Proposal of invariant precursors for boundary lubricated scuffing

Proposal of invariant precursors for boundary lubricated scuffing

Wear 340-341 (2015) 53–62 Contents lists available at ScienceDirect Wear journal homepage: www.elsevier.com/locate/wear Proposal of invariant precu...

8MB Sizes 0 Downloads 10 Views

Wear 340-341 (2015) 53–62

Contents lists available at ScienceDirect

Wear journal homepage: www.elsevier.com/locate/wear

Proposal of invariant precursors for boundary lubricated scuffing Łukasz Wojciechowski a,n, Thomas G. Mathia b a b

Institute of Machines and Motor Vehicles (IMRiPS), Poznan University of Technology, Poland Laboratoire de Tribologie et Dynamique des Systèmes (LTDS)–C.N.R.S., École Centrale de Lyon, France

art ic l e i nf o Article history: Received 30 September 2014 Received in revised form 1 May 2015 Accepted 13 May 2015 Available online 21 May 2015

a b s t r a c t One of the fundamental problems in tribology is the difficulty in predicting final performance and, in particular, the longevity of lubricated or not frictional components. In most cases of the hydrodynamic or elastohydrodynamic lubricated parts, if the load is too high or the motion of the rubbing elements is too slow, boundary lubrication occurs and predicting the performance of the bodies may become complicated due to the scuffing that can appear. The authors of this paper aimed to shear their knowledge and contribute to better understanding of this transitional phase dealing frequently with catastrophic wear. An analysis of various factors (residual stresses, surface free energy and surface morphology) that contribute to and allow for the identification of conditions that lead to the starting phase of the scuffing process will be presented in this paper. The double-blind strategy of experiments was implemented to achieve a higher standard of scientific rigour thus leading to a new concept of the scuffing invariant precursors. & 2015 Elsevier B.V. All rights reserved.

1. Introduction Scuffing is undoubtedly one of the most dangerous potential effects of tribosystems and generally the nature of this tribological wear is still not well understood. According to the ASTM G40 standard, scuffing can be defined as a form of wear that occurs in inadequately lubricated tribosystems which are characterised by macroscopically observable changes in texture with features related to the direction of relative sliding. The mechanism of scuffing has been the object of tribological investigations for many years but the factors that cause its activation are still weakly recognised among scientists. Apart from this fact, it has generally been accepted that the constituent which is necessary but most often insufficient to lead to scuffing, which may occur in lubricated systems, is the breakdown of the elastohydrodynamic or boundary film. Boundary lubrication (BL) can be defined as the regime in which the average lubricant film thickness is so small that the surface asperities come into contact with one another under relative motion, or the regime in which the load is mostly carried by the surface asperities rather than by the lubricant [1]. Generally, if running under conditions of high load and slow speed the tribosystem becomes particularly conducive to this regime. The contact pressure may grow beyond the typical level of the EHL then and may lead to elasto-plastic deformations of the asperities. n

Corresponding author. E-mail address: [email protected] (Ł. Wojciechowski).

http://dx.doi.org/10.1016/j.wear.2015.05.007 0043-1648/& 2015 Elsevier B.V. All rights reserved.

A part of the surfaces may be coated by residual, physically and/or chemically adsorbed full or partial lubricating monomolecular layers. In this case the contact between the asperities is close to that in dry friction and the lubrication effect is obtained by the low shear strength of these layers in the interfacial area. Several theories [2] have been suggested in the literature, yet none of them unequivocally determine the evolution of the abovementioned process. The most relevant theories connect the initiation of this process under the boundary lubrication (BL) process with

 critical temperature initiating instable wear [3,4];  debris generation size [5], kinetics of their accumulation in the      

interface [6] and morphologically trapping space; plastic deformation of asperities [7]; critical friction power [8]; formation and destruction of protective oxide films [9]; desorption of polar constituents of lubricant from metallic surfaces [10–12,23]; lubricant structure decomposition [13,14]; and energetic activation of metallic surfaces [2,15,16] or adiabatic shear instability [17].

Regardless of the above, BL cannot provide long-term protection against scuffing. For this reason it is fundamental to recognise the breaking point of the boundary layer activating the scuffing wear. The authors present a topological paradigmatic approach to

Ł. Wojciechowski, T.G. Mathia / Wear 340-341 (2015) 53–62

54

Nomenclature

γS (SFE) surface free energy [mJ/m2] γSLW Lifsitz–van der Waals component of surface free γS þ /ΘD ΘF ΘW BL EHL Sa

energy [mJ/m2] acid–base component of surface free energy [mJ/m2] diiodomethane contact angle [deg] formamide contact angle [deg] Water contact angle [deg] boundary lubrication elastohydrodynamic lubrication arithmetic mean height, mean surface roughness [μm]

Sku Sp Sq Ssk Sv Sz tSC Vvv

kurtosis of the surface height distribution maximum peak height, height between the highest peak and mean line [μm] root mean square height, standard deviation of height distribution of surface roughness [μm] skewness of the surface height distribution maximum pit height, depth between mean line and the deepest valley [μm] maximum height between the highest peak and the deepest valley [μm] time to scuffing [s] void volume of the valleys [ml/m2]

Fig. 1. Kinematics and geometry of the scuffing investigations (a), procedure of load application (b) and coefficient of friction versus time and load (c).

this problem in which the key role is played by the interaction between the rheological, morphological and physicochemical properties of contacting the surface's layer [2,18,20]. Some specific case is the scuffing under dry friction conditions, limited in practise to only some configurations of materials (very often friction pairs with copper alloy components). The detailed analysis concerning the predisposition of different metals to

scuffing initiated under dry friction conditions was carried out by Kovalchenko et al. [21]. For academic reasons, specific tribological conditions were selected in the paper. The results of the scuffing tests for different burnishing and finishing process conditions are discussed, offering various ground surface properties in order to modify the tribological wear performance of the AISI 4130 cylinders-plane of grey

Ł. Wojciechowski, T.G. Mathia / Wear 340-341 (2015) 53–62 Table 1 Surface free energy and its components depending on the pressure of the burnishing tools on the worked surfaces. Pressure of burnishing tools on machining surfaces [GPa]

Contact anglea [deg] Surface free energyb [mJ/m2]

1.30

ΘW ¼81.5 71.5 ΘD ¼ 56.7 7 5.5 ΘF ¼ 80.9 7 2.4 ΘW ¼76.6 72.2 ΘD ¼ 29.9 7 1.6 ΘF ¼ 68.47 5.8 ΘW ¼63.6 7 4.9 ΘD ¼ 35.17 2.6 ΘF ¼ 65.9 7 6.1 ΘW ¼52.3 7 16.5 ΘD ¼ 34.4 7 1.1 ΘF ¼ 62.7 7 6.9 ΘW ¼52.9 7 16.9 ΘD ¼ 29.2 7 0.5 ΘF ¼ 59.6 7 2.9 ΘW ¼68.3 74.3 ΘD ¼ 30.1 70.5 ΘF ¼ 65.2 7 11.1

1.64

1.87

2.06

2.22

2.36

γS

γSLW

γS

41.1

30.5

10.6

55.0

44.3

10.3

57.5

42.0

15.5

62.6

42.3

20.3

62.1

44.5

17.5

56.9

44.2

12.8

+



a Mean values of contact angles for applied test fluids: water (ΘW), diiodomethane (ΘD) and formamide (ΘF). b Designation of surface free energy and its components: γS – total value, γSLW – + Lifsitz/van der Waals component, γS − – acid/base component, γS+ – acid component and γS− – base component.

Fig. 2. Residual stresses versus pressure of burnishing tools on machining surfaces.

cast iron counter-bodies boundary lubricated with gear oil (with extreme pressure additives). Taking into consideration the topological approach, mutual interactions between the analysed parameters in the context of scuffing were investigated and elucidated in order to propose an invariant precursor concept.

2. Methodology The French Academy of Sciences in 1784 introduced the first so called ‘blind experiments’. More recently, this technique became more popular, however its application in tribology is still limited. Blind testing eliminates unintentional influence or any other bias of person performing experiments or analysis, as they are not aware of expected results. In the presented research here a double blind experimental strategy was implemented. It consists of rigorous way of conducting experiments which eliminate subjective, unrecognised or unsuspected biases carried by scientist and/or technician. In the case of present study the research was carried out separately and then combined in two different locations: France and Poland.

55

The metallic specimens investigated in the tribological experiments were manufactured from AISI 4130 steel in the shape of cylinders with 45 mm external diameter and a width of 12 mm. Cylinders of all specimens were subject to grinding so that their surface morphologies were anisotropic, as profile roughness was relevant to the process. The measured Ra was equal to approximately 0.5 μm. The specimens were then divided into six batches, all of which were subjected to burnishing with six values of pressure (thus creating different levels of energy introduced to the surface layer). Burnishing of the cylinder was performed by two symmetrically spherical sector-shaped rolls of 50 mm in diameter. The particular levels of burnishing pressures complied with the following pressures of burnishing tools on machined surfaces: 1st: 1.3 GPa, 2nd: 1.64 GPa, 3rd: 1.87 GPa, 4th: 2.06 GPa, 5th: 2.22 GPa, and 6th: 2.36 GPa. Kinematic conditions for burnishing were speed – 100 m/min, burnishing feed – 0.08 mm/rev, and number of passes – 2; lubricated by a 1:1 mixture of mineral oil and kerosene. The surface free energy (SFE) of the rubbing bodies was established on the basis of measurements of the static contact angle on the surfaces of the specimens. Three test liquids (water, diiodomethane and formamide), typical of the acid–base method [19], were applied for this purpose. The detailed procedure for determining SFE has been described elsewhere [2]. The residual stresses were measured by the X-ray diffraction method. In order to satisfy the statistical requirements, measurements were conducted in three points every 120° on the cylindrical surfaces of the specimens. Morphological analysis was performed with an optical interferometer on the relevant milimetric area of 1.2  0.9 mm2 in five points every 72° on the cylindrical surfaces of the specimens. The criterion for determining scuffing was an increase of the friction coefficient under constant load (Fig. 1). In order to satisfy the statistical requirements, the scuffing investigation was performed four times for each pair of cylinders and flat blocks. The experiments were performed under single drop lubrication using gear oil with ca. 5% olefin sulphide as an extreme pressure additive. Therefore, the scuffing activation period was measured by the time it took for scuffing to occur, which was equivalent to the case of poor lubrication in the friction node. Scuffing kinetics was performed at sliding speed between the rotating cylinder (AISI 4130) and the stationary block (EN-GJL-300 cast iron with flake graphite) at 0.5 m/s. The load applied incrementally to the friction pair is also presented in Fig. 1.

3. Results and discussion Table 1 presents the physicochemical properties of the AISI 4130 cylinders: SFE together with its components depending on the pressures of the burnishing tools on the machining surfaces of the specimens (including mean values and standard deviation for the contact angle for all testing fluids). An analysis of the presented data points to a significant increase in the value of SFE as a function of the increasing contact pressure of the burnishing tools on the machining elements. This increase was incremental at a level of 21.1 mJ/m2 (from 41.1 mJ/m2 at a pressure of 1.3 GPa to 62.6 mJ/m2 at a pressure of 2.06 GPa). It is worth noting that the optimal value of SFE was observed at “Hertz” pressure slightly lower than its maximum value and equal to 2.06 GPa. At maximum burnishing pressure, i.e. 2.36 GPa, a noticeable decrease (5.7 mJ/m2) in the value of SFE was systematically observed. This fact is probably caused by the socalled critical cold work – when the level of crystallites packing the material is so high that further plastic deformation causes no further strengthening of the surface layer, and its destruction begins. More precise information can be found through an analysis

56

Ł. Wojciechowski, T.G. Mathia / Wear 340-341 (2015) 53–62

Fig. 3. Statistic characteristics, 3D morphological views and Abbott-Firestone curves of AISI4130 cylinders' surfaces burnished under 1st (a), 2nd (b) 3rd (c), 4th (d), 5th (e), and 6th (f) level of burnishing conditions.

Ł. Wojciechowski, T.G. Mathia / Wear 340-341 (2015) 53–62

57

Fig. 4. Vvv (void volume of the valleys) of AISI 4130 cylinders burnished under 1st (a), 2nd (b), 3rd (c), 4th (d), 5th (e) and 6th (f) burnishing conditions.

Fig. 5. Vvv (void volume of the valleys) versus pressure of burnishing tools on machining surfaces.

of separate components of SFE, which allows us to notice an initial increase in the Lifsitz/van der Waals component value (between the 1st and 2nd level of burnishing) and then its stabilisation at the up-close level. This probably comes from the fact that plastic deformations in their first phase strongly affect the orientation of the crystal structure elements. A further increase in the pressure probably has no notable influence on the grains' orientation, but more so on their packaging density and mechanical energy accumulation. Additionally, a rise in the acid/base component value in the range of 10 mJ/m2 can be observed (for "Hertz" pressures of 1.64 to 2.06 MPa). This has an essential importance on the interaction of the surfaces with the lubricating mediums due to extreme pressure additives. These kinds of additives often have a polar character, and that is why the growth of the acid/base component of SFE, (which is responsible for polar interactions with lubricants in reality) can be a perfect starting point for forming more resistant boundary layers. The

Fig. 6. Physicochemical properties of the surface of AISI 4130 cylinders (acid/base component of SFE and residual stresses) versus time to scuffing activation.

direct relationship between surface polarity and resistance to scuffing was found in [2]. The average values of the measured residual stresses for all types of cylinders are summarised in Fig. 2. The increase of work hardening of the surface layer results from the increase of the compressive residual stresses, and can be observed with the growing burnishing pressure level. An exception to this rule is the case of specimens burnished by maximum pressure, where a small decrease of compressive stresses is noticeable. The reason for this observed phenomenon is probable initiation of the surface destruction process caused by the achievement of the critical cold work. However, this effect is not as evident as in the relationship between SFE and the pressure burnishing level. The surface morphology characterisations were done very carefully and by taking into consideration calibration as well as the transfer function and limits of the metrology device. Fig. 3 shows

58

Ł. Wojciechowski, T.G. Mathia / Wear 340-341 (2015) 53–62

Fig. 7. Motifs views, their mean number, area and height of AISI 4130 cylinders burnished under 1st (a), 2nd (b), 3rd (c), 4th (d), 5th (e) and 6th (f) burnishing conditions.

morphological changes for all levels of burnishing pressure. The values of characteristic parameters of the surface (Sa, Sq, Sp, Sv and Sz) significantly changed due to burnishing. The decrease in

values for all of the parameters for the first four levels of burnishing pressures (e.g. average Sq reduced from 0.45 μm to 0.35 μm) can be observed. A further increase in pressure caused

Ł. Wojciechowski, T.G. Mathia / Wear 340-341 (2015) 53–62

59

Fig. 8. Mean number of motifs and their area (a) and height (b) versus pressure of burnishing tools on worked surfaces.

the beginning of surface destruction and a rise in these parameters (e.g. Sq¼ 0.49 μm for a pressure of 2.22 GPa and Sq¼ 0.54 μm for a pressure of 2.64 GPa). Additionally, the values of skewness Ssk were negative, which indicates that most of the material was localised near the peaks of the surface. The kurtosis (Sku) value changed, indicating flattening peaks of the surface and increasing bearing area contact due to plastic deformation. The shape of the Abbott-Firestone curves confirms the flattening of the surface asperities due to burnishing. In contrast to the height parameters, the shape of the curves and the depth of the material concentration do not show symptoms of surface damage. This fact is identifiable only after analysis of the parameters characterising the bearing curves. A parameter that was very well suited for this analysis was Vvv, i.e. void volume of the valleys. According to ISO 25178, the standard Vvv can be defined as a dale volume at mr2 material ratio, where mr2 is usually established at a level of 80%. Fig. 4 presents the bearing curves with a visualisation of the Vvv parameter and its average value for all levels of burnishing pressure. Additionally, Fig. 5 shows the relationship between the Vvv parameter and the pressure of burnishing tools on machined surfaces. Generally, the value of the Vvv parameter decreased due to burnishing, which confirms flattening of the surface asperities. Only the sixth level of burnishing pressure caused re-growth of this characteristic. This fact, coupled with changes in height parameters, indicates the start of surface destruction due to overbearingly large cold work (for this reason further analysis was conducted without this batch of cylinders). The analysis of data obtained in the scuffing investigations allows for a certain trend to be observed in the value of time to scuffing as a function of the physicochemical properties of the surface (Fig. 6). A clear relationship can be recognised between the acid/base component of SFE and scuffing activation. The highest scuffing resistance was observed for specimens with the acid/base component of SFE equal to 20.3 mJ/m2, for which the first scuffing symptoms appeared after 983 s. It is worth noting that there was a significant increase in time in comparison to specimens exhibiting the acid/base component of SFE of 10.6 mJ/m2; the time here being approximately 536 s. Curiously, a further increase in the burnishing pressure level produced no improvement in scuffing. Time to scuffing decreased to almost 820 s with a simultaneous decrease of the acid/base component of SFE equal to 17.5 mJ/m2. There is also a clear dependence between time to scuffing activation and residual stresses. It is evident that the increase in compressive residual stresses causes an increase of the scuffing resistance. The growth of compressive stresses results in enhancement of cohesion forces responsible for maintaining the structure of the metallic surfaces during wear in its entirety. The analysis of morphological changes due to scuffing allows us to identify some invariants that appear and stabilise at the upclose level in this process apart from the starting value. Mean

numbers of motifs, their mean area and height can be a proposition in this case. According to the ISO 12085 standard, motifs can be defined as a portion of the primary profile between the highest points of two local peaks of the profile, which are not necessarily adjacent. A calculation of the motifs' characteristics in 3D is possible thanks to segmentation by use of the watershed algorithm coupled with the innovative “Wolf pruning” algorithm that eliminates insignificant motifs by merging them into larger ones [22]. Fig. 7 presents exemplary motifs' views and their mean numbers and values of height and area for all levels of burnishing pressures. A decrease of the motif number and their height due to increasing burnishing pressure (from 0.437 to 0.408 μm) (Fig. 8b) can be observed. This is a consequence of the surface asperities flattening (due to plastic deformation) which is equivalent to motif flattening and growth of their area (from 0.00471 to 0.00667 mm2) (Fig. 8a). Observation of the same surfaces but after the scuffing investigations allows us to notice that their values are at a similar level. Taking into consideration the first five levels of burnishing conditions, it is visible that the mean number of motifs is equal to 537 4, their mean area is equal to 0.025 70.003 mm2 and their mean height is equal to 2.157 0.2 μm. This indicates that all surfaces reached a similar real contact area directly before scuffing activation (Fig. 9). Unit pressures at the areas of asperities achieved a critical level sufficient to boundary film breakdown and scuffing activation. The next morphological invariant identified in the scuffing investigations was analysed in the technological state of the Vvv parameter. Exemplary values of the Vvv of AISI 4130 cylinders after the scuffing investigations are shown in Fig. 10. The values of the Vvv parameter in the technological state are contained in the range of 0.0236 ml/m2 (from 0.0683 to 0.0447 ml/m2). Due to scuffing, the space characterised by the Vvv parameter significantly increased and formed at an up-close level of 0.352 70.023 ml/m2. On the basis of this information it is possible to indicate the volume of debris produced during the scuffing process. Additionally, knowledge concerning the values of the Vvv parameter may be significant in elucidating the scuffing process activation. The value of Vvv indicates the achievement of some critical volume of the valleys for which the lubricating medium operating in the contact area flows down to the formed free spaces in such a quantity that maintaining fluid friction is impossible. In consequence, the reduced supply of lubricant causes a transition to operation in boundary lubrication conditions and then, consequently, to scuffing activation. The void volumes of the valleys may also be used as a type of “dustbin” for debris produced during friction. As long as the volumes of the valleys contain wear products, the flow of oil in the contact area remains unthreatened. However, if the volume of the valleys is not able to accommodate a larger amount of debris, they will accumulate in the contact area. Thus the oil flow may be

60

Ł. Wojciechowski, T.G. Mathia / Wear 340-341 (2015) 53–62

Fig. 9. Exemplary motifs views, their mean number, area and height of AISI 4130 cylinders burnished under 1st (a), 2nd (b), 3rd (c), 4th (d), 5th (e) and 6th (f) burnishing conditions after the scuffing test.

Ł. Wojciechowski, T.G. Mathia / Wear 340-341 (2015) 53–62

61

Fig. 10. Vvv (void volume of the valleys) of AISI 4130 cylinders burnished under 1st (a), 2nd (b), 3rd (c), 4th (d), 5th (e) and 6th (f) burnishing conditions after the scuffing test.

interrupted and the scuffing process will be activated as a consequence.

them. Further investigations regarding this issue should include the following elements:

4. Conclusions

 On the basis of the experimental observations the authors formulated the following conclusions:

 An increase in compressive residual stresses can be observed









together with an increase of the burnishing pressure level. An exception to this rule is the case of cylinders burnished by maximum pressure, where a small decrease in compressive stresses is noticeable. The reason for this is the start of the surface destruction process caused by the achievement of critical cold work. Generally, the increase of the compressive residual stresses causes an increase of scuffing performance. There are some morphology invariants connected with scuffing process activations. Apart from the starting values, they stabilise at an up-close level, which is characteristic of scuffing process initiation. Mean numbers of motifs and their mean height and volume are at the certain level due to scuffing. This may indicate the shaping of the real contact area of particularly favourable characteristics for scuffing activation. The Vvv (void volume of the valleys) parameter is characterised by similar values for all types of scuffed cylinders. It may indicate that a critical volume of oil has been reached and scuffing can be activated. Due to interactions between characteristic surface parameters, it is possible that improving one of them can cause worsening of another one. Therefore, it is useful to characterise the tribological surface properties in the frame of the topological approach and to look for an optimal compromise between

● Verification of the precursory behaviour of the identified morphologic invariant (motifs and Vvv) for dissimilar configuration of a friction pair’s materials and type of lubricant applied. ● Topological analysis of the surface in terms of the synergism and antagonism of its properties and their influence on scuffing process activation.

Acknowledgements The authors thank G. Meille and J.M. Georges (LTDS) for their long-term cooperation, R. Majchrowski and M. Wieczorowski (Poznan University of Technology) for the topography measurements, and F. Blateyron (Digital Surf) for help in use of Mountainss Software.

References [1] S.M. Hsu, R.S. Gates, Boundary lubricating films: formation and lubrication mechanism, Tribol. Int. 38 (2005) 305–312. [2] Ł. Wojciechowski, T.G. Mathia, Conjecture and paradigm on limits of boundary lubrication, Tribol. Int. 82 (2015) 577–585, http://dx.doi.org/10.1016/j. triboint.2014.02.006. [3] H. Blok, The flash temperature concept, Wear 6 (1963) 483–494. [4] S.C. Lee, H.S. Cheng, Experimental validation of critical temperature–pressure theory of scuffing, Tribol. Trans. 38 (1995) 738–742.

62

Ł. Wojciechowski, T.G. Mathia / Wear 340-341 (2015) 53–62

[5] E. Rabinowicz, Friction seizure and galling seizure, Wear 25 (1973) 357–363. [6] J. Enthoven, H.A. Spikes, Infrared and visual study of the mechanisms of scuffing, Tribol. Trans. 39 (1996) 441–447. [7] L. Jiajun, L. Zhigiang, C. Yingian, The study of scuffing and pitting failure of cam–tappet rubbing pair, Wear 140 (1990) 135–147. [8] R.M. Matveevsky, Friction power as a criterion of seizure with sliding lubricated contact, Wear 155 (1992) 1–5. [9] A.W. Batchelor, G.W. Stachowiak, Some kinetic aspects of extreme pressure lubrication, Wear 108 (1986) 185–199. [10] H.A. Spikes, A. Cameron, Scuffing as a desorption process – an explanation of the Borshoff effect, ASLE Trans. 17 (1974) 92–96. [11] M.W. Bailey, A. Cameron, The effects of temperature and metal pairs on scuffing, ASLE Trans. 16 (1973) 121–131. [12] T.G. Mathia, Étude d'une interface de glissement en lubrification monomoléculaire (Ph.D. thesis), Université Claude Bernard Lyon 1 & Ecole Centrale de Lyon, Lyon-France, 1978. [13] R.S. Gates, K.L. Jewett, S.M. Hsu, A study on the nature of boundary lubricating film: analytical method development, Tribol. Trans. 32 (1989) 423–430. [14] S.M. Hsu, M.C. Shen, E.E. Klaus, H.S. Cheng, P.I. Lacey, Mechano-chemical model: reaction temperatures in a concentrated contact, Wear 175 (1994) 209–218. [15] A.P. Semenov, The phenomenon of seizure and its investigation, Wear 4 (1961) 1–9.

[16] H. Czichos, K. Kirschke, Investigations into film failure (transition point) of lubricated concentrated contacts, Wear 22 (1972) 321–336. [17] J. Hershberger, O.O. Ajayi, J. Zhang, H. Yoon, G.R. Fenske, Evidence of scuffing initiation by adiabatic shear instability, Wear 258 (2005) 1471–1478. [18] T.G. Mathia, F. Louis, G. Maeder, D. Mairey, Relationship between surface states, finishing processes and engineering properties, Wear 83 (1982) 241–250. [19] C.J. van Oss, R.J. Good, M.K. Chaudhury, Additive and nonadditive surface tension components and the interpretation of contact angle, Langmuir 4 (1988) 884–891. [20] K.J. Kubiak, M.T.C. Wilson, T.G. Mathia, Ph. Carval, Wettability versus roughness of engineering surfaces, Wear 271 (2011) 523–528. [21] A.M. Kovalchenko, P.J. Blau, J. Qu, S. Danyluk, Scuffing tendencies of different metals against copper under non-lubricated conditions, Wear 271 (2011) 2998–3006. [22] F. Blayteron, The areal features parameters, in: R. Leach (Ed.), Characterization of Areal Surface Texture, Springer-Verlag, Berlin-Heidelberg, 2013, pp. 45–67. [23] T.G. Mathia, J.M. Georges, Wear initiation of boundary lubricated surfaces, Wear 50 (1978) 191–194.