The tribological behaviour of different clearance ... - SAGE Journals

1061

The tribological behaviour of different clearance MOM hip joints with lubricants of physiological viscosities X Q Hu1*, R J K Wood1, A Taylor2, and M A Tuke2 1

National Centre for Advanced Tribology at Southampton (nCATS), School of Engineering Sciences, University of Southampton, UK 2 Finsbury Development Ltd, 13 Mole Business Park, Leatherhead, Surrey UK The manuscript was received on 13 December 2010 and was accepted after revision for publication on 28 June 2011. DOI: 10.1177/0954411911419061

Abstract: Clearance is one of the most influential parameters on the tribological performance of metal-on-metal (MOM) hip joints and its selection is a subject of considerable debate. The objective of this paper is to study the lubrication behaviour of different clearances for MOM hip joints within the range of human physiological and pathological fluid viscosities. The frictional torques developed by MOM hip joints with a 50 mm diameter were measured for both virgin surfaces and during a wear simulator test. Joints were manufactured with three different diametral clearances: 20, 100, and 200 mm. The fluid used for the friction measurements which contained different ratios of 25 per cent newborn calf serum and carboxymethyl cellulose (CMC) with the obtained viscosities values ranging from 0.001 to 0.71 Pa s. The obtained results indicate that the frictional torque for the 20 mm clearance joint remains high over the whole range of the viscosity values. The frictional torque of the 100 mm clearance joint was low for the very low viscosity (0.001 Pa s) lubricant, but increased with increasing viscosity value. The frictional torque of the 200 mm clearance joint was high at very low viscosity levels, however, it reduced with increasing viscosity. It is concluded that a smaller clearance level can enhance the formation of an elastohydrodynamic lubrication (EHL) film, but this is at the cost of preventing fluid recovery between the bearing surfaces during the unloaded phase of walking. Larger clearance bearings allow a better recovery of lubricant during the unloaded phase, which is necessary for higher viscosity lubricants. The selection of the clearance value should therefore consider both the formation of the EHL film and the fluid recovery as a function of the physiological viscosity in order to get an optimal tribological performance for MOM hip joints. The application of either 25 per cent bovine serum or water in existing in vitro tribological study should also be revised to consider the relevance of clinic synovial fluid viscosities and to avoid possible misleading results. Keywords: total hip joint replacement, metal-on-metal, clearance, friction, viscosity

1

INTRODUCTION

The lower friction levels associated with metal-onmetal (MOM) hip joints can reduce acetabular cup loosening and if the friction is due to a fluid-film or *Corresponding author: National Centre for Advanced Tribology at Southampton (nCATS), School of Engineering Sciences, University of Southampton, SO17 1BJ, UK email: [email protected]

mixed lubrication mechanism it can also reduce the incidence of wear and metal ion release. Although friction can be affected by various factors, in fluidfilm lubricated bearings clearance is probably one of the most influential effects and it should be carefully controlled. Clearance of a hip joint is defined as the difference between the diameters of the acetabular cup and the femoral head. Although there are already many studies on this topic, the influence of clearance on the friction and wear is still not fully Proc. IMechE Vol. 225 Part H: J. Engineering in Medicine

1062

X Q Hu, R J K Wood, A Taylor, and M A Tuke

understood and the selection of clearance is the subject of considerable debate. Data from both friction and wear hip simulator studies have indicated that provided a substantial film of fluid exists, a small clearance can reduce both friction and bedding-in wear in the early postoperative period [1– 11]. This is further supported by theoretical studies based on the analysis of elastohydrodynamic lubrication (EHL) and contact mechanics of MOM hip joints, which demonstrate that a decrease in clearance can increase the predicted lubricating film thickness between the bearings so that the both friction and wear can be reduced [12–19]. While these in vitro studies overwhelmingly support the application of a smaller clearance, they also conclude that the clearance cannot be too small since this may reduce the load-carrying capacity and hence increase wear. At the same time, clinical results have suggested that the relationship between wear and clearance could be more complicated than indicated by simulator studies. McKellop et al. [20] studied three different MOM hip prostheses with diameters ranging between 35 and 42 mm diameter over a 20 year period and observed no significant correlation between the wear and the clearance, except one pair of components with an extraordinary large clearance of 1748 mm and extremely high wear. Reinisch et al. [21] analysed some retrieved 28 mm diameter MOM hip prostheses with early loosening. In this study, the wear rates declined slightly with an increase in clearance in the range between 60 and 96 mm in the early stage of implantation, although the trend was not statistically significant. Retrievals of second-generation MOM hip joints that had been implanted for between 20 and 30 years indicated that they generally had relatively larger clearances (.120 mm) than most current in vitro studies [22–24]. It is clear that further research needs to be performed on the choice of the optimal clearance during the design of MOM hip joints. The effect of the fluid viscosity on the tribological performance of MOM hip joints, and the span of viscosities encountered clinically are still not fully explained. The friction in a lubricated bearing is directly related to the fluid viscosity, and the viscosity is also one of the key properties that determine the formation of an EHL film between the bearing surfaces [8]. The viscosity of a fluid is the ratio of the shear stress produced to the rate of shear strain. Normal synovial fluid is non-Newtonian in nature with a shear-rate-dependent viscosity. The viscosities for normal, osteoarthritic, and inflammatory synovial fluids have been reported to be 0.01–0.4, 0.0025–0.2, and 0.001–0.07 Pa s respectively at a shear rate of 300 s–1 [25–27]. Scholes and Unsworth Proc. IMechE Vol. 225 Part H: J. Engineering in Medicine

[28] reported a viscosity value of 0.006 Pa s for pooled human synovial fluid taken from patients with rheumatoid and inflammatory arthritis using a shear rate of 3000 s–1. Other studies have reported that peri-prosthetic fluids from patients with a primary joint replacement displayed decreased viscosities, but were similar to normal synovial fluid [29]. This recent progress in the understanding of the properties of physiological fluid has shown that to date the range of measured viscosities has not been captured in full during in vitro evaluation of MOM bearings. Currently, the majority of in vitro tribological studies use 25 per cent bovine serum or water as the lubricant, which has a very low viscosity of around 0.001 Pa s [5–11, 30–38]. Some studies use aqueous solutions of carboxy methyl cellulose (CMC) with viscosities between 0.0035 and 0.1152 Pa s, but the relation with the clearance has not been fully investigated [39, 40]. The aim of this paper is to study the lubrication behaviour of different clearance MOM hip joints within the range of human physiological and pathological fluid viscosities. 2

MATERIALS AND METHODS

The components tested in this study were 50 mm diameter MOM hip prostheses (ADEPT, Finsbury Ltd, Leatherhead, UK) with diametrical clearances of 20, 100, and 200 mm that were specifically manufactured for this study (Table 1). Both the heads and the cups were manufactured from medical grade high carbon (C . 0.2 per cent) as cast CoCrMo alloy (ISO5832-4). The surface roughness was controlled to be within Ra \0.02 mm which was measured by a Surftest SV-400 surface tester (Mitutoyo, Kawasaki, Japan). The sphericity was determined within a tolerance of \2 mm in both the polar and equatorial directions using a Roundness machine. The fluids used for the friction tests consisted of different ratios of 25 per cent newborn calf serum and CMC with the addition of 0.1 per cent sodium azide to retard bacterial growth and with a protein content of approximately 11–15 mg/ml. The fluid viscosities ranged from 0.001 to 0.71 Pa s (as determined by a Physica Rheolab MC100 cone-on-plate viscometer (Anton Paar GmbH, Graz, Austria) at a shear rate of g = 300 s–1) thus displaying similar rheological properties to synovial fluid [41] and covering the major range of the viscosities of synovial fluid found in the human body. Friction testing was carried out on a ProSim friction simulator (Simulation Solutions, Leeds UK), see Fig. 1. The friction simulator has two driven axes;

The tribological behaviour of different clearance MOM hip joints with lubricants of physiological viscosities

1063

Table 1 Test matrix of friction measurement Test

Friction measurement with virgin samples

Friction measurement during a simulator wear test

Clearance (mm) Number of samples Viscosities tested (Pa s)

100 4 0.001 and 0.1

20 1 0.001 to 0.71

200 4 0.001 and 0.1

200 2 0.001 to 0.71

Fig. 1 The ProSim friction simulator and configuration of testing samples during the friction measurement

100N

Friction measurement zone

2000N

Dynamic load Moon forward Moon backward

+/-24°

load and rotation. The acetabular component was supported on a pair of externally pressurized oil bearings that created a negligible friction level compared to the friction from the hip bearing. These pedestal bearings were further held on a low-friction carriage, which was supported on externally pressurized bearings, so the friction in the horizontal x– y plane was negligible. A dynamic trapezoidal-form loading cycle was vertically applied to the femoral head with a minimum load during the swing phase of 100 N and a maximum load of 2000 N throughout the stance phase to simulate a simplified form of the human gait cycle published by Paul [42]. A harmonic motion of amplitude 6 24° (total 48°) was applied to the femoral head in the flexion–extension plane with a frequency of 1 Hz. The relationship between the load and motion is illustrated in Fig. 2. The friction force was measured using a piezoelectric transducer fitted between the rotational frame of the acetabular cup and the pressurized bearings during the highest load and the highest rotation speed (the grey strip in Fig. 2). Each measurement was undertaken for at least 120 cycles and a reading was taken every ten cycles. Since any vertical misalignment between the rotational axis of the acetabular carriage and the acetabular cup could not be completely eliminated in situ and since this leads to an additional eccentricity torque, this torque was accommodated by performing two runs, one during the application of the peak load while the head frame was moving forward and the other with the peak load applied while the head frame was moving backwards. The loading cycles each produced a forward torque Tf and backward torque Tb from which the true torque T was calculated as T = ðjTf j + jTb jÞ=2. Three measurements were taken at each forward and backward motion and an average value was calculated for these measurements. Four virgin samples were tested with 100 and 200 mm clearances. The frictional torque was measured with two different viscosities of lubricant: 0.001 and 0.1 Pa s, which represents the lower and higher end of physiological viscosity [25–29, 41]. After the first friction measurement, one 20 mm, one 100 mm, and two 200 mm clearance samples were tested on a wear simulator test using up to 5 400 000 cycles. Samples were removed from the

100 1 0.001 to 0.71

1 sec

Fig. 2 The relation between the load and the motion of the ProSim friction simulator

wear simulator at different stages of the wear test and friction measurements were performed. Viscosity values from 0.001 to 0.71 Pa s were tested for these worn samples. Before each friction measurement, the components were carefully cleaned with detergent, rinsed with water and then in acetone. The wear simulator was an MTS hip simulator system (MTS Systems Corporation, Eden Prairie, MN, USA) with a 6 24° biaxial-rocking motion to Proc. IMechE Vol. 225 Part H: J. Engineering in Medicine

1064


represent the flexion–extension and adduction– abduction movements of the femur during walking [43]. All components were positioned in a physiologically representative manner for the wear testing, i.e. the cup was positioned at 35° to the horizontal above a moving femoral head which was at an angle of 23° to the vertical. The loading profile used during the tests was based upon the data published by Paul [42] with maximum load of 2450 N and minimum load of 50 N. The lubricant used during the wear simulator test was 25 per cent newborn calf serum with the addition of 0.1 per cent sodium azide to retard bacterial growth; it had a viscosity of approximately h300 = 0.001 Pa s. During the friction measurements, the femoral head and the acetabular cup were positioned at the same angles as in the hip wear simulator to ensure the friction was measured over the location of the wear patches (Fig. 1). Frictional torques were measured twice over the whole range of viscosities from 0.001 to 0.71 Pa s, first from high to low viscosities and then from low to high viscosities, and the average frictional torque was taken for each viscosity. After the friction measurement, the samples were put back in the wear simulator, the lubricant was changed, and the wear test continued up to 5 400 000 cycles. The test matrix is presented in Table 1.

3

RESULTS

The mean values of the frictional torques for virgin surfaces of the bearings with 100 and 200 mm clearances are shown in Fig. 3. With the lower viscosity lubricant (0.001 Pa s), the frictional torque of the 200 mm clearance bearing was higher than for the 100 mm clearance bearing. With the higher viscosity lubricant (0.1 Pa s), however, the frictional torque of the 200 mm clearance bearing was lower than the 100 mm clearance bearing. This difference was highly significant (t-test, p\0.01).

Fig. 4 Frictional torque of 100 and 200 mm clearances during the progress of wear

Since the frictional torques of worn samples mainly changed in the lower end of the viscosity range between 0.001 and 0.1 Pa s and levelled off after 0.2 Pa s, Figs 4 and 5 only present the frictional torques obtained in the range between 0.001 and 0.2 Pa s. In Fig. 4, the frictional torque reduces with the progress of wear for both the 100 and 200 mm clearances, which is consistent with the running-in process. In the lower range of viscosity values, the frictional torques obtained at the 100 mm clearance are lower than for the 200 mm clearance. When the viscosity reaches the range between 0.01 and 0.06 Pa s, the frictional torque of the 100 mm clearance bearing surpasses that of the 200 mm clearance bearing and this difference becomes wider with the increase in viscosity level. In Fig. 5, the frictional torque for the 20 mm clearance bearing is higher than for the 200 mm clearance bearing and this difference becomes wider with the progress of wear. Only in the very low range of viscosity do the two bearings have a similar frictional torque level. After 1 900 000 cycles of the wear test, the frictional torque of the 200 mm clearance bearing reduces with the lower viscosity lubricant, but remains constant with the higher viscosity lubricant. With the 20 mm 20um_0.8M 20um_1.3M 20um_1.9M 20um_4.0M

Friction Torque (Nm)

16

200um_0.8M 200um_1.3M 200um_1.9M 200um_4.0M

12 8 4 0 0

Fig. 3 Frictional torque of 100 and 200 mm clearances with two different viscosities

Proc. IMechE Vol. 225 Part H: J. Engineering in Medicine

0.04

0.08

0.12

Viscosity η300 (Pa s)

0.16

Fig. 5 Frictional torque of 20 and 200 mm clearances during the progress of wear


b. 100μm Clearance

12



a. 200μm Clearance

8 4 0 0

40

1065

80

120

Cycles

12 8 4 0

0

40

80

120

Cycles


c. 20μm Clearance 0.001 Pa s

12

0.04 Pa s 8

0.15 Pa s 0.36 Pa s

4

0.71 Pa s

0 0

40

80

120

Cycles

Fig. 6 Frictional torque change during the first 120 cycles of measurement

clearance, the frictional torque increases dramatically for all viscosity levels after 1 900 000 cycles. Figure 6 presents the frictional torque for the first 120 cycles of measurements. The lubrication and friction change from the beginning of the measurement until the steady state is reached can be clearly observed in this figure. For the 200 mm clearance bearing and all viscosities apart from the two highest (0.36 and 0.71 Pa s), the frictional torque remains constant throughout the test, and the higher the viscosity the lower the frictional torque (Fig 6(a)). For the two highest viscosities (0.36 and 0.71 Pa s), the frictional torque starts from a very low value and increases gradually to reach a steady state at about 100–120 cycles, and the higher the viscosity the steeper the gradient of the curve. For the 100 mm clearance bearing, the frictional torque starts from a very low value and increases rapidly to reach a steady state after the first 20 cycles for all viscosities except for the 0.001 Pa s test and the higher the viscosity the higher the steady state frictional torque (Fig 6(b)). For the 20 mm clearance bearing, the frictional torque is constant from the beginning of the measurement (Fig. 6(c)). 4

DISCUSSION

The lubrication between the bearings of hip joints is a dynamic and continuous process during a gait

cycle; it can be divided into two phases: synovial fluid recovery in the swing phase and the formation of an EHL film in the stance phase. During the swing phase, the weight of the body is removed allowing synovial fluid recovery due to the low load and relatively high entraining velocity (‘A’ in Fig. 7). The stance phase starts when the heel strikes the ground. Some of the synovial fluid is squeezed out from between the bearing surfaces reducing the film thickness and then the femoral head begins to rotate at the start of the stance phase relative to the acetabular cup. This relative movement uses the remaining synovial fluid to establish an EHL film between the bearing surfaces which separates or partially separates the bearing surfaces (‘B’ in Fig. 7). Lubrication of MOM hip joints therefore depends on both the formation of an EHL film during the stance phase and the recovery of synovial fluid during the swing phase. If the fluid cannot be fully recovered, depletion of fluid occurs and an EHL film cannot be established. In this case, direct contact will occur between the bearings so that both friction and wear will increase. The significance of lubricant recovery has already been reported in the literature. Williams et al. [44] discovered from their wear study that increasing the swing phase load from 100 to 280 N in the same hip simulator resulted in an over tenfold increase in the wear of MOM hip replacements. From a friction test, Brockett et al. [45] Proc. IMechE Vol. 225 Part H: J. Engineering in Medicine

1066


Fig. 7 The synovial fluid recovery and the formation of an EHL film during swing and stance phases of a gait cycle

concluded that a higher swing phase load results in a depleted lubricating film, hence elevated friction due to increased direct contact. Theoretical and experimental studies have shown that a smaller clearance can promote the formation of an EHL film in MOM bearings [1–11]. However, if the clearance is too small it may restrict or prevent synovial fluid recovery during the swing phase, which will cause more direct contact between the bearings. A ‘suction fit’ between the femoral head and acetabular cup may further explain this mechanism. Suction fit is a commonly observed phenomenon in MOM hip joints [46, 47]; it is a result of a vacuum between the two components, especially when the clearance is small and the viscosity is high. This suction force can hold the head and the acetabular cup together so that the recovery of synovial fluid is impossible even when the hip is not loaded during the swing phase. Although an increased lubricant viscosity can improve the formation of an EHL film, it can also enhance the ‘suction fit’, which means the higher the viscosity, the higher the suction force. A higher viscosity level also requires a larger clearance to allow full synovial fluid recovery during the swing phase. It is therefore important that the selection of the clearance should consider the viscosity, especially in the physiological range, to balance both the formation of EHL film and synovial fluid recovery. For a given bearing, a change in the trend in frictional torque level is observed as the viscosity of the fluid is increased due to the two opposing mechanisms: EHL film formation and fluid recovery. In Fig. 3, the lower viscosity (0.001 Pa s) bovine serum Proc. IMechE Vol. 225 Part H: J. Engineering in Medicine

could be recovered for both clearances so that the frictional torque is mainly decided by the formation of the EHL film. In this case, the smaller clearance has a lower frictional torque due to a better EHL film. With the higher viscosity lubricant (0.1 Pa s), however, the recovery of lubricant gradually becomes predominant, especially for smaller clearances. Although a higher viscosity level can improve the formation of an EHL film, this may lead to poor recovery of the lubricant, subsequently causing fluid depletion and direct contact of the bearing so that the frictional torque is increased. The larger clearance of 200 mm allows good recovery of lubricant with increased viscosity and therefore it improves the EHL film under this viscosity level so that the frictional torque is reduced. This explanation is corroborated by the measurements made during the wear process. In Fig. 4, the frictional torque of the 100 mm clearance bearing is lower in the lower range of viscosity values due to a better EHL film. In the higher range of viscosity, the frictional torque of the 100 mm clearance bearing is greater than for the 200 mm clearance bearing at all stages of the wear process. Figure 6 illustrates the lubrication change at the beginning of the frictional torque measurement, i.e. before the onset of the steady state. Before the friction measurement, the bearings are fully covered by the lubricant. Over the subsequent cycles, with the progress of the frictional torque measurement, however, the lubricant is squeezed out by the application of the load and it is recovered during the swing phase (as illustrated in Fig. 7). If the lubricant is fully recovered to the original state, the frictional torque should remain constant from the beginning to the steady state phase of the measurements. If the lubricant cannot be fully recovered, the frictional torque will increase to a steady state condition where the lubricant is recovered to the condition of the previous cycle. The poorer the recovery then the more rapid is the increase in frictional torque. In Fig. 6(a), the lubricant of the 200 mm clearance bearing can be fully recovered from the very beginning of the measurement for viscosity levels of 0.001, 0.042, and 0.15 Pa s since the frictional torque curves remain constant throughout the torque measurement. There is an increase in frictional torque for the 0.36 Pa s viscosity value, indicating that full recovery does not occur in this case. For the 100 mm clearance bearing (Fig. 6(b)), the lubricant is squeezed out quickly as intimated by the sharp increase in frictional torque in the first 20 cycles for most of the viscosity values. Only for the 0.001 Pa s viscosity is the lubricant fully recovered and the frictional torque remains constant during whole torque measurement. For the


20 mm clearance joint (Fig. 6(c)), all the friction curves are high from the very beginning. Since the first measurement of the frictional torque test was taken at the tenth cycle for each measurement, the depletion of lubricant for the 20 mm clearance may have already occurred before the tenth cycle. The curves in Fig. 6(c) therefore represent lubrication with poor recovery during the unloaded phase. All these results indicate that the recovery of lubricant and the formation of an EHL film are both essential for the lubrication of MOM hip joint bearings. It is therefore important to consider both factors during the selection of an optimal clearance value. In the meantime, the application of 25 per cent bovine serum in existing in vitro tribological studies should be revised because its viscosity is lower than clinic synovial fluid viscosities; otherwise misleading results may be produced. The significance of viscosity in deciding the lubrication conditions may also explain some clinic observations, such as some high wear retrievals even though the implant was well positioned.

3

4

5

6

7

8

5

CONCLUSIONS

The selection of the extent of the clearance (the difference between the diameters of the acetabular cup and the femoral head) is crucial for the lubrication and tribological performance of MOM hip joints. The chosen value for the clearance is a function of the viscosity of the physiological lubricant in the human body. A small clearance can increase the EHL film thickness, but can also prevent fluid recovery during the unloaded phase of walking and therefore compromise lubrication. Increasing the clearance of the MOM bearing improves lubrication for higher viscosity fluids since lubricant recovery is facilitated. It is therefore important that the selection of the clearance value should balance both the formation of the EHL film and the fluid recovery based on the physiological viscosity in the human body to create an optimal tribological performance of MOM hip joints.

Ó Authors 2011

9

10

11

12

13

REFERENCES 14 1 Chan, F. W., Bobyn, J. D., Medley, J. B., Krygier, J. J., and Tanzer, M. Wear and lubrication of metalon-metal hip implants. Clin. Orthop. Rel. Res., 1999, 369, 10–24. 2 Hu, X. Q., Isaac, G. H., and Fisher, J. Changes in the contact area during the bedding-in wear of

15

1067

different sizes of metal on metal hip prostheses. Bio-Med. Mater. Engng, 2004, 14, 145–149. Rieker, C. B., Scho¨na, R., and Ko¨ttig, P. Development and validation of a second-generation metalon-metal bearing. J. Arthroplasty, 2004, 19(8 Suppl. 3), 5–11. Kang, L., Jin, Z. M., Isaac, G., and Fisher, J. Long term wear modelling of metal-on-metal hip resurfacing prosthesis: effect of clearance. The 52nd Annual Meeting of the Orthopaedic Research Society, Chicago, Illinois, 2006, poster no. 0501. Scholes, S. C., Green, S. M., and Unsworth, A. The wear of metal-on-metal total hip prostheses measured in a hip simulator. Proc. IMechE, Part H: J. Engng Med., 2001, 215, 523–530. Dowson, D., Hardaker, C., Flett, M., and Isaac, G. H. A hip joint simulator study of the performance of metal-on-metal joints. Part II: design. J. Arthroplasty, 2004, 19(8 Suppl. 3), 124–130. Rieker, C. B., Scho¨na, R., Konrada, R., Liebentritta, G., Gnepfa, P., Shena, M., Roberts, P., and Grigoris, P. Influence of the clearance on invitro tribology of large diameter metal-on-metal articulations pertaining to resurfacing hip implants. Orthop. Clin. Nor. Am., 2005, 36(2), 135– 142. Dowson, D. Tribological principles in metal-onmetal hip joint design. Proc. IMechE, Part H: J. Engng Med., 2006, 220, 161–171. Isaac, G., Hardaker, C., Hadley, M., Dowson, D., Williams, S., and Fisher, J. Effect of clearance on wear and metal ion levels in large diameter metalon-metal bearings in a hip joint simulator study. The 53rd Annual Meeting of the Orthopaedic Research Society, San Diego, California, 2007, poster no. 1779. Smith, S. L., Goldsmith, A. A. J., and Dowson, D. The effect of diametral clearance, motion and loading cycles upon lubrication of metal-on-metal total hip replacements. Proc. IMechE, Part C: J. Mech. Engng Sci., 2001, 215, 1–5. Brockett, C., Williams, S., Isaac, G., Flett, M., Jin, Z., and Fisher, J. Effect of clearance on friction and squeaking in large diameter metal-on-metal bearings. The 53rd Annual Meeting of the Orthopaedic Research Society, San Diego, California, 2007, poster no. 1655. Jin, Z. M., Dowson, D., and Fisher, J. Analysis of fluid film lubrication in artificial hip joint replacements with surfaces of high elastic modulus. Proc. IMechE, Part H: J. Engng Med., 1997, 211, 247–256. Jagatia, M. and Jin, Z. M. Analysis of elastohydrodynamic lubrication in a novel metal-on-metal hip joint replacement. Proc. IMechE, Part H: J. Engng Med., 2002, 216, 185–193. Udofia, I. J. and Jin, Z. M. Elastohydrodynamic lubrication analysis of metal-on-metal hip-resurfacing prostheses. J. Biomech., 2003, 36, 537–544. Besong, A. A., Lee, R., Farrar, R., and Jin, Z. M. Contact mechanics of a novel metal-on-metal total hip replacement. Proc. IMechE, Part H: J. Engng Med., 2001, 215, 543–548. Proc. IMechE Vol. 225 Part H: J. Engineering in Medicine

1068


16 Mabuchi, K., Sakai, R., Ota, M., and Ujihira, M. Appropriate radial clearance of ceramic-on-ceramic total hip prostheses to realize squeeze-film lubrication. Clin. Biomech., 2004, 19, 362–369. 17 Yew, A., Udofia, I. J., Jagatia, M., and Jin, Z. M. Analysis of elastohydrodynamic lubrication in McKee–Farrar metal-on-metal hip joint replacement. Proc. IMechE, Part H: J. Engng Med., 2004, 218, 27–34. 18 Chan, F. W., Medley, J. B., Bobyn, J. D., and Krygier, J. J. Numerical analysis of time-varying fluid film thickness in metal-metal hip implants in simulator tests. In Alternative bearing surfaces in total joint replacement (Eds J. J. Jacobs, and T. L. Craig), 1998, pp. 111–125 (American Society for Testing and Materials, West Conshohocken, Pennsylvania). 19 Dowson, D., Wang, F. C., Wang, W. Z., and Jin, Z. M. A predictive analysis of long-term friction and wear characteristics of metal-on-metal total hip replacements. Proc. IMechE, Part J: J. Engng Tribol., 2007, 221, 367–378. 20 McKellop, H., Park, S., and Chiesa, R. In vivo wear of 3 types of metal on metal hip prosthesis during 2 decades of use. Clin. Orthop. Rel. Res., 1996, 329S, S128–S140. 21 Reinisch, G., Judmann, K. P., Lhotka, C., ¨ller, K. A. Retrieval study Lintner, F., and Zweymu of uncemented metal–metal hip prostheses revised for early loosening. Biomater., 2003, 24, 1081–1091. ¨ ttig, P. In vivo tribological per22 Rieker, C. B. and Ko formance of 231 metal-on-metal hip articulations. Hip Int., 2002, 12, 73–76. 23 Long, W. T. The clinical performance of metal-onmetal as an articulation surface in total hip replacement. Iowa Orthop. J., 2005, 25, 10–16. 24 Tuke, M. A., Scott, G., Roques, A., Hu, X. Q., and Taylor, A. Design considerations and life prediction of metal on metal bearings: the effect of clearance. J. Bone Joint Surg. Am., 2008, 90(Suppl. 3), 134–141. 25 Rainer, F. and Ribitsch, V. Viscoelastic properties of normal human synovia and their relation to biomechanics, Z. Rheum., 1985, 44, 114–119. 26 Schurz, J. and Ribitsch, V. Rheology of synovial fluid. Biorheology, 1987, 24, 385–399. 27 Conrad, B. The effects of glucosamine and chondroitin on the viscosity of synovial fluid in patients with osteoarthritis. MEng Thesis, Department of Biomedical Engineering, University of Florida, 2001. 28 Scholes, S. C. and Unsworth, A. Comparison of friction and lubrication of different hip prostheses. Proc. IMechE, Part H: J. Engng Med., 2000, 214, 49– 57. 29 Mazzucco, D., McKinley, G., Scott, R. D., and Spector, M. Rheology of joint fluid in total knee arthroplasty patients. J. Orthop. Res., 2002, 20, 1157–1163. 30 Bishop, N. E., Waldow, F., and Morlock, M. M. Friction moments of large metal-on-metal hip joint bearings and other modern designs. Med. Engng Phys., 2008, 30(8), 1057–1064. Proc. IMechE Vol. 225 Part H: J. Engineering in Medicine

31 Banchet, V., Fridricia, V., Abrya, J. C., and Kapsaa, P. Wear and friction characterization of materials for hip prosthesis. Wear, 2007, 263, 1066– 1071. 32 Gisperta, M. P., Serroa, A. P., Colacxoc, R., and Saramago, B. Friction and wear mechanisms in hip prosthesis: comparison of joint materials behaviour in several lubricants. Wear, 2006, 260, 149– 158. 33 Yan, Y., Neville, A., Dowson, D., Williams, S., and Fisher, J. Effect of metallic nanoparticles on the biotribocorrosion behaviour of metal-on-metal hip prostheses. Wear, 2009, 267, 683–688. 34 Affatato, S., Spinelli, M., Zavalloni, M., MazzegaFabbro, C., and Viceconti, M. Tribology and total hip joint replacement: current concepts in mechanical simulation. Med. Engng Phys., 2008, 30, 1305–1317. 35 Morillo, C., Sawae, Y., and Murakami, T. Effect of bovine serum constituents on the surface of the tribological pair alumina/alumina nanocomposites for total hip replacement. Tribol. Int., 2010, 43, 1158–1162. 36 Mavrakia, A. and Cann, P. M. Lubricating film thickness measurements with bovine serum. Tribol. Int., 2011, 44, 550–556. 37 Ortega-Sa´enz, J. A., Herna´ndez-Rodrı´guez, M. A. L., Pe´rez-Unzueta, A., and Mercado-Solis, R. Development of a hip wear simulation rig including micro-separation. Wear, 2007, 263, 1527–1532. 38 Fama, H., Kontopouloua, M., and Bryant, J. T. Method for friction estimation in reciprocating wear tests. Wear, 2011 271, 999–1003. 39 Scholes, S. C., Unsworth, A., and Goldsmith, A. A. J. A frictional study of total hip joint replacements. Phys. Med. Biol., 2000, 45, 3721–3735. 40 Scholes, S. C., Unsworth, A., Hall, R. M., and Scott, R. The effects of material combination and lubricant on the friction of total hip prostheses. Wear, 2000, 241, 209–213. 41 Cooke, A. F., Dowson, D., and Wright, V. The rheology of synovial fluid and some potential synthetic lubricants for degenerate synovial joints. J. Engng Med., 1978, 7, 66–72. 42 Paul, J. P. Loading on head of human femur. J. Anat., 1969, 105, 187–188. 43 Medley, J. B., Chan, F. W., Krygier, J. J., and Bobyn, J. D. Comparison of alloys and designs in a hip simulator study of metal on metal implants. Clin. Orthop. Rel. Res., 1996, 329S, S148–S159. 44 Williams, S., Stewart, T. D., Ingham, E., Stone, M. H., and Fisher, J. Metal-on-metal bearing wear with different swing phase loads. J. Biomed. Mater. Res., Part B: Appl. Biomater., 2004, 70, 233–239. 45 Brockett, C., Isaac, G., Jin, Z. M., Williams, S., and Fisher, J. Influence of bearing couple, lubricant and swing phase load conditions on the friction of 28 mm hip replacements. The 52nd Annual Meeting of the Orthopaedic Research Society, Chicago, Illinois, 2006, poster no. 0494. 46 Clarke, M. T., Lee, P. T. H., and Villar, R. N. Dislocation after total hip replacement in relation to


metal-on-metal bearing surface. J. Bone. Joint. Surg. Br., 2003, 85, 650–654. 47 Tuke, M., Hu, X. Q., and Taylor, A. Suction force of metal on metal hip joints with different clearances and viscosities. J. Bone. Joint. Surg. Br., 2010, 92(Suppl. 4), 613.

Tb Tf g h300

1069

frictional torque when head frame moves backward frictional torque when head frame moves forward shear rate fluid viscosity at 300 s–1 shear rate

APPENDIX Notation Ra T

arithmetical mean roughness frictional torque

Proc. IMechE Vol. 225 Part H: J. Engineering in Medicine