Wear studies on prosthetic materials using the pin-on-disc machine

Wear studies on prosthetic materials using the pin-on-discmachine K W.J. Wright,H. S. Dobbsand J. T. Scales Department of Biomedical Engineering, Hospital, Stanmore, Middx., UK (Received31

March

1981; revised 1 July

Institute

of Orthopaedics

(University

of London/,

Royal National

Orthopaedic

198lJ

The wear of ultra high molecular weight polyethylene in combination with cobalt-chromium-molybdenum alloy was investigated by pin-on-disc experiments in which the alloy pin was loaded against the rotating polyethylene disc. In some experiments the pin was stationary, but in others it rotated about its axis. The effect of lubricant type, pin rotation, magnitude of the applied load, magnitude of the relative surface velocity, and disc storage in various environments was investigated. Wear of the polyethylene was assessed at 100 h intervals by both volumetric and gravimetric methods. Each experiment lasted 1000 h. It was concluded that of the lubricants tested bovine serum was the most suitable and that pin rotation produced more suitable polyethylene surfaces thn did the stationary pin. The wear rate increased slightly with load (possibly owing to creep), was insensitive to speed, and was not affected by disc storage. The dynamic coefficient of friction increased with running time. Results obtained by the two wear assessmentmethods were comparable and the reproducibility of the results was good. Keywords:

wear tests, polyethylene,

friction,

total joint

replacement,

It is necessary to assessthe friction and wear characteristics of material combinations that are considered potential candidates for use in total joint replacement prostheses. As discussed elsewhere’,’ the hip joint is subject to a complex system of loading, movement and environmental conditions. Many attempts have been made to simulate these conditions in equipment which would enable the friction and wear characteristics of actual joint prostheses to be studied in the laboratory. However, the necessary resources and capacity to examine all the various material combinations which may be useful do not exist. Therefore research workers have adopted various simplified methods to screen or rank material combinations such that only the most promising combinations undergo the more expensive and rigorous examinations provided by the simulation technique. The pin-on-disc method described in this paper is one of many techniques which have been employed for screening potential candidate materials. Originally developed by British Petroleum, the method has been further developed at Stanmore over many years. While obviously not providing working conditions which exactly simulate those found in the body, the method incorporates features which, with present day knowledge, are considered to be appropriate. These are: (i) the materials are exposed to a physiological environment; (ii) the polymeric component is subjected to cyclical stressing; (iii) the pin-on-disc bearing configuration resembles that used in total joint replacements in Go; and (iv) the pin and disc specimens

are of simple form

of a size which can be implanted

and

into an animal.

0 1982 Butterworth & Co. (Publishers) Ltd. 0142-9612/82/010041~8

UHMWE.

The object of the present work was to develop a reliable, simple, non-clinical method of examining the wear and frictional characteristics of prosthetic material combinations using a pin-on-disc machine. Using cast cobaltchromium-molybdenum (CoCrMo) pins and ultra high molecular weight polyethylene discs the following aspects were investigated: 1) the effect of the type of lubricant used; 2) the effect of the type of motion of the bearing surfaces; 3) the effect of a variation in the magnitude of the applied load; 4) the effect of a variation in the relative velocity of the surfaces; and 5) the effect of storage of the discs in various environments. The worn surfaces of the pins and discs were studied and compared with the worn surfaces of total hip prostheses removed from patients. Volumetric and gravimetric methods of wear assessment were compared.

MATERIALS

AND METHODS

Pin specimens The pins employed in the experiments were heat treated cast CoCrMo alloy complying with the relevant British Standard (B.S. 3531 Part 1:1968). The microstructure of the heat treated alloy exhibited carbides and was characteristic of partially solution treated material. Only one end of each pin was used in the experiments and that end was machined to a 200 mm spherical radius and polished to a surface finish better than 0.05 microns C.L.A. The pins were autoclaved prior to use.

0 isc specimens The polyethylene

employed

in these experiments

was

$03.00 Biomaterials

1982,

Vol3

January

4 1

Wear of UHMWPE: K.W.J. Wright et al.

medical grade RCH 1000, supplied by Ruhrchemie A.G., Oberhausen-Holten, West Germany. The material was supplied in four blocks each measuring 55 mm x 55 mm x 760 mm. These blocks were derived by the manufacturer from a single master block measuring 800 mm x 400 mm x 80 mm by removing the top, bottom and 4 edges. Disc specimens were prepared from the 4 blocks. Only one surface of each disc was used in the experiments and this was turned to a finish of better than 1.0 micron C.L.A. Each disc was triple packed in polyethylene and gamma irradiated to 2.5 Mrads. The specimens were randomly selected for use in the experiments. The properties of the material were characterized as indicated in the accompanying article13. The density was found to be 942 + 1 kgmm3. It varied slightly with position in the block and it increased slightly (less than 1%) after irradiation. The material was found to be impurity-free (less than 200 ppm), but contained trace elements of calcium and aluminium, and a small quantity of silicon as mentioned in an earlier report3. The material had a high molecular weight and a low porosity. The hardness (DIN 53456) was 36 MNm-*. Test equipment and experimental procedure The Stanmore Mk 2 pin-on-disc machine was employed for all of the experiments, see Figure 7. With this equipment the RCH 1000 disc could be rotated in a lubricant bath at a constant predetermined speed. The cast CoCrMo pin specimen was held in a spindle assembly which was mounted in a combined linear and radial bearing in a support arm. The spindle assembly could also be rotated at a constant predetermined speed. Load was applied to the pin-on-disc bearing by applying weights to the spindle shown in the figure. In order to measure the frictional resistance of the pin-on-disc bearing, a special ‘friction’ pin-ondisc machine was constructed. This machine incorporated the same pin and disc mounting assemblies as the non-friction measuring machine and a general view of both is given in Figure 2. With the friction measuring machine a strain gauged cantilever beam was employed to determine the frictional force generated between the pin and disc specimens. From this information the dynamic coefficient of friction could be derived. In addition, the static coefficient of friction could be obtained by adding lead shot to a weigh pan attached to the support arm. The static coefficient of friction was derived by dividing the minimum frictional effort required to cause the pin to slide over the stationary disc surface

Figure 2

General view of the MK 2 pin-on-disc machine

by the load applied to the pin spindle. The machine and method will be more fully described elsewhere. The duration of each experiment was 1000 h, but at 100 hour intervals the experiments were stopped for quantitative and qualitative wear assessment. The discs were removed from the equipment, washed in water and dried on a paper towel. Similarly, the spindle assemblies were removed from the equipment and the ends of the pins washed in water and dried on a paper towel. The pins were not removed from their respective spindle assemblies. After wear assessment and component examination the specimens were again washed in water and dried on paper towels prior to their reassembly into the pinondisc equipment. After each assessment period new lubricant was employed. Quantitative wear assessments Quantitative wear of the discs was assessedusing two methods. In the first method Talysurf traces were taken across the wear tracks on the discs at four separate stations at 90” to each other. From the trace records obtained, the volumetric wear was computed using a Simpson’s rule technique. The traces were taken immediately after the discs were removed from the equipment. In the second method, each disc under test was weighed prior to testing and was then reweighed after each 100 h of the experiment. Each disc under test was matched with a similar disc as a control. The control disc was necessary to compensate for the increase in the weight of the test disc due to liquid absorption from the lubricant or from the air. Each control disc was therefore weighed prior to testing and was then reweighed at the same time as the test disc after storage in the appropriate lubricant (or in air). A titanium spindle was made to enable the control disc to be fully immersed in the lubricant. The weight measurements were made with a Mettler analytical balance (Type H IO) with precision and readability quoted as f 0.05 mg and 0.1 mg respectively. In practice successive measurements were reproducible to within 0.2 mg. Qualitative wear assessments

Figure 1

42

Schematic diagram of the MK2 pin-ondisc machine

Biomaterials 1982, Vol3 January

At each 100 h interval the pins and discs were photographed using light microscopy. The pins were photographed at low magnification to enable the size of the contact area to be determined. Details on both surfaces were examined and photographed at magnification up to 500x. On some discs Bex film replicas (using acetone as a solvent) were taken across the wear track. These were gold-palladium coated and examined in a scanning electron microscope.

Wear of UHMWPE:

K.W.J.

Wrighr et al.

RESULTS The first and second experiments studied the effects of different lubricants and of different motions of the bearing surfaces under the following standard test conditions: Pin Load Disc Speed The different

10kgf 38 revs. min-’

lubricants investigated were:

a) no lubricant - components run dry; b) bovine serum plus 1% of sodium azide solution*; Cf 0.5% gelatine in distilled water plus 1% of sodium azide solution*; and

d) 0.9% injection B.P. sodium chloride solution plus 1% of sodium azide solution”. In the first experiment the pins were held stationary while in the second the pins were rotated at a constant speed of 1 rev. min-‘. Because the pitch circle diameter of the wear track was 25 mm, the mean relative sliding velocity of the bearing in the former case was approximatefy 3 m. min-‘. For the latter case, however, although the mean relative sliding velocity of the bearing was still approximately 3m. min-‘, with the 8 mm diameter pin specimen contrarotating at 1 rev. min-‘, the relative sliding velocity distribution was modified as shown in figure 3. Typical Talysurf traces taken of the wear tracks and employed in the volumetric wear evaluation are shown in Figure 4. Typical volumetric wear vs time curves are plotted

Relative

Sliding Velocity of Pin on Disc Bearing

TEST DURAll”b,iio”rS1

Figure 5 Volumetric in Figure 4

wear vs time curves obtained

from the data

Figure 6 Gravimetric wear vs time curves obtained experiment as in Figure 5

for the same

in Figure 5 and typical gravimetric wear vs time curves are shown in Figure 6. The third and fourth experiments studied the effects of different pin loads and disc speeds under the following standard test conditions: Lubricant Pin Speed Pin Stationary

Pin Rotating

1rev.min-’

Figure 3 Relative sliding velocity distribu Cons across the 8mm diameter pin for the two motions studied

-

Serum -+ 1% sodium azide solution* 1 rev.min-’

The combinations of disc speed and pin load examined were 19,38 and 76 revs. min-’ and 10, 15 and 20 kgf respectively. The fifth experiment studied the effects of storage of the RCH 1000 discs under the following standard test conditions: Lubricant Pin Speed Disc Speed Pin Load

- Serum-t 1% sodium azide solution* - 1 rev. min-’ - 38 revs. min-’ - IOkgf

The conditions examined were: a) storage in the dark on a shelf for 1 year; b) storage in a goat for 1 year; and c) storage in serum in an incubator at 37°C for 1 year. Because all the wear vs time graphs exhibited an approximately linear relationship, a least squares linear regression analysis was performed on the data obtained from each run to determine the slope and the intercept of the linear regression line. For each experiment the slopes obtained for each of the two runs were averaged and the Figure 4 Tatysurf wear track profiles of discs used in the experiment with the pin stationary (Expt. la 1 Run 2).

*Sodium azide solution growth inhibitor

8% concentration:

Biomaterials

used as a bacteriological

1982,

Vol3

January

43

Wear of UHMWPE:

Table 1 Expt.

Average

No.

K.W.J.

Wright et al.

volumetric

and gravimetric

Run No.

Head Nos.

wear rates per cycle for the series of experiments Lub.

Load (kgf)

Disc speed (revs min-1)

Volumetric wear rate

Storage

mm3X

Gravimetric wear rate*

10-E

mgXlO+

cycle la1

l&l 2&2

l&2

3&3 4&4 l&2

la2

gel Na Cl

10 10 10

dry serum

serum

none

10

38 38 38 38

l&l 2842 3&3 4&4

drv

10 10 10 10

gel Na Cl

cycle

none none

5.68 0.43 3.52 2.93

3.36 -0.14 2.48 4.29

38 38 38 38

none none none none

1.57 1 .I0 0.52 6.06

0.73 -0.05 -0.40 4.45

la3

2

l&3 2&4

serum serum

15 20

38 38

none

1 .90 2.36

2.29 0.47

la4

1

l&3 2&4

serum serum

10 20

76 76

none none

2.38 2.81

1.70 1 .48

la4

2

l&3 l&4

serum serum

10 20

19 19

none none

2.46 3.10

0.97 1 .I 1

la5

1

l&3 2&4

la5

2

1-4

serum serum serum

10 10 10

38 38 38

shelf incub goat

1.37 1.68 1.73

0.41 0.66 0.57

*A negative value indicates a weight gain.

average slope was divided by the appropriate number of cycles to give the average wear rate per cycle as shown for both the volumetric and the gravimetric assessments in Tab/e 7. For each run the value of the wear after 1000 h was obtained from the slope and the intercept and is given in Figures 7- 10. A separate experiment was performed using the ‘friction’ pin-on-disc machine to study the friction between the cast CoCrMo pin and RCH 1000 disc. Before commencement of the experiment the static coefficient of friction was determined for this material combination under both dry and serum lubricated conditions. The frictional effort required to just slide the pin over the disc surface for various pin loads was plotted for both conditions. The static coefficient of friction was obtained from the slope of the graph and the values are given in Tab/e 2, During and without interrupting the experiment, the dynamic coefficient of friction was determined by measuring the frictional resistance to motion developed between the bearing surfaces. The resulting dynamic coefficient of friction vs time curve is plotted in Figure 7 1. Some values are also given in Table 2.

Talysurf instrument. However, as the Simpson’s rule technique was carried out manually, one disadvantage of the method was that it was both time consuming and labour intensive. A further disadvantage was that any cold flow and/or distortion of the discs would be included in the amount of wear determined. This latter aspect, however, is now under investigation and preliminary results suggest that little creep, probably less than IO%, was associated with the wear tracks. The gravimetric method of wear assessment was simple to carry out and the results compared reasonably

Imm 31

KEY

10 0 1c

10 8 c L 2

Method of wear assessment

0

44

GELATIN.!

I NaCl

DRY

Gravlmetrlc Assessment

for

Dynamic Serum lubricated

Serum lubricated

Initial value 0.06

-

of friction

Static Dry

L

SERUM

The volumetric method of wear assessment has been employed satisfactorily at Stanmore for many years, The sensitivity of the method can be varied to suit the amount of wear by selecting the appropriate magnification on the

Coefficients

I

Rotattng Pmlrevmln”

12

DISCUSSION

Table 2 Summary of the values of the coefficient of friction the cast CoCrMo/RCH 1000 pin+n-disc material combination

Statlanary Pm

0.084

0.11

Biomaterials

1982,

Vol3

Minimum value after 15 h 0.07

January

Final value after 1050 h 0.175

-2 Figure 7 motions

SERUM Disc wear after

GELATINE 1000h:

NaCl

effect of different

DRY lubricants

and

Wear of UHMWPf:

Volumetric Assessment imm’l

1L

1Okgf

1

12 i

‘1

101 B6

15kgf

1 2

-

0 -L-_-d 19 38

36

75 Growmetnc

19

38

75

lrevmln’l

19

38

76

(rev

Assessment

Cmgl 1Okgf

19 Figure 8

38

38

76

Disc wear after

rnIri1

1000 h: effect of speed and load

Volumetric

K.W.J.

Wrightet

al.

the assessment of materials such as ceramics, which wear even less than high density polyethylene. In theory, these disadvantages are not insurmountable and the difficulty with the controls could be overcome by presoaking the discs4, and greater accuracy could be achieved using a more sensitive balance. Other possible errors5 are the adherence to the disc surface of worn particles and wear caused by insertion and removal of the disc from the machine. Both these errors are expected to be small. In metal pin/polymer disc combinations wear is generally confined to the polymeric component. However, for material combinations such as metal/metal or ceramic/ ceramic, which produce pin wear, this can be determined by employing the Talysurf technique. With the present design of pin-on-disc equipment the pins cannot be removed from their collets and accurately reinserted during an experiment. Consequently the gravimetric method cannot be employed to assesspin wear. Using either method of assessment there was good reproducibility between runs. Comparing, for example, the 1000 h values (Figures 7-70) it can be seen that in all cases except one the difference between the values obtained in the two runs was less than 20% of the mean of the values. In the one case where this was not so (Expt. 1 Runs 1 and 2, NaCl solution) the large value obtained on the first run was

Assessment 76rev mm

I

I mm’)

“1

- n

n

lncub

Shelf

None

Goat

-

OJ10 Grovlmetrlc

15

2G

10

20

lkgfi

Assessment 7trev

Grovwnetrlc

Assessment

Img)

nvn-

L -

I mgl

0

2

6

] ‘ 1

0None

2

Figure

10

Disc wear after

Shelf

lncub

Goat

10OOh: effect of disc storage

0 ! 10 Figure 9

2c

Disc wear after

1c 1000h:

15

?C

1G

20

lkgfl

effect of speed and load

well with those determined by the volumetric method as shown for example in Tab/e 7. The method, however, had two important disadvantages: 1) the fact that in some cases a weight gain was recorded and 2) the fact that the minimum weight change that could be recorded was 0.2 mg. The first disadvantage presumably resulted from the fact that the use of a control disc provided only partially adequate compensation for liquid absorption. Clearly adequate compensation is difficult to achieve, since each disc weighed approximately 12 gm and after 1000 hours gained 0.5 mg in air, or 2 mg in a lubricant, and lost less than 10 mg due to wear. The second disadvantage meant that only large weight changes could be measured accurately and consequently the method as it stood would be unsuitable for

TEST

D”RAiiDN

Figure 11 Dynamic coefficient of friction CoCrMo/RCH 1000 material combination

Biomaterials

vs time curve for cast

1982,

Vol3

January

45

Wear of UHMWPE:

K. W.J.

Wrightet al.

thought to have been the result of a prodigious NaCl build-up on the pin. Reproducibility is an important factor in determining the suitability of a test method, and the good reproducibility obtained in the present experiments was an encouraging feature.

Relative motion of the bearing surfaces When the pins were held stationary, the metal pin bearing surface was scored and exhibited a transfer film, as shown in Figure 72. This appearance is uncharacteristic of the metal femoral head of prostheses removed from patients. In addition, the wear tracks in the RCH 1000 discs exhibited rough profiles as shown by the Talysurf traces in Figure 4. With the rotating pin, however, the pin surface remained smooth as shown in Figure 73. The wear track profiles were also smooth and the appearance of the wear tracks as shown in Figure 74 was more characteristic of the worn surfaces of RCH 1000 acetabular components removed from patients as shown in Figure 746. The scoring of the metal pin and the excessive transfer may, in part, be responsible for the larger scatter obtained in the wear results for the stationary pin experiments.

Effect of lubricant

type

Referring to Figure 7 it can be seen that the 1000 h wear value was considerably affected by lubricant type. Considering only the rotating pin experiments, since the stationary pin experiments gave uncharacteristic wear as discussed above, it can be seen that the lowest values were obtained

Figure 13 Appearance and profile of rotating pin after wear test in bovine serum (Expt. la2 Run 1)

Figure 12 Appearance andprofile of stationary test in bovine serum IExpt. la 1 Run 2)

46

Biomaterials

1982,

Vol3

January

pin after

lOOOh

lOOOh

with NaCl and gelatine, the highest value without lubricant (dry), and intermediate values with bovine serum. Both NaCl and gelatine produced a deposit on the disc surface, whereas the absence of a lubricant produced evidence of melting as shown in Figure 75. The fact that the lubricant acted as a coolant was demonstrated in other experiments in which the pin temperature, measured at a distance of 5 mm from the bearing surface, was found to be 45°C greater in the absence of a lubricant. This result is consistent with the observed evidence of melting. The disc in bovine serum exhibited neither a deposit nor melting. The appearance of a rotating pin after a 1000 h wear test in bovine serum was shown in Figure 73. A transfer film can be seen to have formed and a similar transfer film was found to occur on the pins in NaCI, in gelatine and dry. The transfer films in gelatine appeared to be the thinnest and least abrasive. Since a suitable lubricant for future experiments was considered to be (a) one which closely resembled that thought to occur in vivo, and (b) one which produced metal and polymer surfaces worn in a manner characteristic of surfaces worn in vim, it was decided that bovine serum was the preferred lubricant of those investigated. This conclusion is supported by previously published work. Duff-Barclay and Spillman’ recognised the importance of protein complexes as boundary lubricants in vivo, while in vitro Charnley7 obtained more realistic results on PTFE using protein-containing lubricants, Dumbleton et al.8 used blood plasma, and McKellop et al.’ used serum as the lubricant of choice.

Wear of UHMWPE:

K. W.J. Wright et al.

Effect of load Referring to Figure 8, it can be seen that the volumetric assessment demonstrated for the three speeds tested a slight increase in wear volume with increasing load. This effect was not found with the gravimetric assessment, which demonstrated a random behaviour as a function of load. The increase observed with the volumetric assessment may be an artefact caused by polymer creep as discussed above. Whether this was the case or not, an increase in the applied load had a surprisingly small effect on the wear rate. A different result was obtained by McKellop et a/. In their experiment, in which a polyethylene pin was tested against a stainless steel counterface, the wear rate doubled when the load doubled. Their mean contact stress, however, did not exceed 6.9 MPa whereas our stresseswere in the range 7.8 MPa (IO kgf) to 15.6 MPa (20 kgf). This difference may account for the different behaviours.

Figure 15 Optical micrograph of wear track in polyethylene after lOOOh wear test with rotating pin without a lubricant (x361. fExpt. la2 Run 1)

disc

Effect of speed Figure 9 appears to indicate an increase in both volumetric and gravimetric wear with increasing disc speed. This effect, however, is the result of the fact that at the three speeds tested, 19,38 and 76 r.p.m., the distance travelled is respectively 1, 2 and 4 times greater than at the lowest speed. Taking this into account, the wear rate per cycle as shown in Table 1 can be seen to be little affected by speed. Thus it can be concluded that, within the range of speeds tested, speed was not a significant factor affecting the wear rate, whereas the number of cycles or the distance travelled was.

Effect of storage Referring to Figure 70, it can be seen that similar wear rates were obtained for the four methods of storage investigated. It may be concluded that (a) storage for a year whether in a goat or the laboratory, has little effect on the wear rate as compared with material ‘as received’ and (b) storage in a goat was not significantly more severe than on the shelf or in an incubator. These conclusions are supported by the work of Amstutz and Ludwig” who implanted UHMW polyethylene in dogs for up to three years and observed no increase in wear rate after implantation.

The wear equation Previous workers have used the following wear equation originally proposed by Archard” for metal contacts: W = K.P.X. where W is the amount of wear, P is the applied load, X is distance travelled and K is the wear coefficient. Since our experiments demonstrated that the wear volume or weight was relatively insensitive to the load or the speed, but depended on the number of cycles, we concluded that within the range of loads and speeds tested using our method the following wear equation is applicable: W = k.X.

Figure (stored lb) the patient

14 SEM micrograph of (al wear track on RCH 1000 disc in goat for 1 year) after lOOOh wear test in bovine serum; surface of a worn acetabular component removed from a after a 26 month implantation period (x42.51

Values of k, the wear rate per cycle, were given in Table 1. Our wear rates can be compared with published data. Brown et al.” reported the results of pin-ondisc experiments in which the pin was polyethylene, the disc was metal, no lubricant was used and the range of loads was from 25-l 25 N.

Biomaterials

1982,

Vol3

January

47

Wear of UHMWPE: K.W.J. Wright et al.

They obtained a wear rate of 2.2~ IO-’ mm3/Nm. Although in their experiments the polymer was continuously loaded, whereas in ours it was not, their value can be compared with ours provided an estimate can be made of the distance slid per cycle for the polymer disc in our experiments. Two possible estimates may be considered: 1) the contact area diameter (5 mm), namely the distance slid by the pin over each element of the disc, and 2) the circumference of the wear track (78 mm), namely the total distance slid by the pin. Assuming a volumetric wear rate of 2 x 10m6mm3/cycle and a load of 100 N (typical of the values shown in the table), the former estimate gave a wear rate of 4 x 10m6mm3/Nm and the latter a value of 2.6 x 10c7mm3/Nm. The value obtained by Brown et al.‘2 lies between these values, but because of the difference in test method an entirely satisfactory comparison cannot be made. McKellop eta/.g compared published wear data for polyethylene on the basis of the depth worn per year. This quantity they determined by multiplying the calculated wear rate (wear per unit load per unit sliding distance) by a typical contact stress (3.45 MPa) and by a typical yearly sliding distance (50 km). From thirteen reports including their own they obtained values from 1 to 1600pm/year. Our values of 45 and 690/..rm/year (derived from the values given above) clearly fall within this range. These values can also be compared with the average value of in viva cup wear measured by Charnley’ from X-rays, namely 150/_rm/year.

Coefficients of friction It was interesting to observe that the static coefficient of friction for the serum lubricated case was higher than the value obtained for the dry case. The theory which has been proposed for this phenomenon is that in the dry situation only contacting asperities have to be sheared, whereas in the lubricated case both the asperities and any intervening film have to be sheared. The initial dynamic coefficient of friction (serum lubricated) was less than the static value for the serum lubricated case and was similar to the dry static value (see Tab/e 2). However, during the course of the experiment the dynamic value increased. The main reason for this increase was probably the build-up of the transfer film on the metal pin surface. The dynamic coefficient of friction was measured both immediately before and immediately after each wear assessment. There appeared to be no consistent variation in the coefficient which could be attributed to either the interruption of the experiment or the use of fresh lubricant. An increase in friction was also reported by McKellop et al.’ but their value rose as the transfer film increased in thickness and fell as it was removed.

CONCLUSIONS The pin-on-disc method has been further developed by the incorporation of pin rotation. This development produced worn surfaces which were more similar to those produced in vivo. However, only one material combination (cast CoCrMo/RCH 1000) has been examined and further experiments are required to determine the applicability of the method to other material combinations. From the experiments it was concluded that: 1) the use of a lubricant was essential; 2) boyine serum was the most suitable of the lubricants investigated;

48

Biomaterials 1982, Vol3 January

3) there was only a small increase in the wear rate as a result of a two fold increase in load; 4) there was no effect on the wear rate over the range of speeds studied; 5) there was no appreciable effect on the wear rate following storage for 1 year either in the laboratory or in the animal; 6) the amount of wear was proportional to the distance travelled. The absence of an appreciable effect due to the variations in load and speed was surprising. The absence of an effect due to storage was perhaps not surprising in view of the established clinical acceptability of RCH 1000. The dynamic coefficient of friction at the start of the experiment was low but increased with time due to the deterioration of the bearing surfaces.

ACKNOWLEDGEMENTS We would like to thank our colleagues in the Department of Biomedical Engineering, particularly Mr. J.D. Wood for helping with the running of the experiments and Mr. J. Meswania for manufacture of the equipment and the specimens. We are grateful to the Institute Librarians and the staff of the Medical Photographic Department. Finally we are grateful to the Department of Health and Social Security for the provision of a research grant (R/E1049/ 39STSB5) to support this work.

REFERENCES 1

2

3

4 5

6

7

8

9

Wright, K.W.J. and Scales, J.T., Stanmore hip joint simulators for study of total hip prostheses, Evaluation of Artificial Joints, (Eds. V. Wright and D. Dowsonj, Biolog. Eng. Sot. 1977, Ch. 1, pp. 9-l 7 Wright, K.W.J. and Scales, J.T., The use of hip joint simultators for the evaluation of wear of total hip prostheses. In Evaluation of Biomaterials, (Eds. G.D. Winter, J.L. Leray and K. de Groat), Wiley, 1980, pp.135146 Dobbs, H.S., Scales, J.T., Wright, K.W.J. and Braun, G., Silicon containing particles in ultra high molecular weight polythene. J. Bioengng, 1977,1, (2) 113-I 14 Clarke, l.c., Wear of artificial joint materials, 1 Friction and wear studies, Engng. Mad. 1981,10, (31.115-122 Seedhom, 8.8.. Dowson, D. and Wright, V., Wear of solid phase formed high density polyethylene in relation to the life of artificial hips and knees, Wear, 1973,24,3551 Duff-Barclay, I. and Spillman, D.,Total human hip joint prostheses: a laboratory study of friction and wear, Proc. Inst. Mech. Eng. 1967.181, Part 3J. 90-103 Charnley, J., The wear of plastics materials in the hip joint. In Plastics in Medicine and Surgery, The Plastics and Rubber Institute, London: 1975, p. 3.1 Dumbleton, J.H., Shen, C. and Miller, E.H., A study of the wear of some materials in connection with total hip replacement, Wear 1974,29,163-171 McKellop, H., Clarke, l.c., Markoff, K.L. and Amstutz, H.C., Wear characteristics

of U.H.M.W.

polyethylene.

J. Biomed.

Mater. Res. 1978,12,895-927 10

11

12

13

Amstutz, H.C. and Ludwig, M., Wear of polymeric bearing materials: the effect of in viva implantation. J. Biomed. Mater. Res, 1976, lo,2531 Archard, J.F., Contact and rubbing of fiat surfaces, J. Appl. Phys. 1953,24, (8) 981-988 Brown, K.J., Atkinson, JR., Dowson, D., and Wright, V., The wear of ultra high molecular weight polyethylene with reference to its use in prostheses in Plastics in Medicine and Surgery, The Plastics and Rubber Institute, London, 1975, p. 2.1 Dobbs, H.S., Characterization of ultra-high molecular weight polyethylene, Biomaterials 1982, 1, 49