Elsevier

Wear

Volume 268, Issues 9–10, 25 March 2010, Pages 1157-1169
Wear

Influence of the friction coefficient on the contact geometry during scratch onto amorphous polymers

https://doi.org/10.1016/j.wear.2010.01.003Get rights and content

Abstract

The scratch behaviour of an amorphous polymer was investigated to determine how the characteristics of the material affect the scratch resistance. A thermoplastic polymer (polymethylmethacrylate) was studied using both an experimental device allowing in situ observation of the contact area during scratching with spherical tips and a numerical approach using finite element modelling (FEM). The rheological properties of the PMMA surface were approximated by a simplified bilinear law, while the friction at the interface between the indenter and the material was modelled with Coulomb's local friction coefficient μad varying between 0 and 1, for each computed ratio a/R from 0.1 to 0.6 (where a is the contact radius and R the radius of the tip). FEM results for elastic–plastic contacts indicate that the contact geometry is directly related to the plastic strain field in the deformed volume beneath the indenter during scratching. The dimensions of the plastically deformed volume and the gradient of plastic strain both are shown to depend on the ratio a/R and also on the local friction coefficient μad. An equivalent average plastic strain is calculated by FEM over a representative plastically deformed volume. The average plastic strain increases with a/R, as predicted by Tabor's empirical rule, and with the local friction coefficient for a given ratio a/R. Clear correlations are demonstrated between the average plastic strain and the geometrical parameters classically used to describe the shape of the contact area.

Introduction

For many industrial applications, materials are more and more coated to modify their surface properties, and thick polymers films or varnishes (paintings for automotive industry or optical devices) are often used to allow protection against environmental influences, due to their chemical or appearance properties [1], [2]. The assessment of wear and damage on the protected surface appears to be very important. When two rough engineering surfaces are in contact, local asperities at the microscopic level will experience both normal and tangential loads, whereupon the softer material (the polymer films, in many cases) will show as a function of the mechanical properties of both materials and of different contact parameters (asperity geometry, loading rate, sliding velocity) either elastic deformation, plastic deformation and severe damage, such as cracks, as described in Fig. 1. Any form of scratch tests is a form of controlled abrasive wear and thus it seems reasonable to use such academic tests, first as a simple means of ranking material for their likely resistance to abrasion in service, but also as a reproductive means of assessing the mechanical properties of surfaces [3], [4]. These experiments allow a normal and tangential loads to be applied on a controlled way, and then to evaluate the mechanical response of given surfaces as a function of different experimental parameters, in order to reproduce the nature of the contact in service [5]. However, scratch test is considered as complex system testing, because surfaces of bulk or coated materials have to withstand a complex stress state generated by the sliding indenter under normal force. As described analytically for elastic contact [6], [7], scratch testing leads to (i) compressive stresses in front of the indenter and (ii) tensile stresses behind the indenter, and (iii) shearing strain caused by friction between the indenter and the tested surface [5]. Due to the inhomogeneous state of strain and strain rate, a good interpretation of scratch tests requires at present time numerical simulations. During the last decades, as compared to static normal indentation, few studies using finite element approach [8], [9], [10], [11], [12] have systematically investigated the mechanics for scratching with conical indenter geometry and often for frictionless contact, in the case of elastic–plastic materials. Recently, similar studies coupling experimental results and finite element analysis have described the mechanical behaviour of amorphous polymeric surfaces, especially the transition between quasi elastic and ductile ploughing regimes during scratch with spherical indenter, accounted for the influence of the mean contact strain a/R and the local friction coefficient μad [13], [14], [15]. The main goal of these previous studies consists in the prediction of a strong connection between the frictional sliding process and the material elastic–plastic properties [11], [12], [13], [14], especially the ability to strain harden.

However, a key point for a better understanding of mechanics during scratch process deals with the evaluation of the elastic strain component, the definition of a representative plastic strain and then a total scratch strain [16], that could be connected to the average contact pressure, as for all reverse methods proposed for interpretation of load–displacement curves obtained during indentation tests [17]. As demonstrated by Gauthier et al. [18] using specific microscratch device, the level of deformation during scratch with spheres of polymeric surfaces is governed in first approximation by the relation proposed by Tabor [19] for indentation of metals:ε¯0.2acR2ac2=0.2tanβwhere ac is the true radius of the contact area, R is the radius of the scratching tip, and β is the attack angle. Due to the large elastic recovery of polymers, by comparison with metals and high local friction coefficient greater than 0.15–0.2 classically measured for metals, the coefficient 0.2 in the previous equation does not hold for polymers [20]. Hence, for spherical indentation and for given local friction coefficient, the level of deformation imposed during scratch remains proportional to the ratio a/R, named the mean contact strain [13]. Moreover, as for purely elastic contact [6], [7], the local friction coefficient modifies clearly the elastic and plastic strain fields in shape, size and intensity [14], [15]. Even if it is difficult to predict the influence of the local friction without any calculations, it is not obvious to think that the total scratch strain may be dependent of the friction. According to pioneering work of Challen and Oxley [21], about scratching with wedge (2D case) of rigid perfectly plastic materials, an increase of the friction produces a very marked decrease in the thickness of the strained layer and finally a simultaneous increase of the average plastic strain in this layer.

In this paper, using a three-dimensional finite element modelling (FEM), the influence of the local friction during scratch tests with spherical indenter is studied. The material is modelled as an elastic linear hardening plastic material, whose rheological parameters correspond to mechanical properties of an amorphous thermoplastic polymer (polymethylmethacrylate). The main objective is to define an average plastic strain during scratching process as a function of the mean contact strain a/R and the local friction coefficient μad. As shown in Fig. 2, several geometrical parameters used to describe the contact geometry between the moving tip and the deformed surface, especially the pile-up formation and the elastic recovery at the rear part of the contact, are studied as a function of the scratching conditions (a/R, μad). These results obtained with FEM are compared to experimental results obtained with the built-in microscratch device developed in the laboratory, that allows in situ observations of the contact and the residual groove left on the surface during test [22].

Section snippets

Microvisioscratch setup

The experimental device for the scratch test, called the ‘microvisioscratch’, has been described previously [18], [22]. It consists of a commercial servomechanism bearing a small, temperature controlled transparent box which contains the sample and the scratching tip. Control of the moving tip and recording of the normal FN and tangential FT loads, scratching speed Vtip and temperature T are computer driven. A built-in microscope allows in situ observation and measurement of the groove left on

Experimental in situ observations

In scratch tests using a spherical indenter, for polymeric surfaces the geometry of the contact between the moving tip and the deformed surface is much more complex to predict than for other classes of materials like for example ceramics, which mostly display elastic contacts, or metals which rapidly exhibit fully plastic contacts. In the case of polymers, the shape of the contact area changes with the temperature, tip speed and applied normal load for a given tip radius [13], [18]. As

Elastic and plastic strain fields

During scratch tests, elastic and plastic energy is consumed mainly in two distinct zones: (i) a very thin layer at the interface between the rigid indenter and the surface, which is subjected to extremely high shear strain due to the high strain rate imposed by the relative movement and adhesive slipping arising from local friction, and (ii) a larger and deeper volume, where elastic–plastic yielding occurs. By analogy with indentation, the second volume is assumed to be roughly spherical and

Discussion

The above figures indicate that the local friction coefficient plays a similar role to the geometrical ratio a/R in the evolution of the plastic strain during a scratch test with a rigid spherical indenter. The FE results reveal that with increasing local friction there is

  • (i)

    an increase in the level of plastic strain imposed beneath the moving tip and in the residual groove;

  • (ii)

    an important modification of the shape and size of the plastically deformed volume and hence of the corresponding plastic

Conclusion

A detailed study of the effect of the local friction coefficient on scratch testing was carried out. The results of numerical simulations reveal the influence of the local friction coefficient on the plastic deformation in scratch experiments with a spherical indenter. The local friction between the moving spherical indenter and an elastic–plastic surface not only increases the maximal value of the equivalent plastic strain, but also modifies the shape of the plastically deformed volume and the

References (27)

  • C. Gauthier et al.

    Elastic recovery of a scratch in a polymeric surface: experiments and analysis

    Tribology International

    (2001)
  • S. Lafaye et al.

    A surface flow line model of a scratching tip: apparent and true local friction coefficients

    Tribology International

    (2005)
  • K.L. Johnson

    The correlation of indentation experiments

    Journal of the Mechanics and Physics of Solids

    (1970)
  • Cited by (35)

    • Scratch behavior of polycarbonate by Rockwell C diamond indenter under progressive loading

      2020, Polymer Testing
      Citation Excerpt :

      Scratch-induced surface damage can change morphology/aesthetics and functionality of polymeric surface. Scratch experiments [5–7] together with finite element simulation [8–10] and theoretical modeling [11–16] have been widely adopted to study various scratch-induced damage patterns [17] (e.g. adhesive failure of multi-layer polymeric systems [18]) as well as tribological properties of different polymers (e.g. polypropylene/PP [19,20], ductile polycarbonate/PC [21], thermoplastic and brittle polymer polymethylmethacrylate/PMMA [22] and thermoset resin CR39 [23]), which are complicated and little understood [24] due to the complex contact condition together with pronounced material pile-up, instantaneous material recovery in scratch of viscoelastic polymer of inherent rate-dependent nature [25], and a large number of factors influencing scratch-induced damage [26] such as indenter size [27], indenter geometry (e.g. Berkovich [28], spherical [29], conical [30] and sphero-conical indenters [31,32]), residual stress on the surface [33], interfacial friction, thermal history [21], local surface/sub-surface flaws [34,35], surface texture [21,36], stiffness of scratching instrument [37], scratch speed [9,35,37–40], strain rate [41], normal load [42,43], media [44] and material properties (e.g. constitutive model [9], plastic friction/slippage [10], crystallinity [45], elastic modulus [46], strain hardening [47], yield strength [8,10,48] and viscoelastic characteristics [49–52]). Stress and strain fields cannot be experimentally evaluated [10].

    View all citing articles on Scopus
    View full text