Gaussian distribution, and the contact of each rough body conforms to the Hertz elastic contact theory. On the basis of these assumptions, Greenwood and Tripp,43Whitehouse and Archard,44Tsukizoe and Hisakado,45and Bush et al.46have proposed different elastic contact models by considering different geometric shapes and non-Gaussian distributions.A linear relationship between the contact area and the applied load is obtained and validated by the later numerical method of researchers.47,48In addition,due to the difference between contact stress and material properties, researchers have further proposed a contact model for describing the plastic deformation of a rough interface49,50and an elasto-plastic deformation contact model.51,52

The fractal contact theory was established for the reason that scholars realized that wear surface feature parameters obtained based on the statistical contact theory are obviously limited by the influences of the resolution and sampling length of an instrument. Rough surface self-similarity and scaleindependent properties allow the fractal geometry function to provide all the surface feature information on the fractal surface in all scale ranges.53Thus, the fractal contact theory is more suitable to characterize a rough surface. Majumdar et al.54,55established an M-B model by introducing the fractal geometry function into the engineering field,making the rough surface fractal model gaining widespread attention.Blackmore and Zhou56proposed a model describing an isotropic rough surface; Yan and Komvopoulos57proposed a threedimensional fractal mechanics theory to describe the elasticplastic contact of a rough surface; some other corresponding theoretical research and engineering application have also received more and more scholars’ attentions.58-62In addition,an adhesive effect is proposed owing to the fact that the existing wear theory is limited by the observation scale, which is different from the classical Hertz contact theory.63Proposed elastic contact theories considering the adhesive effect mainly include the JKR theory and the DMT theory, which are used to predict the adhesion force between an elastic sphere and a smooth plane.64By analyzing these two theories, Tabor pointed out that the two theories could be applied to these two extreme cases, and a dimensionless number μ was proposed to reflect the ratio between the elastic deformation energy and the surface energy.Johnson and Greenwood65then proposed an M-D model of a cylindrical surface contact,while Wu66also used the Lermard-Jones potential law to analyze adhesion problems between cylindrical surfaces. Barthel67discussed the development of elastic contact mechanics for smooth surfaces in a subsequent study. Adhesive contact theories on rough surfaces were further developed on the existing basis. Combining statistical functions and fractal geometry functions reflecting the surface roughness, more and more adhesive contact models have been proposed, which are suitable for the wear of tribological systems with different interfaces and functional requirements.68,69

From this above reviewing work, a series of contact interface models including spherical contact interface, rough contact interface, and sliding contact interface is introduced.The main process of an interface contact model is based on the micro-mechanical contact theory,which combines different types of interface characteristics and interface contact behaviors. When they are utilized further for establishing a contact model for the sliding/impact wear, some discussions should be done to prove their availability.

4.2. Analysis of the coupling mechanism

The sliding/impact wear is distinguished by test results. Some scholars have been making a qualitative investigation for the actual damage mechanism. From references, we can see that wear is a surface phenomenon involving a complex process that is affected by the physical,mechanical,and chemical characteristics of opposing elements, the medium, and the surrounding environments. It says that the sliding/impact wear includes two or more wear mechanisms with coupling relationships.Coupling usually refers to the interrelationships between the elements of a system, between systems, and between systems and environments. The coupling action is usually transmitted by coupling variables, either unidirectional,bi-directional, or random.70The coupling mechanism is focused on exploring the natural description and mathematical characterization of the coupling relationship between different mechanisms. Because there are many differences between different types of mechanisms, for a simple measurable coupling relationship,it can be directly verified by experimental testing;for a multi-dimensional coupling relationship, we only use indirect detection means to verify it.Therefore,this characterization can be artificially set or automatically evolved, which should be explained specifically.

At present, the coupling mechanism is ubiquitous. For example, the corrosive fatigue crack propagation of a metal is significantly different from that of conventional fatigue.Through experiments,it is considered that cyclic loading,geometric characteristics, and metallographic and environmental factors have important effects on the corrosive fatigue crack propagation.Wei and Simmons71utilized a fracture mechanics method to characterize the crack propagation rate; a few superposition models with different correction forms were proposed under different loading conditions with corrosive environment. Austen and Mcintyre72argued that mechanical fatigue and stress corrosion were not a simply superposition in some cases, but rather a faster evolution process represents the corrosion fatigue crack propagation process,which we call the competition model. In addition, fatigue and creep, from the microscopic point of view,the microporous growth caused by creep damage and the internal cracks caused by fatigue damage have a coupling interaction, and this effect shows a nonlinear cumulative feature. Zhang73analyzed the creepfatigue coupling damage characteristics of aluminum alloy,and used combined cumulative damage models to describe the coupling characteristics of these two mechanisms. In addition,some scholars74thought that the impact of creep on fatigue life performance was also related to the frequency.Therefore, by introducing a frequency correction factor in the Coffin-Manson or Morrow model describing the low-cycle fatigue life, a new model could be established to describe the interaction between creep and fatigue.

The study of the above coupling mechanism can be summarized as superposition, competition, combination, induction,etc.,75,76which gives us a specific direction about how to describe the coupling mechanism. The superposition can be expressed as the addition form of degradation rates that satisfies where dP/dt is the system degradation rate, dPi/dt is the degradation rate caused by the ith degradation mechanism,and n is the number of degradation mechanisms. Different from this, the competition indicates that the degradation behavior is determined by the dominant degradation mechanism with the highest degradation rate under a specific stress condition, so it is can described as

The combination is suitable for the situation of simultaneous existences of various mechanisms.Each mechanism causes certain damage to a unit, and the accumulated damage between different mechanisms can cause a unit failure. The unit’s time-to-failure (TF) equals to

where Direpresents the damage caused by the ith degradation mechanism. The induction refers to the induction of a certain type of mechanism to another type of mechanism. The reason is generally that the mechanism causing the induction changes the internal or external factors of the induced mechanism,thereby accelerating the degradation behavior of the induced mechanism. Therefore, it can be expressed as

where dPj/dt is the degradation rate caused by the jth degradation mechanism. α（·） is the inducing factor. Mi（I，E） and Mj（I，E） separately reflect the ith and jth degradation mechanisms related to the internal factor vector (I) and the external factor vector (E).

It can be seen that the wear mechanism caused by a sliding behavior can be expressed as fatigue wear, adhesive wear,abrasive wear, and corrosion wear. Among them, adhesion and abrasive grains are the main factors.77,78These above two kinds of wear mechanisms may occur at the same time,but the proportion is different in different wear periods.Among them, the adhesion wear involves many factors, not only with material properties, but also closely related with environmental factors; for the abrasive cutting process, the existing research is limited to the simulation calculation of some simple geometric shapes of hard particles,so it is difficult to give a relatively complex calculation of grinding wear modeling.The impact wear is the collision in a short time.The wear mechanisms caused by the impact can be summarized as the following four categories79: brittle fracture and crack (surface fatigue debris); tough squeeze pressure; plastic deformation;oxidized abrasive wear. The coupling damage caused by both sliding and impact motions has been investigated by many researchers. The mechanisms occurring in a given impact case are determined by the stress and sliding conditions within the contact.Potential coupling wear mechanisms are illustrated in Fig. 16.

Fig. 16 Classified wear process with different motions and mechanisms.

By the existing research,surface fatigue leads to the formation of spalls.Repetitive impact of metallic components at low stress levels, however, is more likely to result in small, submicron wear particles and a highly oxidized surface state. Under moderate conditions of normal or compound impact, it has been shown that wear is caused by a process similar to that of adhesive wear during sliding. This indicates that the adhesive wear mechanism can dominate the whole wear process in some cases. More severe conditions at the contact lead to the occurrences of surface fatigue, subsurface crack growth,and spalling. Very high impact energies bring about failures due to surface fracture and produce rapid material removal.This evolution of wear mechanisms, as the conditions in an impact contact become more severe, is illustrated in Fig. 17.Oxidative wear tends to occur at low stress levels and can be found in both normal and compound impact contacts.Besides,it is particularly prevalent in contacts with a small amount of sliding.Wear rates are low as contacting surfaces are separated by oxide films,which tend to have better lubricating properties than those of bare metal surfaces.Wear debris is generally fine and predominantly made up of metal oxide. The hardness of oxidative wear debris is basically higher than those of contact friction pairs, causing abrasive wear happen on the contact surface. The following content will introduce the existing research on the coupling wear mechanism based on different experiments.

Fig. 17 Coupling sliding/impact wear mechanisms.

Based on the above analysis,the damage caused by the coupling sliding/impact behavior should focus on how to characterize the mathematic relationship between the abrasive grains and the adhesion caused by the sliding wear and the fatigue or abrasive grains caused by the impact load. The potential coupling mechanism summarized as superposition, competition, combination, and induction can be a reasonable choice for describing the coupling relationship between the sliding and impact wear behaviors.

4.3. Wear equations for prediction

Based on the above-mentioned interface contact theory and the continuous refinement and deepening of wear mechanisms,wear calculation and prediction modeling have become a new prospect.At this time,it is necessary to establish a quantitative wear equation for the sliding/impact wear behavior.Meng and Ludema80divided wear equations (predicting the wear rate or wear) into three categories through a lot of literatures: empirical equations, equations based on contact mechanics, and equations based on material failure mechanisms. The establishment of an entire wear equation depends on how this quantitative model is used to describe the wear process and which parameters should be included in the model.

An empirical equation is determined by data fitting,but few documents prove that these variables are independent of each other. Typical empirical equations are only valid in the test range and more accurate than theoretical equations. Most of the theoretical equations solve a wear problem under fixed sliding conditions, usually without considering the temperature,surface roughness, and other factors.

With the development of the interface contact theory,wear equations based on contact mechanics have been proposed.Different scholars have proposed their own models (including correction models) and computational methods to explain the wear mechanism at the theoretical level.In the last several decades, there have been more than 300 related formulas being proposed,77but even the best formula has a lot of limitations.In addition, many contact calculation formulas or finite element calculations simply consider the friction coefficient as a constant, and do not take into account the effects of wear,which greatly restricts their application areas.

An equation based on the material failure mechanism is mainly due to the fact that researchers have realized that abrasion resistance is not an inherent property of a material.Therefore, they focus on a combination of involving more material parameters like material flow, fracture toughness, fracture strain,and so on.While these models are very complex and with no general form,one study is just specified in one scenario.

The wear patterns or models discussed above are subject to some limitations with different degrees, although there is a lot of literature that has a substantial discussion of wear mechanisms, but only at the qualitative level, resulting in a defective model still being used. Meng and Lumeda80argued that the deficiencies of existing formulas are mainly reflected in the fact that the variables in a formula are not independent of each other and lack the influences of the interdependence, time dependence, and material properties that exist between them.

Based on the establishment of a wear equation,the feasibility of the prediction and calculation of the sliding/impact wear degree of the contact interface of a system is ensured. It can provide a theoretical basis for predicting the long life index of the system. Currently, most of the traditional wear models can only be developed for a single wear mechanism, just like the Archard equation, the three-body abrasive model, and so on. However, from existing experimental studies for the sliding/impact wear, we can see that the sliding/impact wear process is mostly determined by several wear mechanisms (i.e.,fatigue, abrasive, adhesion, etc.), and these involvements and interactions produce a complex wear evolution process which cannot be covered by traditional wear models. The variations of the sliding/impact wear degree under specific conditions are totally different from the prediction by traditional wear equations. For instance, in the actual sliding/impact wear, an increase of the impact force can decrease the wear degree;sometimes, lubrication can increase the wear volume. These abnormal phenomena cannot be explained by the classic Archard equation. In our opinions, traditional empirical methods are not applicable for sliding/impact wear modeling. It urges us to develop a new sliding/impact wear equation. The difficulty for the establishment mainly includes two terms. Firstly,a clear coupling relationship between two or more wear mechanisms should be explained by the micro contact theory.However, different mechanisms have different features, such as the subsurface micro crack propagation of the fatigue wear, the surface asperities’ micro cutting of the adhesive wear, etc., so it is hard to find a unified parameter that can be defined as the coupling feature to satisfy different characteristic requirements of these coexisting wear mechanisms. Secondly, an effective description of the macroscopic statistical wear characteristics should be developed,but with the improvement of prediction accuracy,traditional statistical hypothesis based on the Gauss distribution is no longer applicable due to the limitation of the scale dependence. Furthermore, the dynamic variation of surface morphology that can affect the wear process presents a greater challenge to statistical modeling. Therefore, it is difficult to develop a new macroscopic wear model to reflect the comprehensive characteristics of those micro wear processes.

4.4. Wear life prediction

Life prediction for a system involving a sliding/impact wear process is very important for guaranteeing operational safety.81,82With regard to characterization of the sliding/impact wear life, it is necessary to consider the change of the system mechanical behavior caused by the interface damage of the system, which can lead to a failure of the system function performance. Therefore, the determination of the wear threshold is critical and worthy of study. At present, most of the wear life characterization methods can be divided into two categories: one directly from a measurement of the wear volume; the other is indirect through external output signals to identify the system change caused by the wear or performance degradation state.

The wear amount can be directly reflected in the wear degree,83and current wear representations of mechanical parts are concluded into three forms,that is,the grinded material volume,84quality,and wear surface thickness.This can only reflect the specific wear volume, lacking the distribution, depth, etc.Some researchers have studied the shape features of wear particles generated from wear track that usually contain plenty of information about the wear states of a machinery operational condition.85,86However,they cannot reflect the system function or performance parameters influenced by the wear.That is,the system’s failure criterion cannot be directly linked with the degree of wear.Some literature just directly gives the maximum wear threshold without an analysis process. For example,Kawakubo et al.87evaluated the wear life of the contact head of a hard disk based on the wear thickness. The quantitative relationship was not discussed between the wear volume and the hardware’s functional failure criterion.Lu et al.88and Pan et al.89evaluated the wear life based on the maximum wear amount in predicting the gear duration time,although the maximum wear was mentioned as the wear threshold, but not the mechanical response of the system or system failure criteria.

Besides, some scholars have tried to qualitatively explain the wear effect from the point of view of system function or performance degradation.90-92For instance, Yin et al.92utilized a simulation method to describe the surface spall evolution process of gear systems and proposed a lifetime prediction model based on degradation performance parameters.However, most of them lack a theoretical quantitative analysis process, combining only simulation or experimentation to give some qualitative explanations by fitting test data. In addition, there is few corresponding research concerned about the uncertainty of environmental conditions. Uncertainty of environmental conditions can lead to fluctuations in the wear process,93-95leading to the system function or performance showing uncertainties at the same time. Therefore, there is a lack of a basic theoretical approach to describe the relationship in the whole transmission chain between the interface wear and the system failure behavior,which leads to that the product’s life design based on the damage mechanism cannot be improved. Thus, a corresponding life characterization study can predict the wear life under specified working conditions, further providing a basis for relevant testing activities, maintenance, and protection strategy. From our perspective, a practical coupling sliding/impact wear life should be defined as the moment when the system loses its partial function or performance and cannot operatenormally due to the coupling sliding/impact wear. The corresponding wear life prediction process is presented in Fig. 18.Firstly, based on the system dynamic analysis, an interface contact calculation was performed by a contact model. Then,combing with the sliding/impact wear model, the wear degree can be determined. Besides, the relationship between the wear degree and function/structure parameters can be investigated to modify the function/structure model, and the system performance response P should be studied, which reflects the system performance change caused by the sliding/impact wear. Finally, the coupling sliding/impact wear life t*can be calculated given that the system performance threshold equals to P0(if P ≥P0, the system fails). Furthermore, uncertainties existing in these models (such as I,E, P0,etc.) need to be considered in the actual sliding/impact wear life prediction.

5. Conclusions

Current studies on the sliding/impact coupling behavior basically focus on the observation of wear phenomena, not giving a reasonable quantitative description. Therefore, in order to describe the damage mechanism and life prediction about the sliding/impact coupling behavior, this paper discussed the interface contact, coupling damage mechanism, quantitative wear prediction, and life characterization. It can clarify the development trend of future research on the sliding/impact wear behavior, pointing out the improvement direction of existing theories or methodologies. The significance can be provided as follows:

(1) Clearly understanding wear failure in a contact interface. The coupling mechanism between wear mechanisms is effectively interpreted by describing system characteristics such as the historical correlation and changes of the system characteristics.

(2) Supporting the implementation of high-reliability design requirements for systems. Based on the recognition of the wear process, a deep description of the wear behavior and wear mechanism is realized.From the viewpoint of high-reliability design, it can help find the potential weakness of a structural system, optimize the design parameters, and provide a basis of related materials or structures in the process,lubrication medium,and material selection.

(3) Providing a theoretical basis for predicting the long life index of systems. Based on the establishment of a wear equation,the feasibility of the prediction and calculation of the wear degree of the contact interface of a structural system is ensured. A corresponding life characterization study can predict the wear life under specified working conditions,and thus further provide a basis for relevant testing activities, maintenance, and protection strategy.

Acknowledgement

This study was supported by the National Natural Science Foundation of China (No. 51675025).