Scratch hardness and deformation maps for polycarbonate and polyethylene.
The widest application of polymers as engineering materials requires a better understanding and control of their surface mechanical properties. This is particularly important when polymeric materials are used to improve contact or tribological performance; that is, where polymers are adopted in optical, coatings, and plastic engineering applications.
This paper presents results of an experimental study of the surface deformation mechanisms of two solid polymers when they undergo scratch deformations. The experiment involves drawing a rigid indenter of a conical shape along planar, smooth surfaces ([less than]5 [[micro]meter] c.l.a.(*)) of the specimens. The parameters, which influence the measured response of the materials, are the imposed strain (cone angle), normal load under which the indenter is moved, the bulk temperature of the samples, and the sliding or scratching velocity. The results obtained are used to construct deformation maps that show the regions of the different modes of surface damage for the various experimental conditions. The scratch geometry and the normal load are used to compute the scratch hardness of the materials; a plastic deformation characteristic. Commercial ultrahigh molecular weight polyethylene and polycarbonate resin samples were primarily utilized in this study as examples of polymers that have different deformation characteristics. In addition, these materials are commonly employed in applications where scratch deformations undermine the longevity of the components manufactured from these polymers.
Two techniques that were adopted to observe the surfaces after scratching were scanning electron microscopy and laser profilometry. Soon after the production of the scratch deformation, scanning electron micrographs of the scratched samples were taken in order to subjectively assess the deformation modes. Subsequently, laser profilometry scans were performed on the specimen surfaces in a direction orthogonal to scratch direction. The measured parameters, the scratch depth and width, were used to evaluate the scratch hardness of the materials. In addition, the geometry of these scratches provided a characterization of the damage processes, such as the material removal and the debris generation mechanisms, as well as the wear and the mechanical dissipation properties of the material.
Studies on the production of scratch deformation maps for polymers have been few. The present authors (1) have produced scratch maps for PMMA, which showed the effects of the normal load and the scratch velocity upon the deformation mechanisms. They also correlated the normal hardness with the scratch hardness values for lubricated and unlubricated contacts. Briscoe et al. (2-4) have shown the dependence of scratch data upon the imposed strain for [Lambda]-irradiated PTFE, PMMA, and PE. The dependence of scratch hardness of PMMA on the contact temperature was also illustrated, and the same authors (5) have used the scratch hardness data for the evaluation of cure temperature for a glass fiber reinforced polyester. Recently, Briscoe et al. (6) have reported on the scratch hardness of a model paste material and its dependence upon the contact conditions. The behavior of this system represents the pure extreme of a model plastic response for a polymeric material. Lim and Ashby have presented wear maps for metals (7). The scratch behavior of highly brittle solids is characterized by extensive cracking modes and is a topic that has been extensively studied; a recent review by Lawn (8) provides a comprehensive account. Lamy (9) utilized a pendular sclerometer to investigate the ductile to brittle transitions in the scratch deformation of a range of materials, including PMMA. These studies illustrate that the major feature that separates polymers from materials, such as metals and ceramics, at ambient temperature, in terms of deformation mechanisms, is the fact that polymers show a wide range of scratch deformation characteristics within a modest range of the variations of strain, strain rate, and temperature.
Scratch hardness data have also been used for the modeling of abrasion phenomena for a number of materials (10), and the scratch test has been widely adopted to obtain information about the wear resistance and the adhesion characteristics of coatings (11-14). In these cases the hardness dependence upon the contact strain was usually not investigated and additional techniques were used (e.g. acoustic emission, etc. (15)) to characterize the surface mechanical properties.
The general principle on which the scratch method is based relies upon the fact that different materials show different resistances to deformation upon scratching and hence, different modes of deformation. The data obtained provides several surface properties such as the scratch hardness, the fracture toughness, the abrasion resistance, and the identification of the various deformation modes that are specific to the material being examined.
In a quasi-static normal indentation the deformation is symmetrically accommodated by the specimen and the indenter is supported uniformly over the contact area. In contrast, in a scratch deformation where a significant amount of plastic deformation occurs, the indenter is only fully supported by the specimen in the front or leading part while the recovered material, behind the indent, partly supports the indenter in the rear half [ILLUSTRATION FOR FIGURE 1 OMITTED]. The amount of recovered material that will support the indenter depends upon the nature of the material. For a perfectly plastic material there is no recovery after deformation and hence, there is no load bearing support in the rear half of the indenter surface. For an elastic or visco-elastic material there is partial support of the indenter in the trailing zone. Therefore, for a general case of viscoelastic plastic response, the scratch hardness, [H.sub.s], may be defined (by analogy with indentation hardness) as follows:
[H.sub.s] = q 4W/[Pi][d.sup.2] (1)
where W is the normal load applied on the indenter, d is the recovered width of the scratch [ILLUSTRATION FOR FIGURE 1 OMITTED], and the parameter q varies according to the type of material tested and how the material supports the indenter (q [congruent] 2 is for rigid plastic materials, q [greater than] 1 for viscoelastic plastic materials). The computed scratch hardness value is primarily a reflection of the plastic yield character of these polymers.
The strain, in quasi-static indentation, is taken according to Johnson's assumption (16) such that, for a conical plastic indent, the effective strain is proportional to the tangent of the slope of the cone; this relationship appears to be valid for ductile metals, but the proportionality constant may be different for viscoelastic materials (17). Hence, effective strain, [[Epsilon].sub.s], is given as
[[Epsilon].sub.s] = 0.2 tan [Theta] (2)
where [Theta] is the angle between the slope of the cone and the sample surface (attack angle); see Fig. 1. Effective strain rate in scratch hardness may be defined as follows (1):
[Mathematical Expression Omitted] (3)
where v is the scratch velocity and d the measured scratch width, The reciprocal of [Mathematical Expression Omitted] corresponds to the time required for the indenter to traverse a plastically deformed contact diameter.
EXPERIMENTAL APPROACH AND MATERIALS
Figure 2 shows a schematic drawing of the scratching apparatus. It consisted of an adjustable lever arm at one side of which was a thin rigid holder into which the indenter was attached. A micro-step motor stage, which was controlled by a computer, provided the movement of the samples under the indenter. For the scratch studies at different material bulk temperatures, a thermally controlled cell was used. The heating equipment consisted of a metallic sample holder, which was heated by electric resistance cartridges and controlled by an electronic device. The temperature was maintained constant within [+ or -] 0.2 [degrees] C. The polymer samples were fixed inside the cell and the experiments were performed in situ. Before performing the experiments, a laser distance measurement device was used to ensure that the sample surfaces were perpendicular to the axis of the indenter ([+ or -] 2 [degrees]).
Rigid tool steel indenters of conical shape were utilized and, in order to vary the strain of the contact, a range of included angles were chosen between 35 [degrees] and 150 [degrees]. The tip radius of the indenters was ca. 5 [[micro]meter]. The thickness of the samples was ca. 5 mm. The contacts were unlubricated.
To study the dependence of the surface mechanical properties and the nature of the scratch deformation modes, upon the sliding speed and the strain rate at the ambient temperature (21 [+ or -] 2 [degrees] C), scratch studies at two different velocities were carried out; 0.0026 mm/s and 2.6 mm/s. Furthermore, the influence of the sample temperature was also investigated for PC at the scratch velocity of 0.0026 mm/s and at 30 [center dot] C. After each scratch experiment, SEM images of the scratches were recorded to identify the mechanisms of deformations. From the scanning electron micrographs subjective assessments of the damage were drawn. Hence, "scratch maps" for PC and PE were constructed.
Scratched sample surfaces were also examined by using a commercial laser stylus profilometer (Rodenstock, Germany). Laser profilometry provided cross sectional profiles of the scratches from which rather more objective assessments of the damage could be drawn.
Commercial ultrahigh molecular weight polyethylene (PE) (Goodfellow, UK) and MFR 80 polycarbonate (PC) (Dow Chemical Co., USA) resins were used. These materials were originally provided in injection molded sheets in a generally good surface condition; surface roughness ca. 5 [[micro]meter] c.l.a. They were washed in an aqueous commercial detergent solution and distilled water prior to use. No attempt was made to polish the sample surfaces prior to use.
Scratch Hardness of Polycarbonate and Polyethylene
Figure 3 shows the experimental data for the scratch hardness of PC and PE as plotted against the cone angle (strain variation; Eq 2). The scratch hardness values were. evaluated using Eq 1 with q = 1; viz. assuming the fully elastic deformation extreme. The arbitrary choice of q as unity is not regarded as a serious limitation, as the computed values will correspond to effective strains and strain rates, which are themselves only approximate estimates. In addition, during the deformation a prow of material is displaced ahead of the indenter [ILLUSTRATION FOR FIGURE 1 OMITTED], which also influences the value of load supporting area. The data are plotted for different scratch velocities and, in the case of PC, for two bulk temperatures (ambient and 30 [degrees] C). For comparison purposes, PMMA data, taken from ref. 1, have also been included in the figure. In Fig. 3, the PC and PMMA systems show an increasing trend for the hardness in the lower cone angle range and a decreasing trend for the higher cone angle range. This is In contrast to the case for PE where the scratch hardness continues to decrease as the cone angle is changed from lower to higher cone angles. The smaller values of scratch hardness for the lower cone angles indicates that for this condition the material removal process takes place by a more energy efficient method and hence the energy required for the removal process is less; this is the case for the micromachining modes (see later). This process of material removal is called machining, which takes place by plastic micro chip formation accompanied by brittle cracking. Figure 4 shows a micrograph of the PC when scratched by a 45 [degrees] cone. The image shows regular crack formation inside the plastically formed scratch groove. The material deforms in both a ductile and a brittle manner. For the higher angles, there is also a combination of brittle and ductile failure. The data for PC in Fig. 3, show that its scratch hardness increases as scratch velocity is increased. This indicates a strain rate strengthening of the material at the higher scratch velocity. In scratch hardness, the effective strain rate is usually defined as the ratio of the scratch velocity to the scratch width. The strain rate hardening effect means that the material provides a greater resistance to plastic deformation. This is mainly a bulk response as, during scratching, considerable bulk deformation of the material takes place. However, for the higher scratching velocity case, interfacial heating may be contributing to the trends; see for example reference 18 for typical calculations and West and Senior (19), who also reported on an interfacial melting effect and changes in crystallinity during the sliding of steel upon a polyethylene surface. This interfacial heating is confined mainly to the contact surface regions. As the scratching is a dynamic process, there is less time available for the dissipation of heat energy into the bulk of the material at the higher sliding velocities. The interfacial heating gives rise to a predominantly ductile deformation of the material in the contact zone; this has been observed in the micrographs to be presented later in this paper. It may also be the case that interfacial heating produces an effective reduction in the shear stress; the interface shear stress, or friction, decreases with increasing temperature (20). In effect, the contact is lubricated by a softened interfacial layer; this effect has been observed for pastes (6).
The scratch hardness data for PE show only a decreasing trend with the increasing of the cone angle and the decrease in the contact strain. This may reflect the strain dependence of the yield process but this polymer is known to undergo extensive structural reordering and strain softening in interfacial deformations (see for example reference 20). However, it is noted later that for this polymer the main deformation processes involved almost entirely ductile flows. The data shown in Fig. 3 also indicate a time dependence effect, the influence of velocity, which is anticipated on the basis of a strain rate dependent yield stress.
In order to examine the effects of temperature upon the deformation mechanism, a few experiments were conducted at higher bulk temperatures. A specimen of PC was heated in the thermal cell and scratched while the material was hot. The data for these experiments are also shown in Fig. 3. The hardness value decreases considerably for an increase of the bulk temperature by 9 [degrees] C. This is mainly due to the bulk softening of the material. Hence, the scratch hardness increases for the high scratch velocity case despite the interfacial heating effect. This is because the frictional heat does not appreciably dissipate into the bulk of the material, whereas, in the case of external bulk heating, the scratched hardness is lower due to the bulk softening.
For the PE sample, it is observed that the scratch hardness continues to decrease as the cone angle is increased. At the higher scratch velocity the material shows an increase in the hardness values due to a strain rate induced hardening.
Deformation Maps for Polycarbonate
Figure 5 presents results of the scratch deformation on the polycarbonate shown in terms of the subjectively assessed deformation modes. The data were drawn from the characteristics of many sample scratches. A range of normal loads (0.5 to 3.5N) and included angles (35 [degrees] to 150 [degrees]) were used, and scratches were produced at a sliding speed of 2.6 mm/s. The data show the dependence of the deformation mechanisms upon both the normal load and the imposed strain (cone angle). For the lower loads and the sharper cone angle cases, the material responds primarily with a ductile ploughing deformation. Ductile ploughing may be defined as a mainly plastic deformation where the material accommodates the indenter motion without any evidence of discrete fractures. Similar deformation characteristics are also found in the scratches produced by the 60 [degrees] and 90 [degrees] cone angles at the higher normal loads. An elastoplastic deformation accompanies the ductile ploughing mode within the region of the lower loads for the 60 [degrees] and 90 [degrees] included cone angles. When the load is increased for the sharpest cones (35 [degrees]), "edge crack" formation is encountered in the vicinity of the edges of the scratches [ILLUSTRATION FOR FIGURE 4 OMITTED]. As the scratch condition becomes more severe, with further loading, brittle machining and chip formation are visible [ILLUSTRATION FOR FIGURE 6 OMITTED]. For the 35 [degrees] degree cone and for normal loads over 2N deep grooving accompanied by a clear brittle behavior is the main feature of the deformation. For blunt cones ([greater than]90 [degrees]) the deformation moves from the ductile to the so-called ironing mode, where the indenter slides on the surface by mainly deforming the material elastically with some plastic deformation. Evidence of the indenter passage may be observed from laser topographical measurements as a result of the smoothening or polishing effect on the original sample surface asperities. For the lower loads and the blunt cones there is no detectable deformation found on the sample surface; the process is entirely elastic. For these cases, a fully elastic deformation occurs during scratching and the material recovers completely to its original shape. However, some evidence of the smoothening of the asperities is still detectable.
Figure 7 shows a similar deformation map to that shown in Fig. 5, but for a lower velocity of scratching (0.0026 mm/s) in order to investigate the influence of the scratching velocity upon the response of the material. The main difference observed for the PC at the two different scratch velocities is that at the very low velocity the material shows a more brittle-like nature. Consequently, the area in the map corresponding to ductile ploughing reduces; see Fig. 7. This increase in the ductile behavior at higher scratch velocity appears to be in contradiction to the expectation that at higher velocity the material should behave in a more brittle manner due to the increase in the strain rate. One explanation is that at the higher velocity the rate of energy dissipation at the contact point is higher, which may lead to a local adiabatic heating of the material. A rise In the temperature of the polymer will lead to the suppression of brittle failure. This proposition is consistent with the deformation map obtained at slightly higher temperature (T = 30 [degrees] C) for the same conditions of load and cone angle, and, at the lower scratching velocity. This deformation map is described further in the next section.
Effect of Temperature for Polycarbonate
Figure 8 is a deformation map drawn for the PC at a scratch velocity of 0.0026 mm/s and for the bulk temperature of 30 [degrees] C. The most striking change in the deformation pattern between Figs. 7 and 8 is that, with the increase of the temperature, the brittle failure is suppressed. For example, for the 35 [degrees] cone indenter the crack formation starts at a load of [approximately]1 N for the case of 30 [degrees] C. At this temperature, the brittle machining and chip formation regions are observed mainly for the lower cone angles, that is, for the higher strains and at the higher loads. One of the features of the deformation at the higher temperature is that the ductile ploughing mechanism is now a dominant mode of deformation. This is mainly due to the softening effect of the temperature on the bulk of the polymer. Figure 9 shows a micrograph of the PC scratched by the 45 [degrees] cone at 30 [degrees] C. This figure shows a reduction in the crack formation while the surface in the scratched area itself appears to be more smooth.
Deformation Maps for Ultrahigh Molecular Weight Polyethylene
The deformation mechanism for the PE has similar characteristics to those observed for other polymers. However, the transitions between the different mechanisms are very dependent upon the external parameters. Figures 10 and 11 show the deformation maps for PE for two different velocities of scratching. Clearly, these maps show that the ductile ploughing mechanism is a dominant mode of deformation for this material except at the very high strain values. An increase in the load tends to produce microcracks that become more extensive with the further increase of the load. This indicates that the deformation mechanism for the polymer is largely affected by the normal load or the stresses involved. This is true especially for the case of the higher strains (the lower cone angles). Unlike PC, for the PE case there is not much change in the deformation mechanisms for two different velocities adopted. However, there seems to be some reduction in the crack density for the higher scratch velocity case. This appears to be the result of an increase in the interface temperature because of the frictional heating.
The present study of scratching deformations has shown that strain, strain rate, and temperature are some of the important factors in deciding the material deformation mechanism and hence, the material removal process for two organic polymers; a PC and a PE. The measured response of the materials depends mainly upon the mechanism of the material deformation at the contact points. Hence, a polymer shows different scratch hardness values under different imposed conditions. This is because the energy dissipation or the energy required to scratch the surface of a polymer depends upon the nature of the deformations, viz. elastic, plastic, visco-elasto-plastic, brittle, etc. For PC, and also PMMA (1), the scratch hardness exhibits a maximum at intermediate levels of contact strain. The hardness values are also a function of velocity and ambient temperature. These data are partially consistent with the expected behavior based upon the premise that the bulk yield stress or ductile-brittle transition is a function of strain, strain rate, and temperature. Indirect evidence of the influence of adiabatic heating in the contact zone is inferred from the influence of the sliding velocity on the deformation processes. The effects of adiabatic frictional heating appear to be significant even at the relatively low velocities employed in the present work. In principle these effects may be assessed by first order calculations; see Lancaster (18) albeit for a different contact geometry, but for organic polymers. These calculations are accepted as being imprecise and direct measurement is not a viable alternative. However, the first order calculations suggest temperature rises in the contact zone of several degrees, which is consistent with the observed behavior of these systems. Similar data are noted for the PE, although here the nature of the deformation is such that it is not greatly changed during the variation of the contact strain; it remains largely plastic in its nature.
More generally, the scratch deformation maps provided in this paper illustrate the wide range of material flow and disruption phenomena that occur during the scratching process for two typical organic polymers.
The authors would like to thank Dow Chemical Company (USA) for the financial assistance provided for this project. They also acknowledge the support and encouragement provided by Dr. A. Willem deGroot and Dr. Frank Knoll of the same company.
q = A parameter in the calculation of the scratch hardness.
W = Normal load.
d = Scratch width.
[[Epsilon].sub.s] = Effective strain.
[Alpha] = Cone semi angle.
[Theta] = Attack angle.
[Mathematical Expression Omitted] = Effective strain rate.
v = Scratch velocity.
1. B. J. Briscoe, P. D. Evans, S. K. Biswas, and S. K. Sinha, Trib. Int., 29, 93 (1996).
2. B. J. Briscoe, E. Pelillo, and S. K. Sinha, "Scratching Maps for Polymers," data presented in the Institute of Physics Meeting on Predictive Methods in Tribology: I Mapping in Tribology, London (UK), Imperial College, 12 Sep 1995 (to be published elsewhere).
3. B. J. Briscoe, P. D. Evans, and J. K. Lancaster, J. Phys. D: Appl. Phys., 20, 346 (1987).
4. B. J. Briscoe, D. Briggs and D. G. Rance, in Comprehensive Polymer Science, Pergamon Press, Oxford, United Kingdom (1988).
5. B. J. Briscoe and P. D. Evans, Compas. Sci. & Technol., 34, 73 (1989).
6. B. J. Briscoe, S. K. Biswas, S. K. Sinha, and S. S. Panesar, Trib. Int., 26, 183 (1993).
7. S. C. Lim and M. F. Asby, Acta Metall., 35, 1 (1987).
8. B. Lawn, in Fundamentals of Friction: Macroscopic and Microscopic Processes, I. L. Singer and H. M. Pollock, eds., NATO ASI Series, Kluwer Academic Publishers, The Netherlands (1992).
9. B. Lamy, Trib. Int., 17, 35 (1984).
10. T. S. Barrett, G. W. Stachowiak, and A. W. Batchelor, Wear, 153, 331 (1992).
11. S. J. Bull, D. S. Rickerby, A. Matthews, A. Leyland, A. R. Pace, and J. Valli, Surf. Coat. Technol., 36, 503 (1988).
12. P. Burnett and D. S. Rickerby, Thin Solid Films., 154, 403 (1987).
13. S. J. Bull, Surf. Coat. Technol., 50, 25 (1991).
14. P. Hendquivst, M. Olsson, S. Jacobson, and S. Soderberg, Sur/f. Coat. Technol., 41, 31 (1990).
15. S. Lingard and K. K. Ng, Wear, 130, 367 (1989).
16. K. L. Johnson, Contact Mechanics, Cambridge University Press, England (1985).
17. B. J. Briscoe and S. K. Sebastian, Proc. R. Soc. Lond A, 452, 439 (1996).
18. J. K. Lancaster, in Polymer Science, A.D. Jenkins, ed., North Holland (1972).
19. G. H. West and J. M. Senior, Wear, 19, 37 (1972).
20. B. J. Briscoe, in Friction and Traction, pp. 81-93, D. Dowson, M. Godet, C. M. Taylor and D. Berthe, eds., Westbury House, IPC Press, Guildford, England (1981).
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||International Forum On Polymers - 1996|
|Author:||Briscoe, Brian J.; Pelillo, Enrico; Sinha, Sujeet K.|
|Publication:||Polymer Engineering and Science|
|Date:||Dec 1, 1996|
|Previous Article:||On the correlation between the viscosity of partially hydrolyzed polyacrylamide solution and the diffusion coefficient in the semidilute...|
|Next Article:||Rheological properties of phenophthalein poly(ether ether ketone).|