Available online at www.sciencedirect.com Wear 265 (2008) 143–149 Wear of materials used for artificial joints in total hip replacements L.V. Wilches, J.A. Uribe, A. Toro ∗ Tribology and Surfaces Group, National University of Colombia, Medellı́n, Colombia Received 19 July 2007; accepted 19 September 2007 Available online 26 November 2007 Abstract Two tribological pairs used for artificial joints were studied. Wear tests were performed in a pin-on-disc type machine modified to allow lubrication with bovine serum. The pins were made of either AISI F138 stainless steel or ASTM F136 titanium alloy, while the discs were manufactured with ultra-high molecular weight polyethylene (UHMWPE). The tested pairs were ASTM F138 stainless steel–UHMWPE and ASTM F136 titanium alloy–UHMWPE. The sliding velocity was fixed to 0.58 m/s for all the tests and the variation of friction force was registered as a function of time and normal load. Also, an electrochemical setup allowed monitoring the variation of the corrosion potential between the metallic pins and the bovine serum. The lower values of friction coefficient were measured when a thin film of polymer was transferred to the metallic surface, which led to a smooth interface and avoided debris generation. The viscoelastic behavior of UHMWPE was responsible for the increase in friction coefficient after a testing period, which depended on the normal load and the specific pair tested. © 2007 Elsevier B.V. All rights reserved. Keywords: Total hip replacements; Wear mechanisms; Wear debris; Corrosion potential; Friction coefficient 1. Introduction The use of alloys and polymers for total hip and knee replacements has increased significantly in recent years, mainly as a consequence of their good combination of mechanical and surface properties together with recent improvements in biocompatibility and bioactivity [1–4]. Metallic materials are highly appreciated due to their good wear resistance and high mechanical properties such as hardness and ductility, while polymers usually show low friction coefficients and excellent chemical stability when in contact with physiological fluids. Nevertheless, a significant number of total joint replacements fail long before the end of their expected life causing severe traumas to the patients . These failures are related to surface damage of the parts in contact, generally represented as wear of the polymers and corrosion of the alloys. Most of the implants used in Colombia for total hip replacements are manufactured either with stainless steel–UHMWPE or Ti6Al4V–UHMWPE pairs, so the knowledge of the surface damage mechanisms that act on these surfaces when put in contact to the human body is crucial to develop better materials ∗ Corresponding author. Tel.: +57 4 425 5339; fax: +57 4 425 5339. E-mail address: [email protected] (A. Toro). 0043-1648/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.wear.2007.09.010 as well as to improve the surgical practices and post-operative rehabilitation. Although a considerable amount of information is available regarding tests in joint simulators [6,7], standard test methods for wear testing of materials used in joint prostheses  and materials response to particular wear–corrosion process in the human body [9–11], the knowledge of the wear mechanisms acting on the surfaces in contact is still limited, in particular when factors such as surface finishing, chemical composition, stress distribution and medical record of the patient have to be considered from a comprehensive approach. The purpose of this work is to study the mechanisms of surface deterioration of materials used for total hip joint replacements when submitted to conditions close to those found in the human body, as well as to suggest improvements for future applications based on friction coefficient and corrosion potential results. 2. Experimental procedure 2.1. Materials AISI 316 LVM Stainless steel (ASTM F138 standard) bars with 6 mm in diameter, Ti6Al4V extra low interstitial (ELI) alloy (ASTM F136 standard) with 6.35 mm in diameter and ultra 144 L.V. Wilches et al. / Wear 265 (2008) 143–149 Table 1 Chemical composition, hardness and microstructure of the metallic materials studied Chemical analysis carried out by optical spectrometry (wt%). Fig. 1. Modified pin-on-disc wear testing machine. (a) General aspect of the device, (b) modified disc, and (c) detail of the positioning of pins and rings and configuration of electrochemical cell. L.V. Wilches et al. / Wear 265 (2008) 143–149 145 Table 2 Initial roughness of the surfaces in contact Material Ra (m) Rq (m) ASTM F138 (AISI 316LVM) pin ASTM F136 (Ti6Al4V) pin UHMWPE ring 0.04 ± 0.01 0.04 ± 0.01 0.24 ± 0.02 0.07 ± 0.01 0.07 ± 0.01 0.32 ± 0.03 high-molecular weight polyethylene, UHMWPE (TIVAR 1000) sheets with 6 mm in thickness were used in this investigation. The ASTM F138 and ASTM F136 bars were manufactured by Sandvik, while Quadrant Engineering Plastics Products provided the UHMWPE sheets. The measured chemical composition of the metallic materials, as well as their average hardness and microstructure are indicated in Table 1. The measured average hardness of the UHMWPE was 60–65 SH D. 2.2. Wear tests The wear tests were performed in a modified pin-on-disc machine , which allowed lubrication of the contact surfaces with bovine serum with a protein content of 0.030 ± 0.001 g/cm3 (measured by the Biuret method). The bovine serum was used to provide chemical and biological characteristics similar to those of synovial fluids found in natural joints . It was obtained from adult Holstein cows having an average weight of 650 kg, which were fed with unfertilized Kikuyo grass plus a 6% P mineral supplement. Blood collection was made by using a Vacutainer system; the serum was incubated at 37 ◦ C for 30 min and then centrifuged at 3000 rpm for 5 min. The whole process of blood collection, centrifugation and protein content measurement was performed in the Animal Production Laboratory of the National University of Colombia in Medellı́n. In each test, the pin was fixed to a rigid arm and put in contact with a rotating ring under the application of a normal load by dead weights, as shown in Fig. 1. All the pins were cylinders with 6 mm in diameter and 12 mm long, and the rings had 170 mm of internal diameter, 190 mm of external diameter and a thickness of 6 mm. The homogeneous contact between the pin and the ring was verified by checking the flatness of the surfaces with a dial indicator having a reading accuracy of 0.002 mm. The relative slip speed between pins and rings was fixed to 0.58 m/s and the normal loads were 5, 15, 30, and 50 N, which correspond to mean contact pressures of 177, 531, 1060 and 1767 kPa, respectively. Three tests were performed for each condition of normal load and tribological pair. The total testing time was 7 h for all the pairs studied and the temperature of the ambient was controlled to 25 ◦ C. The tribological pairs tested were as follows: (a) ASTM F138 stainless steel pins against UHMWPE rings, and (b) ASTM F136 alloy pins against UHMWPE rings. The surface of metallic pins was polished with emery papers and abrasive cloths with diamond particles of 0.1 m in diameter, which led to the roughness values shown in Table 2. On the other hand, the surface finishing of the UHMWPE rings was obtained through lathe machining (see Table 2). All the samples were washed, dried and isolated after washing in order to avoid contamination. The microstructure of the stainless steel Fig. 2. Variation of friction coefficient with testing time for different normal loads. AISI 316LVM pin sliding against UHMWPE ring. was composed of austenite grains with average size of 25–30 m and the titanium alloy showed an ␣-␤ distribution with a volume fraction of ␤ of circa 20%. The time-variation of the corrosion potential between the working electrode (pin) and the bovine serum during the wear tests was measured with the aid of the arrangement shown in Fig. 1c. Before and after each test, the samples were ultrasonically cleaned, dried in warm air and characterized by optical and scanning electron microscopy in order to identify the damage mechanisms at the surface. The friction coefficient and corrosion potential were registered every second with the aid of a data acquisition card and software Labview 5.1 provided by National Instruments under an educational contract. 3. Results and discussion Figs. 2 and 3 show the variation of the friction coefficient of ASTM F136–UHMWPE and ASTM F138–UHMWPE pairs as a function of testing time for different normal loads. It can be observed that the friction coefficient has a trend to increase suddenly at a testing time that depends on the magnitude of the normal load. For both systems studied, when the normal load was 5 N no significant changes in friction coefficient were observed, but for a normal load of 50 N the effect was noteworthy. This particular response of the surfaces in contact can be related to adhesive phenomena at the interface, since UHMWPE Fig. 3. Variation of friction coefficient with testing time for different normal loads. ASTM F138 pin sliding against UHMWPE ring. 146 L.V. Wilches et al. / Wear 265 (2008) 143–149 Fig. 4. Evidences of adhesion of UHMWPE to metallic pins after 7 h of testing. Normal load: (A) 5 N; (B) 15 N; (C) 30 N; (D) 50 N. The polymer layers correspond to the clearer areas in the image and the black arrows indicate the sliding direction. layers were transferred to the pin’s surface during the tests, being greater the amount of material transferred when the normal load increased (see Fig. 4). For a normal load of 5 N the amount of polymer transferred to the pin was small, but in the case of normal load of 50 N practically all the pin’s surface was covered by the UHMWPE layer, especially in the case of UHMWPE–AISI 316LVM pair. In addition, for a given normal load the increase of friction coefficient at a specific interval during the tests is also related to the adhesion of polymer to the metallic surface, as can be seen in Fig. 5 for AISI 316LVM samples tested using a normal load of 50 N, which were analyzed in SEM before and after the sudden increase in friction coefficient reported in Fig. 3. The increase in the amount of transferred polymer to the pin’s surface with normal load is stimulated by a marginal lubrication regime at the effective area of contact, which is a result of localized high contact pressures that prevents a hydrodynamic regime. The images in Figs. 4 and 5 suggest that for the most part of the testing time in the metal–polymer pairs the contact was of polymer–polymer type, so the measured friction force can be related to cohesive failures at the sub-surface of the ring instead of being a consequence of the adhesive forces at the stainless steel–UHMWPE interface, which leads to delamination of polyethylene. A similar observation was made by Saikko et al.  in CoCr alloy–UHMWPE pairs tested in a three-axis knee wear simulator with ball-on-flat, where polymer wear debris between 0.1 m and 1 m were reported. Furthermore, the fluctuations observed in Figs. 3 and 4 could be a consequence of periodic stick–slip cycles caused by adhesion and subsequent cohesive failure at the sub-surface of the polymer. Fig. 5. Pin’s surface of ASTM F138 stainless steel tested against UHMWPE rings under a normal load of 50 N, before (A) and after (B) the sudden increase of friction coefficient. Parallel marks indicate the sliding direction. L.V. Wilches et al. / Wear 265 (2008) 143–149 147 The effect of wear of UHMWPE–metal pairs on osteolysis and early loosening in total hip replacements has been reported in literature ; a polymer transfer layer was observed by Gispert et al.  in UHMWPE–ASTM F138 pairs and by Li et al.  in UHMWPE–ASTM F136 pairs after pin-on-disc tests similar to those described in the present work. Moreover, Watters  also found these transfer layers at the surface of ASTM F136 femoral heads in contact with UHMWPE liners, after experiments carried out in a conventional hip simulator. Regarding the wear mechanisms, Buford et al. , Gongde , Liu  and Ahlroos  independently concluded that the formation of adhesive joints and subsequent cohesive subsurface failures govern the initial stages of surface degradation of UHMWPE in contact with metals, while localized fatigue effects become more significant for longer testing times. The results found in the present work support these conclusions. 3.1. Effect of testing conditions on corrosion potential The variation of corrosion potential with testing time for the metal–polymer systems studied is shown in Fig. 6. In the ASTM F138–UHMWPE pair two regions can be distinguished: the first region is characterized by a strong decrease in the values of potential (circa 350 mV during the first 30 min of the test) and the second one shows a slight decrease to a stability value of about −650 mV after 7 h of testing. In contrast, the ASTM F136–UHMWPE system shows a quite different behavior: there is an initial increase of the corrosion potential during the first 2 h of the test, followed by a period in which no significant variations are observed. Since the corrosion potential established in titanium alloy–polyethylene pairs was higher than that of stainless steel–polyethylene pairs, the surface of the titanium alloy can be considered more resistant to uniform corrosive attack under the specific corrosion–wear conditions employed. Nevertheless, surface examination of both materials after the tests did not reveal any evidences of localized corrosion, which means that the passive layer was unaffected in all cases. The fluctuations in the curves can be associated to changes in the thickness of the diffusive layer close to the metal surface and variations of the concentration of the solution. Small temperature variations and high contact pressures can induce the precipitation of proteins which are temporally adhered to metal, changing its electrochemical response and therefore affecting the mea- Fig. 6. Variation of corrosion potential with testing time for different normal loads. Metallic pins sliding against UHMWPE ring. sured corrosion potential. Although no significant changes in colour and texture of the bovine serum were detected after naked eye observation, the microscopic examination of the surfaces after the tests revealed the presence of tiny grains adhered to the metallic pins (Fig. 7), whose chemical composition is related to that of the bovine serum. Also, the changes in effective contact area between the pin and the ring due to adhesion of polymeric layers to metal slow down the electrochemical reaction since these layers act as an isolating material. 3.2. Surface damage and wear mechanisms The predominant wear mechanism in metal–polyethylene pairs was adhesion, which caused thin polymer films to be transferred to the surface of the pins since the beginning of the tests. This film probably covers the surface of the pins almost com- Fig. 7. (a) Grain precipitated from the bovine serum, which was observed at the surface of ASTM F138 stainless steel pin and (b) EDS spectra of the grain. 148 L.V. Wilches et al. / Wear 265 (2008) 143–149 Fig. 8. Evidences of plastic deformation (a) and cracking (b) of UHMWPE films transferred to the surface of ASTM F138 pins. Load 50 N. Testing time 7 h. Fig. 9. Wear marks and signals of plastic deformation at the surface of UHMWPE ring sliding against an AISI 316LVM pin. Load 50 N. Testing time 7 h. pletely during the first stages of the tests and acts as a lubricant afterwards. However, when the normal load and testing time increase the adhesive forces at the interface overcome the cohesive forces inside the polymeric ring, which leads to failure beneath the interface. The material that is removed from the ring is deposited onto the pin’s surface and from that point on the contact is mainly of polymer–polymer type. For longer testing times another important change is observed at the surface of the pins, as can be seen in Fig. 8. The polymer film shows cracks and signals of plastic deformation such as prows, especially in the areas close to the borders of the contact zone. It is proposed that cracks could be a consequence of thermal fatigue related to the difference between the thermal expansion coefficients of the metal surface and the polymeric film, while plastic deformation is caused by the high contact pressures achieved at the interface. On the other hand, analysis of the worn surfaces of the rings revealed plastic deformation and damage to the texture left by the machining process prior to wear tests (Fig. 9). 3.3. Viscoelasticity effects The effect of viscoelastic response of the UHMWPE surface sliding against ASTM F138 stainless steel can be seen in Fig. 10. The average value of friction coefficient did not change considerably when the normal load was gradually reduced from 50 N to 5 N, which means that the relationship Ff = W applies to Fig. 10. Variation of friction coefficient with normal load during a single test for contact between AISI 316LVM and UHMWPE. Load 50 N (a) the load is reduced during the test. (b) The load is increased during the test. L.V. Wilches et al. / Wear 265 (2008) 143–149 the tribo-system, i.e., the friction force depends mainly on the number of plastically deformed asperities (the higher the normal load, the higher the number of deformed asperities) . On the other hand, when the normal load was gradually increased during a single test, an increase in friction coefficient with time was observed. The change in friction coefficient could be the result of a change in the relation between the number of plastically deformed asperities and the normal load, and/or a substantial change in mechanical properties of the polymer due to the high temperature at the interface, which is clearly more feasible when the normal load is higher [22,23]. 4. Conclusions • The most important wear mechanisms in the metal–polymer pairs under the specific conditions studied were the adhesion of the polymer to the metallic surface and subsequent failure at the sub-surface of the polymer. • The transference of a thin polyethylene layer to the metallic surfaces promoted a lubricating effect at the interface for low normal loads and short testing times. When this condition was achieved the friction coefficient was always lower than 0.04. • The amount of polymer adhered to the surface increased with the normal load and was affected by the viscoelastic response of the UHMWPE, especially for normal loads over 30 N. For higher normal loads and longer testing times the lubricating effect was not predominant and significant plastic deformation of the polyethylene was observed. • The measured corrosion potential of ASTM F136 titanium alloy in bovine serum during the wear tests was higher than that of ASTM F139 stainless steel, which indicates a better resistance of the former to uniform corrosion under the specific conditions studied. Acknowledgement The authors thank to Research Division of National University of Colombia (Medellı́n) for financial support, project no. 20301004575, Code DIME 30805798. References  A. von Reccum, Handbook of Biomaterials: Evaluation—Scientific, Technical and Clinical Testing of implant materials, Taylor & Francis, London, 2002, pp. 18–37. 149  A. Ravaglioli, A. Krajewski, Bioceramics Materials—Properties and Applications, Chapman & Hall, New York, 2001.  M. Long, H.J. Rack, Titanium alloys in total joint replacement—a material science perspective, Biomaterials 19 (1998) 1621–1639.  S.W. Zhang, State of the art of polymer tribology Tribology International, vol. 31, Elsevier Science Ltd., Great Britain, 1998, p. 53.  O. Makoto, Control of crystalline orientation to enhance the wear resistance of ultra-high molecular weight polyethylene crystallization cups for artificial joints, Wear 255 (2003) 1045–1050.  A. Essner, G. Schmidig, A. Wang, The clinical relevance of hip joint simulator testing: in vitro and in vivo comparisons, Wear 259 (2005) 882– 886.  J. Geringer, B. Forest, P. Combrade, Fretting–corrosion of materials used as orthopaedic implants, Wear 259 (2005) 943–951.  ASTM F-732-00 Standard Test Method for Wear Testing of Polymeric Materials Used in Total joint Prostheses, ASTM, USA, 2000.  J. Yao, T. Blanchet, D. Murphy, M. Laurent, Effect of fluid absorption on the wear resistance of UHMWPE orthopedic bearing surfaces, Wear 255 (2003) 1113–1120.  G.M. Bartenev, V. Lavrentev, Friction and Wear of Polymers, Elsevier, New York, 1981, pp. 1–29.  B. Shi, O. Ajayi, E. Fenske, H. Liang, Tribological performance of some alternative bearing materials for artificial joints, Wear 255 (2003) 1016–1023.  V. Saikko, T. Ahlroos, O. Calonius, A three-axis knee wear simulator with ball-on-flat contact, Wear 249 (2001) 310–315.  Z.M. Jina, et al., Mini-symposium: biomechanics for the frcs orth exam chapter of biotribology, vol. 20, Curr. Orthop., 2006, pp. 32–40.  M.P. Gispert, et al., Friction and wear mechanisms in hip prosthesis. Comparison of joint materials behaviour in several lubricants, Wear 260 (2006) 149–158.  X. Li, et al., New insights into wear of Ti6Al4V by ultra-high molecular weight polyethylene under water lubricated conditions, Wear 250 (2001) 553–560.  E.P. Watters, Wear of artificial hip joint material, Chem. Eng. J. 112 (2005) 137–144.  A. Buford, et al., Review of wear mechanisms in hip implants: Paper I, Gen. Mater. Des. 25 (2004) 385–393.  L. Gongde, A study on sliding wear mechanism of ultra high molecular weight polyethylene/polypropylene blends, Wear 256 (2004) 1088– 1094.  C.Z. Liu, Tribological behaviour of PA/UHMWPE blend under dry and lubricating condition, Wear 260 (2006) 109–115.  T. Ahlroos, Effect of Lubricant on the wear of Prosthetic Joints Materials, Acta Polytechnica Scandinavica Mechanical Engineering Series No. 153, Doctoral Thesis, Helsinki University of Technology, 2001.  I.M. Hutchings, Tribology: Friction and Wear of Engineering Materials, Edward Arnold, London, 1992, p. 22.  K.C. Ludema, Friction Wear Lubrication: Textbook in Tribology, CRC Press, Inc., New York, 1996, p. 141.  E. Rabinowicz, Friction and Wear of Materials, John Wiley & Sons, New York, 1965, pp. 167–191.