luis 14
TRANSCRIPT
309
A hip wear simulator with 100 test stationsV SaikkoDepartment of Mechanical Engineering, Helsinki University of Technology, Laboratory of Machine Design, PO Box 4300,Helsinki, Fin-02015 HUT, Finland. email: [email protected] manuscript was received on 3 September 2004 and was accepted after revision for publication on 4 April 2005.
DOI: 10.1243/095441105X34301
Abstract: A novel high-capacity hip wear simulator of the pin-on-disc type was designed,built, and validated. This so-called Super-CTPOD (circularly translating pin-on-disc) device hasas many as 100 separate test stations, being an advanced version of the previously validated12-station CTPOD. A validity test was done so that in all stations the specimens and the testconditions were as similar as possible. Hence, for the first time in this field, an adequatenumber of similar tests was done for a proper statistical analysis of wear data. The pins wereconventional, gamma-sterilized ultra-high molecular weight polyethylene, and the discs werepolished CoCr. The lubricant was diluted calf serum and the test length 3 million cycles. Inthe course of the test, the pins became highly polished, whereas the discs remained practicallyunchanged. The majority of the polyethylene wear particles were rounded, with a meandiameter of 0.25 mm. The 100 wear factor values computed from the 100 steady state wear ratevalues of the pins were normally distributed, the mean ±95 per cent confidence interval being1.63±0.017×10−6 mm3/N m. The standard deviation was 5.4 per cent of the mean. Therewere no outliers. The wear mechanisms and the wear factor agreed well with clinical findings.Altogether, the Super-CTPOD test system was shown to be a unique combination of validity,low variation, capacity, efficiency, reliability, productivity, economy, ease of operation, andcompact size.
Keywords: hip wear simulation, ultra-high molecular weight polyethylene, wear datastatistics, wear test validation, high-capacity wear test
1 INTRODUCTION run, it can be stated that the scarcity of valid wear
testing capacity has been a true bottleneck in the
It is characteristic of wear tests that the results show
evolution of orthopaedic biomaterials.
a considerable variation, wear testing of orthopaedic In 1996, the present author designed and built abiomaterials being no exception. The variation is
12-station circularly translating pin-on-disc (CTPOD)
basically due to the fact that there are numerous device that was shown to produce wear highly similarfactors affecting the wear mechanisms, and the
to that occurring in total hip prostheses in vivo, being
phenomena lying behind the tribological behaviour the first truly simple design to reach this goal [1].of surfaces are complex. Some of the important
Hence, the CTPOD is called a hip wear simulator, in
variables, such as the degradation of the serum- distinction from hip joint simulators that are usedbased lubricant with time, are still poorly understood
for wear testing of actual hip prostheses. The crucial
and difficult to control. It is therefore most desirable factor was found to be the type of relative motion.to run a large number of similar tests for the statistical
In the circularly translating motion, the direction
analysis of test data and for the comparison ofof sliding changes continually, as is the case in thedifferences in means between materials, test con-
hip joint in walking. This type of sliding together
ditions, etc. Most of the existing wear test devices with a serum-based lubricant results in adhesivefor prosthetic joints and their materials are very
wear manifested as a polished appearance of the
expensive and have 12 test stations at the most. Since polyethylene wear surface and the production of poly-there are not too many of such devices, even world-
ethylene wear particles mostly in the 0.1–1 mm size
wide, and since one test takes at least 6–8 weeks to range. These observations are in good agreement with
H05704 © IMechE 2005 Proc. IMechE Vol. 219 Part H: J. Engineering in Medicine
310 V Saikko
clinical ones [2–4]. Moreover, the wear factor values against polished CoCr discs so that the specimensproduced by the CTPOD for conventional ultra-high and the test conditions in all 100 stations were
asmolecular weight polyethylene (UHMWPE) slidingsimilar as possible. This material combination wasagainst a polished CoCr counterface range from 1 to
chosen because it is the most commonly used com-
2×10−6 mm3/N m, being close to those measured bination in prosthetic joints, and there are plenty offor retrieved conventional metal/polyethylene total
scientific publications on its tribological behaviour.
hip prostheses [5]. All this is true despite the fact thatthe CTPOD test employs a flat-on-flat contact insteadof a ball-in-socket contact of the actual prosthetic 2 MATERIALS AND METHODShip. Interestingly, the CTPOD motion is analogous tothat of the most widely used hip joint simulator, the The Super-CTPOD design (Fig. 1) consists of threebiaxial rocking motion design, originally introduced
main modules: loading, motion, and pin guiding
by McKellop and Clarke [6]. The type of loading was(Fig. 2). The loading and pin guiding modules canfound to be clearly less important than the type of
be easily demounted, after which the specimens
motion [7]. (Fig. 3) can be disassembled with special tools (Fig. 4)In the present paper, the second generation offor cleaning, examination, and wear measureme
nt.the CTPOD design is introduced. This so-called The loading module is a rigid aluminium box thatSuper-CTPOD has 100 separate test stations to make
weighs 21 kg and contains 100 individual pneumatic
possible, for the first time, samples of sufficient size actuators. The maximum operating pressure of thefor a proper statistical analysis. The capacity of
actuator is 0.1 MPa, which generates a load of 165 N.
100 stations is unparalleled (Table 1). Despite the The motion module consists of a base plate, afact that the number of stations in so high andmotion plate, and the circular translation motio
nthe specimens are not miniaturized, the device ismechanism between them. The motion plate makesof very compact size, 384 mm×339 mm×333 mm the electromechanically driven circular translation in(w×d×h). With the set of special tools, the assemblythe horizontal plane. The motion plate is made ofand disassembly of the specimens is extremely quickand easy, although the test chambers are packedin the minimum of space. The motion, load, andspecimen size are identical to those of the 12-stationCTPOD. A new feature is the temperature controlsystem, which keeps the lubricant temperatureuniform within 1 °C in all 100 stations. The volumeof the lubricant in each test chamber is 16 ml, 33 percent higher than that in the 12-station CTPOD, toreduce the risk of excessive protein precipitation,which could compromise the validity of the wearsimulation. The only lacking feature of the Super-CTPOD compared with the 12-station CTPOD is thefriction measurement, which was dictated by spacelimitations.To see whether the device produces valid wear
results with reasonable variation, conventional Fig. 1 Super-CTPOD, 100-station hip wear simulatorgamma-sterilized polyethylene pins were tested with multidirectional motion
Table 1 Contemporary multidirectional motion pin-on-disc devices for hip wearsimulation
Type of device Motions Number of test stations Reference
Boston Two translations 1 with 6 specimens [8]Durham One translation, one rotation 4 [9]Winterthur Two rotations 4 [10]Ortho-POD Two rotations 6 [11]Leeds One translation, one rotation 6 [12]CTPOD Two translations 12 [1]Super-CTPOD Two translations 100 Present study
Proc. IMechE Vol. 219 Part H: J. Engineering in Medicine H05704 © IMechE 2005
A hip wear simulator with 100 test stations 311
Fig. 2 Super-CTPOD main modules (loading, motion, Fig. 4 Tools for assembly and disassembly of specimens,
and pin guiding) shown separated. Test chambers their holders, and test chambersfilled with serum-based lubricant and surroundedby coolant water are installed in the motionmodule. Pins are attached with press-fit sleevesto pin guiding shafts
test chamber and a piece of PVC (polyvinyl chloride)tube forms the brim with a height of 35 mm andinternal diameter of 28 mm (Fig. 3). The brim is fixedto the disc with a silicone O-ring. The chambers areplaced on the motion plate, locked horizontally bytwo guide pins per disc, and surrounded by the waterbath. The O-ring prevents the mixing of the lubricantand the circulating water. It was found in the systemtests that the lubricant temperature was one degreeabove the water temperature. For instance, if theincoming and outgoing water temperatures were18.5 and 19.5 °C respectively, the serum temperatureswere between 19.5 and 20.5 °C. The direction ofcirculation was regularly changed to reduce the tem-perature differences further between the chambers.The chambers are organized in 11 rows so that thenumber of chambers in the rows are 7, 8, 9, 10, 11,10, 11, 10, 9, 8, and 7. The rows are of unequal width
Fig. 3 One Super-CTPOD test chamber consisting ofthe test disc, PVC brim, and O-ring seal, andone test pin attached to guiding shaft with aplastic sleeve. Flanges of shafts protect the
testchambers against contamination from above.
in order to leave space for the pillars in the corners(Fig. 2). In this way, the dimensions of the devicewere minimized. The pillars carry the load and thefrictional force. The drive motor and the temperaturecontrol system are located beneath the installation
Note the diamond coating of the ends of stain- table shown in Figs 1 and 2.less steel shafts preventing crevice corrosion The pin-guiding module, with a total weight of
15 kg, is a thick polyacetal block having 100 precision-
stainless steel. Both of its sides are machined and
made bores for stainless steel precision shafts, the
ground. In addition, the upper side is polished and rotation of which is prevented by levers. A polyacetaldiamond coated. The motion plate is surrounded bysleeve is pressed around the lower end of the sha
fta brim forming a basin for the temperature controland the test pin is pressed into the lower end of the
system. The basin contains water circulated by a
sleeve (Fig. 3). By using this press-fit principle, the
pump through a heat exchanger. With the heat radial dimension of the pin holder is minimized.exchanger working by tap water run by a thermostat
The pin holder must be thin enough so that there is
valve, the water temperature within the basin can be enough space within the lubricant chamber with anaccurately controlled. There are 100 separate, similar
internal diameter of only 28 mm, considering the
test chambers. The test disc forms the bottom of the motion along a circular track of 10 mm diameter.
H05704 © IMechE 2005 Proc. IMechE Vol. 219 Part H: J. Engineering in Medicine
312 V Saikko
The advantages of the high shape of the chamber million cycles using a procedure published elsewhereare that the mixing of the lubricant by the pin [13]. During a weighing stop, the specimens, pinholder is efficient, preventing sedimentation, and the holders, lubricant chambers, and O-rings were care-surface area of the lubricant relative to its volume is fully cleaned. The pins were first vacuum desiccatedsmall, retarding evaporation. The pneumatic actuator for 30 min and then allowed to stabilize for 2 hourspushes the shaft downwards from its upper end.in the room atmosphere before the weighing, whichSince the shaft has only one degree of freedom in was done with a Mettler AT261 DeltaRange balancevertical translation, the contact force between the to a repeatability of ±0.03 mg. The original meanpin and the disc is equal to the force generated by weight of the pins was 717.11 mg (range 712.98–the pneumatic actuator. This is assisted by two 719.64 mg). After the weighing, the specimens werefactors: (a) there is a small clearance between the reassembled, and the test was continued with freshshaft and the bore, and (b) the frictional force with lubricant. The total length of the tests was 3 milliona continual change of direction tends to ‘whirl’ the cycles. The steady state wear rate, in mg per oneshaft within the bore, facilitating its axial movement million cycles, was taken to be the slope of the lineardownwards, as the pin shortens due to wear. regression line in the diagram of weight loss versusThe motion was such that the pin translated,number of cycles. The first half a million cycles
without rotation, along a circular track of 10 mm was omitted in the linear regression because it is adiameter on the disc. Therefore, the direction of transient phase dominated by the removal of thesliding relative to the pin changed continually. One original machining marks. Taking into account thecomplete revolution of the direction of sliding, density of polyethylene, the load, and the slidingwhich took 1.0 s, is called a cycle. The sliding speed distance per cycle, the wear factor k was calcu-was constant at 31.4 mm/s. The diameter of the lated so that the wear rate was divided by 1
a
06×cylindrical polyethylene pin was 9.0 mm and height 0.94 mg/mm3×70.7 N×0.0314 m. The wear factor is12.0 mm. The wear end of the pin was flat and a useful quantity because it makes possible the com-non-chamfered. The CoCr disc had a diameter ofparisons with in vitro and in vivo wear measure-28 mm and thickness of 10 mm. The contact was flat- ments for actual hip prostheses. All the 100 tests wereon-flat. The load was static 70.7 N and the nominal run simultaneously.contact pressure was therefore 1.1 MPa. Polyethylene wear particles were isolated fromThe pins and the discs were supplied by Center-samples of used serum lubricant by NaOH digestion,
pulse Orthopedics, a Zimmer Company. The pins HCl neutralization, and filtration on 0.05 mm pore sizewere made from conventional, calciumstearat-free polycarbonate filters, and examined with scanningultra-high molecular weight polyethylene GUR 1020, electron microscopy (SEM) and image analysis tools,‘Sulene-PE’, ISO 5834-1/-2. After the machiningas described elsewhere [14].from a compression moulded sheet, the pins weregamma-irradiated by 25–40 kGy in nitrogen. The discswere made from CoCrMo wrought alloy ‘Protasul-20’,ISO 5832-12. The wear faces of the discs were
3 RESULTS
machined, ground, and finally diamond-polishedto an arithmetical mean surface roughness value
In visual examination, all pins were highly polished
of 0.01–0.02 mm R . The surface roughness wasmeasured with a Taylor Hobson Surtronic 3+diamond stylus instrument using 0.8 mm cut-off and
(Fig. 5). Microscopy revealed a featureless, flat topo-graphy in every pin [Fig. 6(a)] with some mildcriss-cross scratches [Fig. 6(b)]. The discs were not
4 mm evaluation length. damaged, and no patches or layers formed on them
The lubricant was prepared so that triple 0.1 mm
during the tests. Only seldom some subtle, circular
sterile filtered HyClone Alpha Calf Fraction serum,
scratches were observed. The occurrence of these
Catalog number SH30076.03, lot number AFL5848,
rare, barely visible, scratches did not correlate with
was diluted 1:1 with Milli-Q-grade distilled water. the wear rate of the corresponding pins. The SEM
Hence, the protein content of the lubricant wasof the polycarbonate filters revealed that the over-
21 mg/ml. No additives were used. Each chamber
whelming majority of the wear particles (Fig. 7) were
contained 16 ml of lubricant. The lubricant tem-rounded and their mean equivalent circle diameter
perature during the test was maintained at 20 °C(range 19.5–20.5 °C).
(ECD)±SD (standard deviation) was 0.25±0.10 mm(n=125). Agglomerations, from which individual
The amount of wear was measured by stoppingparticles could not be distinguished with certainty,
the test and weighing the pins at intervals of half a were common.
Proc. IMechE Vol. 219 Part H: J. Engineering in Medicine H05704 © IMechE 2005
A hip wear simulator with 100 test stations 313
Fig. 5 Polyethylene test pins on the handling tray after 3 million cycles. Topographical featuresappear pronounced due to polishing. Reflection of light varies because the pin height onthis handling tray varies slightly
Fig. 6 Optical micrographs from the polyethylenesurface worn for 3 million cycles
The dependence of wear on the number of cycleswas highly linear, the correlation coefficient R2 of the Fig. 7 Scanning electron micrographs of polyethylenelinear regression of the mean weight loss values being wear particles on the polycarbonate filter with0.9982, the slope of the regression line 3.40 mg per a 0.05 mm pore size
H05704 © IMechE 2005 Proc. IMechE Vol. 219 Part H: J. Engineering in Medicine
314 V Saikko
values lay within the mean ±1 SD, i.e. between 1.54
Fig. 10 Normal probability plot of the wear factor k
one million cycles, and the y axis intercept 0.70 mg(Fig. 8). The mean k value computed from theslope was 1.63×10−6 mm3/N m. The same value wasnaturally obtained by performing the linear regressionanalysis for each pin and computing the mean ofthe 100 values. The correlation coefficient R2 of thelinear regression analysis of the gravimetric wear ofindividual pins ranged from 0.9906 to 0.9998 with amean of 0.9973. The 100 k values, computed fromthe 100 steady state wear rate values, were normallydistributed (see Table 2 and Figs 9 and 10). Therewere no outliers. Normality tests (Shapiro-Wilk W,Anderson-Darling, Martinez-Iglewicz, Kolmogorov-Smirnov, D’Agostino skewness, D’Agostino kurtosis,and D’Agostino omnibus) could not reject normalityat the 5 per cent significance level. In a normal distri-bution, 68.3 per cent of the observed values lie within
Fig. 9 Frequency histogram and relative cumulative
frequency of the wear factor k
Fig. 8 Gravimetric wear (weight loss) of polyethylenepins. Each circle represents the mean value of100 pins. Vertical bars indicate the standarddeviation
of weight
loss
since
previous
weighing (10−6 mm3/N m) with a regression line passin
gthrough the first and third quartiles, and 95
perTable 2 Descriptive statistics of the wear factor cent confidence limits
k (10−6 mm3/N m) in the Super-CTPODtest
Statistic Value ±1 SD of the mean. In the present test, 71 of 100 k
Mean±95 per cent confidence interval 1.63±0.017Sample variance 0.0077 and 1.71×10−6 mm3/N m.Standard deviation 0.088Minimum 1.41First quartile (25th percentile) 1.57Median 1.63 4 DISCUSSIONThird quartile (75th percentile) 1.68Maximum 1.86Sample size 100 A novel pin-on-disc hip wear simulator, the Super
-Kurtosis 3.16 CTPOD, was introduced. The most useful andSkewness 0.05
2n−2 distribution that the lowest value of n
valuable characteristics of the device, validity, low
Proc. IMechE Vol. 219 Part H: J. Engineering in Medicine H05704 © IMechE 2005
A hip wear simulator with 100 test stations 315
variation, capacity, efficiency, reliability, productivity,Low variation of the results indicated that the testeconomy, ease of operation, and compact size, are
method was of high quality. Achieving a 95 per cent
discussed below. confidence interval of the mean as small as that inValidity is the most important characteristic of a the present study, 1.63±0.017×10−6 mm3/N
m, iswear simulator. In the present case, the validity is very difficult with low n values, because the lengthjudged by the similarity of the wear produced by the of the interval is proportional to 1/√n. For the firstsimulator to that occurring in prosthetic hip jointstime in this field, it was shown that k is normallyin vivo. The crucial results, the polished appearance distributed. The comparison of the means of tw
oof the worn polyethylene surface, the undamaged samples using the t-test requires that the data arecounterface, the size and shape of the wear particles,
normally distributed. Consider a theoretical example:
and the mean wear factor value, agreed well with the wear properties of two types of polyethylene, Aclinical findings [2–5]. In the absence of severe
and B, are compared with the Super-CTPOD. The
femoral head damage, the main load-bearing zone mean k of type A is found to be 1.63×10−6 mm3/N mof retrieved polyethylene acetabular cups typically
and the mean k of type B is found to be 10 per cent
appears polished, indicating adhesion as the principal higher than that of type A, a difference not toowear mechanism [2, 3]. The majority of wear particles
dramatic but with potential clinical significance.
isolated from periprosthetic tissues are rounded and It is assumed that their wear factors are normallyrange in diameter from 0.1 to 1 mm with a mean of
distributed with SDs equal to that of Sulene-PE,
0.5–0.7 mm [3, 4]. In the present study, the mean ECD 0.088×10−6 mm3/N m. It is easy to calculate usingof wear particles was somewhat smaller, 0.25 mm, the Twhich may be at least partly due to the fact that the resulting in a 95 per cent confidence that type B reallyfilter pore size used was 0.05 mm, whereas the 0.2 mm
wears faster than type A is 4. This would leave an
pore size is the most widely used, and used also abundance of machine capacity for other materialsin the two studies of clinical particles [3, 4]. It has
to be run simultaneously. With new materials, the
been found that the filter pore size has a signifi-sample size must of course be decided before any-cant effect on the observed size distributions of the thing is known about the distributions, but thisparticles [15]. It is recommended that the 0.05 mm example gives an idea of the differentiation power of
pore size is used instead of 0.2 mm, because a signifi-
the Super-CTPOD. Note that in the present study,
cant proportion of particles are below 0.2 mm in (a) the run was successful at the first attempt, (b) alldiameter. The mean clinical wear factor measured
the 100 tests were run simultaneously, (c) there was
for 129 retrieved Charnley polyethylene acetabular not a single outlier pin with exceptionally low or highcups worn against 22 mm diameter stainless steel
k values (see Fig. 10, where all values are between
heads with varying abrasive damage was found to be the 95 per cent confidence limits), (d) none of the2.1×10−6 mm3/N m [5]. In the present study, there
results was rejected, and (e) there were no reruns.
was practically no damage on the discs, which may Moreover, the appearances of all 100 pins and of allpartly explain the slightly smaller mean wear factor
100 discs were highly uniform.
of 1.63×10−6 mm3/N m. Retrieved heads made of The capacity of 100 separate test stations isCoCr, which is harder than stainless steel, typically
superior to all previous devices (Table 1). The high
have retained their original mirror finish and show capacity is perhaps the single most original featureonly occasional scratches [3]. When scratching occurs,
of the Super-CTPOD. Wear testing of orthopaedic
it is often caused by abrasive acrylic particles, both in biomaterials is inevitably time-consuming. For athe cases of stainless steel [5] and CoCr heads [16]. thorough evaluation of, for example, the effect ofIn the present wear simulation, acrylic particles and different sterilization methods, different types ofall other possible contaminants were deliberately lubricants, or different types of counterface on theleft out, because first a valid, clean wear test that wear behaviour of a certain material, or for a com-will serve as a reference test for any future study
parison of the wear behaviour of a large number of
needed to be done. A scratched counterface has different materials, the capacity is of the utmostbeen shown to produce more large, elongated poly- importance. Typically, existing wear simulator an
dethylene particles [13]. The oxidation of polyethylene joint simulator units have 12 test stations at most,is another important issue, as it may seriously impair
making it impossible to implement large test pro-
the wear resistance in vivo [17]. The majority of grammes within a reasonable timescale. If miniaturelaboratory wear tests have been done with non-aged
specimens were used, it would not be too difficult to
material, as in the present study. pack even a much higher number of test stations
H05704 © IMechE 2005 Proc. IMechE Vol. 219 Part H: J. Engineering in Medicine
316 V Saikko
than 100 into one CTPOD device, but it has been polishing of the discs were done manually at the
shown that the wear behaviour of a miniaturized
Helsinki University of Technology (HUT), but the
contact surface is quite different from that of actual shape of the disc lends itself very well to modern,bearing surfaces in vivo [18]. Note that the Super- automatic polishing machines.CTPOD specimens are not miniaturized. On the Ease of operation is important from the point ofcontrary, 9.0 mm is a relatively large pin diameter.
view of efficiency, and special attention was paid to
Moreover, the wear face diameter can easily be this aspect in the design of the device. The Super-increased up to 12.5 mm, the outside diameter of the
CTPOD was the tenth wear test device designed,
pin holder sleeve. built, and operated by the present author, and so theEfficiency is, however, more than sheer capacity.
importance of this aspect was very clear from per-
The downtime for wear measurement, including sonal experience. For instance, all the parts that needdisassembly, cleaning, weighing, and reassembly,
to be cleaned (the specimens, pin holders, lubricant
must be as short as possible so that the entirechambers, and O-rings) are easily cleaned as batchestest can be performed in a minimum of time. Inin an ultrasonic cleaner. During the running, n
oSuper-CTPOD testing, the downtime proved to bemeasures by the operator are needed apart fromonly about 12 hours with one operator (V.S.). Withnormal supervision.a 1 Hz test frequency, a three million cycle test Its compact size, 384 mm×339 mm×333 mmfor 100 specimens with six periodic wear measure- (w×d×h, above the installation table shown inments can therefore be completed in just 38 days, as Figs 1 and 2), makes it easy to move the device andin the present study. No doubt, this is a mark of locate it, for example, in a clean air chamber,
ifefficiency. desired. The compact size makes the device and itsReliability, as an absence of technical malfunction,
modules light, but does not cause any inconvenience
proved to be as good as it can possibly be, since the for the operator thanks to the set of assembly andtest did not need to be stopped even once during disassembly tools for specimen handling (Fig. 4).the six 500 000 cycle test stages. It is essential for As the temperature was maintained at 20 °C, thereliability that the structure of the simulator isevaporation from the open lubricant chambers wa
ssound and robust from the machine design point of so minimal that the device did not need to beview, with all principles, functions, mechanisms,
stopped for the addition of water during the 139 hour
components, machine elements, and parts being periods that each 500 000 cycles took to run. Theoptimized and as simple as possible in relation tofact that the results were so similar to clinical fin
d-the objective, valid hip wear simulation with low ings proves that the lubricant temperature needvariation. not be kept at 37 °C. The lower temperature ha
sProductivity is naturally a result of the high the advantage that the microbial growth and the
capacity, efficiency, and reliability of the Super-
denaturation of the proteins is slower. Hence, the
CTPOD test system. Now it is possible, for the first wear mechanisms remain more stable. It is knowntime, to ‘mass-produce’ valid wear data without any
that the protein precipitation, which increases with
additional effort compared with existing low-capacity increasing temperature, can erroneously change thesimulators. Moreover, the Super-CTPOD design itself
wear mechanisms in a wear test so that precipitated
is ideal for serial manufacture. A set of, say, ten units protein slurry acts as an anti-wear agent [19], whichwith a massive total capacity of 1000 test stations
does not happen in vivo. Due to the microbial
would be a truly powerful hardware in ambitiousgrowth, the use of the 37 °C temperature necessitatesbiotribology research programmes. the use of antibiotics, such as sodium azide, th
atEconomy of the Super-CTPOD is based on its are not only toxic but also may affect the wear
sound design, consisting of simple parts made from
mechanisms. The present study showed that there is
common materials and of industrial componentsno need to use any additives whatsoever in the serum(pneumatics, bearings, electrical drive). The shapes
if the temperature is kept at 20 °C – not even EDTA
of the specimens are also simple, making them (ethylene diamine tetraacetic acid), since there waseconomical (Fig. 3). The pin is a simple cylinder and
no problem of calcium phosphate precipitation.
the only special features in the disc are the groove Insufficient lubricant volume is another possiblefor the O-ring, a recess for the O-ring removal tool,reason for the problematic protein slurry
formationand two 3 mm bores on the lower side for the locking [19]. Since no such problem was encountered inpins of the motion plate. No screws are needed inthe present tests, it can be deduced that 16 ml of
thetheir fixation. In the present study, the grinding and present lubricant, alpha calf serum diluted 1:1, is
Proc. IMechE Vol. 219 Part H: J. Engineering in Medicine H05704 © IMechE 2005
A hip wear simulator with 100 test stations 317
sufficient with a load of 70.7 N and sliding velocity 5 CONCLUSIONSof 31.4 mm/s, at least with the clinically most widelyused combination, polyethylene against CoCr. The
For the first time, the number of similar tests done
average wear factor, 1.63×10−6 mm3/N m, cannot beconsidered too low in relation to clinical findings [5].
for the widely used conventional UHMWPE wassufficiently large (100) for obtaining a 95 per cent
The temperature control system of the Super-CTPODallows the use of a wide range of temperatures. The
confidence interval for a mean wear factor as low as±1.1 per cent of the mean value and for proving that
minimum lubricant temperature that was achieved
the wear factor is normally distributed. This was made
in system tests was 8 °C, as the cold tap water tem-perature of the laboratory was 7 °C. On the otherhand, temperatures above 40 °C proved problematic,as they led to excessive evaporation.
possible by the low variation of results without asingle outlier, which is indicative of a high-quality testmethod. The test conditions were shown to closelysimulate the clinical conditions in the prosthetic
With the 12-station CTPOD, a test has been donehip joint. Moreover, the results indicated that inso that the specimens and the test conditions in all12 stations were similar to those of the present study[20]. The only differences were that the lubricantvolume was 12 ml and the average lubricant bulktemperature was 28 °C. The mean wear factor wasfound to be 1.19×10−6 mm3/N m, 27 per cent lowerthan that of the present study, and the SD was 9.2 percent of the mean. It is logical to assume that thelower k value in the 12-station CTPOD test was dueto the smaller lubricant volume and higher lubricanttemperature, as the amount of precipitated proteininvolved in the wear mechanism was probablyhigher, resulting in a lower wear rate, in conformitywith hip joint simulator studies [19, 21].The nominal contact pressure used, 1.1 MPa, i
slower than the values used by most other researchers.
order to produce wear similar to that known to occurclinically the lubricant temperature of 20 °C is pre-ferable to higher values due to factors related toprotein degradation that is inevitable in vitro incontrast to in vivo conditions.Altogether, the Super-CTPOD wear test system
wasshown to be a unique and superior combination ofvalidity, low variation, capacity, efficiency, reliability,productivity, economy, ease of operation, and com-pact size, making it the most original and usefulinnovation among contemporary wear simulatorsfor prosthetic joint materials. The Super-CTPOD isthe solution to the capacity problem of valid wearsimulation, and has true potential to mark a newepoch in the biotribology research of orthopaedicbiomaterials.
However, as the average load value in a hip jointduring a gait cycle is 1000 N [22], the average contactpressure value on the contact zone with this load
ACKNOWLEDGEMENTS
is likely to within the range 1–2 MPa. A simplifiedformula assuming hemispherical contact, ellipsoidalcontact pressure distribution, and a 28 mm headdiameter yields 0.785×0.386×1000/142=1.5 MPafor the average pressure [23]. The average value is asensible choice for a pin-on-disc device. Excessivepressure is likely to lead to abnormal thermal pheno-mena [18]. Overheating of the tribocontact is a pitfallto be avoided in wear testing. After all, a simulator
must, for practical reasons, run 24 hours a day, which
21 Liao, Y.-S., McKellop, H., Lu, Z., Campbell, P., and
on the effect of rotation on the wear of metal-on- J. Tribol., 2003, 125, 638–642.9 Scholes, S. C. and Unsworth, A. Pin-on-plate studies
on the wear of ultra-high molecular weight poly-20 Saikko, V. Effect of lubricant protein concentration
p. 73.Society, Anaheim, California, 1–4 February 1999,45th Annual Meeting of the Orthopaedic Research
8 Bragdon, C. R., O’Connor, D. O., Lowenstein, J. D.,
degradation and its effect on the outcome of hip
J. Engineering in Medicine, 1999, 213(H), 301–310.19 Wang, A., Polineni, V. K., Essner, A., Stark, C., and
lubricant in wear simulation of polyethylene7 Saikko, V. and Ahlroos, T. Type of motion and
directional motion pin-on-disk device: effects of
1985, ASTM STP 859, pp. 351–368 (American Society
Second Symposium (Eds A. C. Fraker and C. D. Griffin),In Corrosion and degradation of implant materials:wear of ultra-high-molecular-weight polyethylene.
6 McKellop, H. A. and Clarke, I. C. Degradation and
Saikko, V., Lehto, M. U. K., and Ja¨rvinen, M.Pajama¨ki, K. J. J., Halonen, P. J., Nevalainen, J. K.,
17 Puolakka, T. J. S., Kera¨nen, J. T., Juhola, K. A.,
5 Hall, R. M., Unsworth, A., Siney, P., and Wroblewski,
475–483.J. Biomed. Mater. Res. (Appl. Biomater.), 2002, 63,
16 Sychterz, C. J., Engh Jr, C. A., Swope, S. W.,
Agrawal, C. M. Standardized analysis of UHMWPE
counterfaces. It has been shown to produce valid
limited to the study of soft materials against hard
The usefulness of the CTPOD principle is not
The study was supported by the Academy of Finlandthrough a Senior Research Fellowship and Grant43328. The author thanks Dr Claude Rieker (ZimmerGmbH, Switzerland) for providing the polyethylenepins and the CoCr discs, Mr Ilmari Pajama¨ki, Mr TeroKoskenneva, and Mr Kai Koski for manufacturing the
parts of the simulator, Ms Seija Suojakari for grindingand polishing the discs, and Mr Jaakko Kera¨nen forthe SEM of wear particles.
undeniably differs from the activity of a typicalarthroplasty patient. The present results indicatedthat static loading is sufficient. Dynamic loading didnot appear necessary, but it is a feature that can be
REFERENCES
added later.1 Saikko, V. A multidirectional motion pin-on-diskwear test method for prosthetic joint materials
.J. Biomed. Mater. Res., 1998, 41, 58–64.
2 Atkinson, J. R., Dowson, D., Isaac, G. H., andwear also for hard-on-hard couples [24]. With these, Wroblewski, B. M. Laboratory wear tests and clinicalit is recommended that a ball-on-flat contact and observations of the penetration of femoral headsvery low loads are used. into acetabular cups in total replacement hip
joints.
H05704 © IMechE 2005 Proc. IMechE Vol. 219 Part H: J. Engineering in Medicine
318 V Saikko
II: a microscopical study of the surfaces of Charnley and crosslinked ultrahigh molecular weight poly-polyethylene acetabular sockets. Wear, 1985, 10
4,217–224.
3 McKellop, H. A., Campbell, P., Park, S.-H.,Schmalzried, T. P., Grigoris, P., Amstutz, H. C., andSarmiento, A. The origin of submicron polyethylenewear debris in total hip arthroplasty. Clin. Orthop.,1995, 311, 3–20.
ethylene studied with a multidirectional motion
pin-on-disk device. J. Biomed. Mater. Res., 2001, 57,506–512.
14 Calonius, O. and Saikko V. Analysis of polyethylene
particles produced in different wear conditions invitro. Clin. Orthop., 2002, 399, 219–230.
15 Scott, M., Widding, K., and Jani, S. Do current wear
4 Mabrey, J. D., Afsar-Keshmiri, A., Engh, G. A., particle isolation procedures underestimate the
Sychterz, C. J., Wirth, M. A., Rockwood, C. A., and number of particles generated by prosthetic bearing
components? Wear, 2001, 251, 1213–1217.wear particles from failed total joint arthroplasties.
McNulty, D. E., and Engh, C. A. Analysis of prosthetic
femoral heads retrieved at autopsy. Clin. Orthop.,1999, 358, 223–234.
B. M. Wear in retrieved Charnley acetabular sockets.Proc. Instn Mech. Engrs, Part H: J. Engineering inMedicine, 1996, 210(H), 197–207.
Increased volumetric wear of polyethylene linerswith more than 3 years of shelf-life time. Int.Orthop., 2003, 27, 153–159.
18 Saikko, V. and Ahlroos, T. Wear simulation of
UHMWPE for total hip replacement with a multi-for Testing and Materials, Philadelphia, Pennsylvania).
counterface material, contact area, and lubricant.J. Biomed. Mater. Res., 2000, 49, 147–154.
acetabular cup. Proc. Instn Mech. Engrs, Part H:
Dumbleton, J. H. Quantitative analysis of serum
joint simulator wear testing on UHMWPE. In theJasty, M., and Harris, W. H. Development of a newpin-on-disk testing machine for evaluating poly-ethylene wear. In 24th Annual Meeting of the Societyfor Biomaterials, San Diego, California, 22–26 April1998, p. 355.
ethylene sliding againsta CoCr counterface.metal samples. J. Mater. Sci. Mater. Med., 2001, 12,299–303. Benya, P. The effect of frictional heating and
forced10 Rieker, C. B., Schaffner, S., Scho¨n, R., and Konrad, R. cooling on the serum lubricant and wear of UHMWBi-directional screening wear test with clinical polyethylene cups against cobalt–chromium and validation. In 10th Conference of the European zirconia balls. Biomaterials, 2003, 24, 3047–3059.Orthopaedic Research Society, Wiesbaden, Germany, 22 Bergmann, G., Deuretzbacher, G., Heller, M.,
13–15 October 2000, p. O-42. Graichen, F., Rohlmann, A., Strauss, J., and Duda,11 Mazzucco, D. and Spector, M. Effects of contact G. N. Hip contact forces and gait patterns from
area and stress on the volumetric wear of ultrahigh routine activities. J. Biomech., 2001, 34, 859–871.molecular weight polyethylene. Wear, 2003, 254,23 Saikko, V. and Calonius, O. An improved method
of514–522. computing the wear factor for total hip prosthese
s12 Besong, A. A. and Fisher, J. Comparative wear involving the variation of relative motion and con-
of UHMWPE with uni-axial and multi-directional tact pressure with location on the bearing surface.frictional force. In 44th Annual Meeting of the Ortho- J. Biomech., 2003, 36, 1819–1827.paedic Research Society, New Orleans, Louisiana, 24 Saikko, V. and Kera¨nen, J. Wear simulation of16–19 March 1998, p. 773. alumina-on-alumina prosthetic hip joints using
a13 Saikko, V., Calonius, O., and Kera¨nen, J. Effect of multidirectional motion pin-on-disk device. J. Am.
counterface roughness on the wear of conventional Ceram. Soc., 2002, 85, 2785–2791.
Proc. IMechE Vol. 219 Part H: J. Engineering in Medicine H05704 © IMechE 2005