Fi

SI

a

ARRA

KABCFHFA

1

cobtvotiwut

oT

(

1d

Medical Engineering & Physics 33 (2011) 96–105

Contents lists available at ScienceDirect

Medical Engineering & Physics

journa l homepage: www.e lsev ier .com/ locate /medengphy

luid load support and contact mechanics of hemiarthroplastyn the natural hip joint

ainath Shrikant Pawaskar ∗, Eileen Ingham, John Fisher, Zhongmin Jinnstitute of Medical and Biological Engineering, University of Leeds, Leeds LS2 9JT, UK

r t i c l e i n f o

rticle history:eceived 16 March 2010eceived in revised form 12 August 2010ccepted 16 September 2010

eywords:rticular cartilageiphasicontact mechanicsinite elementemiarthroplastyluid load supportctivities of daily living

a b s t r a c t

The articular cartilage covering the ends of the bones of diarthrodial synovial joints is thought to haveevolved so that the loads are transferred under different and complex conditions, with a very high degreeof efficiency and without compromising the structural integrity of the tissue for the life of an individual.These loading conditions stem from different activities such as walking, and standing. The integrity ofcartilage may however become compromised due to congenital disease, arthritis or trauma. Hemiarthro-plasty is a potentially conservative treatment when only the femoral cartilage is affected as in case offemoral neck fractures. In hemiarthroplasty, a metallic femoral prosthesis is used to articulate against thenatural acetabular cartilage. It has also been hypothesized that biphasic lubrication is the predominantmechanism protecting the cartilage through a very high fluid load support which lowers friction. This maybe altered due to hemiarthroplasty and have a direct effect on the frictional shear stresses and potentiallycartilage degradation and wear. This study modelled nine activities of daily living and investigated thecontact mechanics of a hip joint with a hemiarthroplasty, focussing particularly on the role of the fluid

phase.

It was shown that in most of the activities studied the peak contact stresses and peak fluid pressureswere in the superior dome or lateral roof of the acetabulum. Total fluid load support was very high (∼90%)in most of the activities which would shield the solid phase from being subjected to very high contactstresses. This was dependent not only on the load magnitude but also the direction and hence on thelocation of the contact area with respect to the cartilage coverage. Lower fluid load support was found

s nea

when the contact area wa

. Introduction

A person undergoes a series of activities including walking,limbing stairs, rising from a chair in the course of a day. The jointsf the lower extremities have to bear not only the weight of theody but also the forces that are generated due to muscles andheir moments. The hip joint for example, is known to withstandery high loads of 7–9 times body weight [1,2] with the highestf those forces experienced during stumbling. The acetabulum andhe femur of the hip joint also articulate with varying speed depend-

ng upon the activity and this may range between extremes of slow

alking and fast running of an athlete. However, the opposing artic-lar cartilages usually survive the lifetime of an individual despitehe harsh operating conditions they are subjected to. Cartilage may

∗ Corresponding author at: Institute of Medical and Biological Engineering, Schoolf Mechanical Engineering, The University of Leeds, Leeds LS2 9JT, UK.el.: +44 0113 343 5011; fax: +44 0113 242 4611.

E-mail addresses: s.s.pawaskar@leeds.ac.uk, men2ssp@leeds.ac.ukS.S. Pawaskar).

350-4533/$ – see front matter © 2010 IPEM. Published by Elsevier Ltd. All rights reserveoi:10.1016/j.medengphy.2010.09.009

rer the edges where the fluid drained easily.© 2010 IPEM. Published by Elsevier Ltd. All rights reserved.

however break down due to injury, congenital disease or the devel-opment of arthritis [3,71]. Mechanical factors may cause structuralas well as biochemical changes in articular cartilage [4–6].

Loss of cartilage in the hip joint may lead to total hip jointreplacement to alleviate pain and improve quality of life. Whenonly the femoral head is affected, hemiarthroplasty is an option.In hemiarthroplasty only the femoral head is replaced by a rigidmetallic prosthesis, which then articulates with natural acetabularcartilage. To understand the potential effect of hemiarthroplasty onthe acetabular cartilage and improve prosthetic design, it is impor-tant to investigate the conditions under which the joint operatesduring different activities of daily living.

Instrumented prostheses have been used to measure in vivo con-tact forces [2,7–10] and contact stresses [11–15] in both natural andartificial hip joints. Bergmann et al. have used instrumented pros-theses to record the hip contact forces for nine activities [16]. They

found the forces to be as low as 0.26 time body weight (BW) duringrising from a chair and as high as 2.6 times BW during stair descent.

Many numerical studies have concentrated on the contactstresses [17,18] since they play an important role in the tribol-ogy of the cartilage [12,19], cartilage degradation [20–22] and

d.

gineer

“pammdhFsaic

bmTithdcctle

iatdtv

2

(uAao

S.S. Pawaskar et al. / Medical En

preoperative planning and postoperative rehabilitation” [18]. Therediction of contact area from contact mechanics analysis maylso give an indication of the areas of acetabular cartilage which areost susceptible to breakdown. A finite element model of contactechanics of the human hip joint has been experimentally vali-

ated using cadaveric tissue in which cartilage was modelled as ayperelastic material [23]. However, none of the three-dimensionalE/numerical studies, to the best of authors’ knowledge, have con-idered the biphasic nature of the cartilage. Hence, those studiesre not able to account for interstitial fluid pressurization and itsnfluence on tribology and the contact mechanics of the articularartilage within the joint.

Biphasic lubrication, which is due to the load partitioningetween fluid and solid phases with the fluid phase sustainingost of the load, aids in maintaining a very low friction [24–26].

his coefficient of friction of the articulating cartilages usually liesn the range of 0.001–0.02 [24,27,28]. The role of fluid pressurisa-ion within cartilage in reducing the frictional coefficient has beenypothesized by many authors [25–27,29–36]. It has not only beenirectly measured [37–39] but has also been shown to be linearlyorrelated with the coefficient of friction in sliding experiments ofartilage against glass under constant load [39]. It has been shownhat the fluid phase is capable of supporting more than 90% of theoad thus resulting in low solid-to-solid contact and hence a lowerffective coefficient of friction [27,32,37,38].

Thus, the biphasic properties of articular cartilage play anmportant role in the tribology of natural synovial joints. Hemi-rthroplasty may alter the biphasic fluid load support. The aim ofhis study was to investigate the hip joint with a hemiarthroplastyuring several activities of daily living in order to understand theribology and contact mechanics of the biphasic cartilage underarying and complex conditions.

. Material and models

A solid model of male left pelvis created from CT scans [40]

Fig. 1) was used to create finite element (FE) model of the hip bysing I-DEAS (ver. 11, Siemens PLM Software, Plano, TX, USA) andBAQUS (ver. 6.7-1, Dassault Systemes, Suresnes Cedex, France). FEnalyses considering different activities of daily living were carriedut using ABAQUS.

Fig. 1. FE model of hip hemiarthroplasty.

ing & Physics 33 (2011) 96–105 97

A spherical horseshoe shaped acetabular cartilage of uniformthickness of 2 mm was used in all the simulations. The acetabularbone cavity was of radius 30 mm and the inner radius of the acetab-ular cartilage was 28 mm [41,42]. The acetabular cartilage wasapproximately divided into four regions; viz. lateral roof, medialroof, anterior and posterior horns [18]. The superior dome wasassumed to be at the centre covering parts of lateral and medialroofs. The cortical bone layer of the pelvis was assumed to havethickness of 1.41 mm [43]. A femoral prosthetic head of radius27.5 mm was used throughout to give a radial clearance of 0.5 mm.The centres of both the cartilage surface and prosthetic head wereat the origin of the coordinate system.

Biphasic poroelastic elements were used for the acetabular car-tilage whereas bones were assumed linearly elastic. The elementsused for each component of the hip joint along with their mate-rial properties are given in Table 1. There was a minimum of threeelements through the cartilage thickness [23]. The permeability ofthe cartilage was k = 9.83 × 10−16 m4/N s [44] and the water contentwas 80% [45].

The cartilage solid phase was modelled as neo-Hookean, withelastic strain energy potential given in Eq. (1) [46].

W(C) = G

2(I1 − 3) + K

2(J − 1)2 (1)

where, C, right Cauchy–Green deformation tensor; G, Shear mod-ulus; I1, first deviatoric strain invariant; K, Bulk modulus; J, totalvolume ratio when linear thermal expansion strain is not consid-ered.

Bulk modulus, K and shear modulus, G are related to Young’smodulus, E and Poisson’s ratio, � by Eqs. (2) and (3).

The surfaces of the acetabular cartilage and metallic prosthetichead were slave and master respectively. The head was coarselymeshed due to strict master-slave algorithm which prevents slavenode penetration into the master [46]. The mesh was finalized aftersensitivity analysis to ensure that the error in predictions was lessthan 5% when the consecutive meshes of different densities wereused.

E = 3K(1 − 2�) (2)

E = 2G(1 + �) (3)

The pelvis was pinned by restricting all three translationaldegrees of freedom of the nodes of the sacro-iliac joint and those ofthe contralateral side of the pubic symphysis. The peripheral sur-faces of the cartilage through its thickness were always exposedand hence free flow was prescribed on these surfaces. The back sur-face of the acetabular cartilage was tied to the impermeable lunatesurface of the acetabular cavity. Fluid flow on the contacting carti-lage surface was imposed depending upon developing contact [47].Frictionless contact was assumed.

One cycle each of the nine activities of daily living as shownin Table 2 was simulated with their respective load vectors [16].A long-time duration of five cycles was analysed only for normalwalking. Femoral rotation was already accounted for in these loadvectors. The hip joint contact force data used in this study alsoaccounted for muscle forces [16]. The pelvis rotates about boththe transverse axes during activities [16]. The load vectors wererotated to take into account this pelvic orientation. The loads wereapplied at the centre of the head, as shown in Fig. 1. The entire anal-ysis was quasi-static with different load vectors being applied oneafter the other in each step without changing any of the boundary

conditions.

Another set of models, representing all the activities were anal-ysed using non-linear void ratio dependent permeability (VRDP).This was calculated using Eq. (4) [48,49]. Material parameters, Mand � used in this equation were 4.638 and 0.0848 respectively

98 S.S. Pawaskar et al. / Medical Engineering & Physics 33 (2011) 96–105

Table 1Elements used for various components of hip joint along with their material properties.

Component Element Elastic modulus, E (MPa) Poisson’s ratio, � Reference

Type Quantity

Cancellous bone (pelvis) C3D4 12,165 70 0.2 [72]Cortical bone (pelvis) C3D6 3468 17,000 0.3 [72]Subchondral bone (pelvis) C3D4 247 2000 0.3 [72]Acetabular cartilage C3D8RP 14,772 1.072 0.011 [44]Prosthetic head C3D6 432 220,000 0.3 [73]

C3D8 2304

Table 2List of activities with their start and end [16].

Activity Starts at Ends at Cycle time (s)

Slow walk (0.98 m/s) Heel strike Ipsilateral heel strike 1.248Normal walk (1.09 m/s) Heel strike Ipsilateral heel strike 1.103Fast walk (1.46 m/s) Heel strike Ipsilateral heel strike 0.953Stand Up (chair height – 500 mm) Beginning of getting up Standing position 2.489Sit down (chair height – 500 mm) Standing position Sitting in a relaxed position 3.719Down stairs (stair height – 170 mm) Toe off Ipsilateral toe off 1.439Up stairs (stair height – 170 mm) Heel strike Ipsilateral heel strike 1.593Knee bend Standing position Standing position 4.665Stand 2-1-2 leg Two legged stance Two legged stance 6.703

Fig. 2. (a) Peak contact pressure, (b) peak fluid pressure, (c), (d) acetabular contact area and (e) total fluid load support during first cycle of slow, normal and fast walking.

S.S. Pawaskar et al. / Medical Engineering & Physics 33 (2011) 96–105 99

Table 3Maximum peak contact pressure with corresponding peak fluid pressure, contact area and total fluid load support (TFLS) for different activities and where and when theyoccurred.

Activity Peak contact pressure (MPa) Peak fluid pressure (MPa) Where Contact area (%) TFLS (%) Cycle time (%)

Slow walk 2.97 2.77 Superior dome 49.64 90.11 16.5Normal walk 2.78 2.59 Superior dome 52.00 91.21 15.5Fast walk 2.99 2.77 Superior dome 53.10 90.95 12.5Stand up 2.98 2.53 Posterior horn 55.13 84.10 44.5Sit down 2.57 2.19 Posterior horn 49.88 85.39 45.0

LateSupMedLate

[wud

k

F

Down stairs 4.63 3.85Up stairs 3.00 2.77Knee bend 2.42 2.10Stand 2-1-2 leg 4.40 3.71

50]. Initial permeability (k0) and initial void ratio (e0) representingater content, were constant permeability and void ratio val-es respectively, that were used in the models without void ratioependent permeability.

= k(

e

e0

)�

exp

(M

2

[(1 + e

1 + e0

)2− 1

])(4)

ig. 3. Contours of contact stresses (MPa) in acetabular cup during different phases of fir

ral roof 38.88 81.80 88.0erior dome 53.79 91.07 16.0ial roof nearer posterior side 46.77 86.99 53.5ral roof 40.43 83.42 46.0

3. Results

The variation of peak contact pressure on the acetabular car-tilage surface for one cycle of slow, normal and fast walking is

depicted in Fig. 2a. In the case of normal walking, the maximumpeak contact pressure of 2.78 MPa was found in the superior domeof the acetabulum at 15.5% of the cycle (Fig. 2a and Table 3). Thepressure distribution was in the antero-posterior direction butslightly towards the posterior side similar to that observed at 15%

st cycle of normal walking (A – anterior; P – posterior; M – medial; L – lateral).

100 S.S. Pawaskar et al. / Medical Engineeri

Table 4Average total fluid load support (TFLS) for different activities.

Activity Average TFLS (%)

Slow walk 89.28Normal walk 90.96Fast walk 90.08Stand up 87.74Sit down 87.18Down stairs 88.45Up stairs 89.82Knee bend 88.97Stand 2-1-2 leg 88.23

a5p2

tttvtHtrs

nbta

tion forces for the stance phase [7,15]. The peak contact pressure

Fig. 4. Total fluid load support for five normal walking cycles.

nd 20% cycle time (Fig. 3). The corresponding contact area was2.00% of the total potential contact area and total fluid load sup-ort was 91.21%. The maximum fluid pressure at this instant was.59 MPa.

The variation of peak contact pressure, peak fluid pressure andhe contact area with respect to time was similar to that of con-act force for all three walking speeds (Fig. 2a–d). The variation ofotal fluid load support in general seemed to follow contact forceariation for slow and fast walking whereas it deviated towardshe last 35% of the cycle in the case of normal walking (Fig. 2e).owever, it always remained high and on an average it was found

o be 89.28%, 90.96% and 90.08% for slow, normal and fast walkingespectively (Table 4). The total fluid load support decreased onlylightly (0.17%) over five cycles of normal walking (Fig. 4).

The contours of the contact pressure at different stages of one

ormal walking cycle for the first cycle are shown in Fig. 3. It cane seen that throughout the stance phase when the load was high,he contact was mostly maintained in the superior dome of thecetabulum. It then started moving towards the medial roof during

Fig. 5. Contours of contact stresses

ng & Physics 33 (2011) 96–105

the swing phase. It moved more medially in cases of slow and fastwalking compared to normal walking (Fig. 5).

The predictions of important parameters for all the activitiesthat were analysed are listed in Tables 3 and 4. The predicted peakcontact pressure, peak fluid pressure, contact area and total fluidload support for standing up and sitting down are shown in Fig. 6.The variation of parameters for going down the stairs and climbingup the stairs is shown in Fig. 7 as a function of percentage cyclewhereas the results for knee bending and standing on one leg areshown in Fig. 8. The peak contact pressure, peak fluid pressure andacetabular contact area for different activities generally showedthe same time-dependent trend as that of contact force. However,while going down the stairs and in one-legged stance, the contactarea variation was found to be somewhat deviating from that ofthe contact force (Figs. 7c, d and 8c, d). The variation of total fluidload support was different to that of the contact force for almostall activities except for some similarity in slow and fast walking(Fig. 2e).

When void ratio dependent permeability was used, the vari-ations of all the parameters of interest remained similar to thepredictions with constant permeability. The maximum differencewith respect to constant permeability predictions was 3.87% in totalfluid load support during standing up. In normal walking this dif-ference was even less and was around 1.93%. The variation of totalfluid load support during normal walking for void ratio dependentmodel and the one without is shown in Fig. 9 which shows a closesimilarity between the two curves.

4. Discussion

A limited number of FE/numerical studies of the contactmechanics of the hip joint exist. However, none of these studieshave investigated the role played by interstitial fluid pressuriza-tion in the contact mechanics and tribology of the cartilage in awhole joint model. Therefore nine different activities of daily liv-ing were simulated to investigate the extent of effect that the fluidphase has in the tribological functioning of the hip joint after hemi-arthroplasty. The methodology used in analyzing these models wasvalidated using porcine acetabular cups loaded with a rigid metallicprosthesis [70].

The variation of acetabular cartilage peak contact pressure fol-lowed that of contact force in all the activities of daily living asobserved previously [17]. A similar correlation was observed byPark et al. with respect to hip joint forces [15] and ground reac-

was maximal when going down the stairs just before toe-off. Thiswas the most strenuous of all the activities investigated. Standingon one leg was the next demanding activity. In both these activi-ties the corresponding contact areas were smaller and were on the

at 84% of first walking cycle.

S.S. Pawaskar et al. / Medical Engineering & Physics 33 (2011) 96–105 101

F area a

lshTt

attssdorptfltiIa

ig. 6. (a) Peak contact pressure, (b) peak fluid pressure, (c), (d) acetabular contact

ateral roof of the acetabulum. The contact stresses for descendingtairs were found to be higher than those during stair ascent asas been also observed in clinical [14] and analytical studies [51].he variation of peak fluid pressure also showed patterns similaro those of contact forces for all the activities of daily living.

The total fluid load support in all the activities was found to beround 90%. This reduced the load supported by the solid phase ofhe cartilage which would reduce the effective coefficient of fric-ion [26,27,29,31,34,39]. This would then reduce the frictional sheartresses thus protecting the cartilage from wear. This fluid loadupport was high over most of the cycle for all the activities. Therop in the fluid load support was very small even after five cyclesf normal walking. The migration of contact probably helped inehydration of the cartilage thus maintaining high interstitial fluidressurization [52,19]. It has been also hypothesised that the con-act migrating faster than the diffusive velocity of the interstitial

uid (∼10−4–10−6 mm/s) does not allows enough time for the fluido flow to the regions of low pressure and may help in maintain-ng higher fluid pressurisation over longer time duration [53,36].t reduced only when the contact moved towards the edges of thecetabular cartilage. This was because, the acetabular labrum was

nd (e) total fluid load support during first cycle of standing up and sitting down.

not modelled in the present study and free flow was prescribedon the edges of the cartilage. Thus more fluid exudation occurredwhen the contact moved towards any edge. In the presence of thelabrum which has lower permeability than the cartilage [54], thisedge effect will reduce. The anomaly seen in total fluid load supportbetween 60% and 95% of the normal walking cycle in Fig. 2e wasdue to the contact being slightly away from the medial edge as seenin Fig. 5.

The contact area was generally small in spite of the conform-ing contacting surfaces. The contact areas for all activities did notexceed 55.59% which happened at 47.5% of the standing up cycle.The contact moved medially but slightly towards the posterior hornduring this time. This was lower than that observed by Yoshida andcolleagues in their discrete element analysis study [18]. However,the clearance used by Yoshida et al. was not mentioned. Acetabularfit has been cited as an important parameter in the prevention of

acetabular erosion [55,56]. The current study used a radial clear-ance of 0.5 mm between the acetabular cartilage and the metallicprosthetic head which might have reduced the contact areas. Thesmallest contact area was 23.77% at 7% of the standing up cyclewhen the contact moved to the posterior horn of the acetabular cup.

102 S.S. Pawaskar et al. / Medical Engineering & Physics 33 (2011) 96–105

F area as

Iach

wuaptcttihmpbml

ig. 7. (a) Peak contact pressure, (b) peak fluid pressure, (c), (d) acetabular contacttairs.

t should be noted that the commercial femoral head prosthesesre available in increments between 1 and 2 mm [57]. Thus a radiallearance of 0.5 mm represented the smallest realistic clearance foremiarthroplasty for spherical acetabular cup assumption.

The acetabular cavity was assumed to be spherical in this studyhich is not the case in an anatomical joint. The non-spherical artic-lar surface would, in turn, introduce variable clearance in the jointnd adversely affect both the contact pressures and fluid load sup-ort. The total fluid load support in this case might be reduced ifhere was a greater area available for fluid exudation (as in higherlearances) or the permeability was increased (when water con-ent increases as in case of OA) [45,58]. This would then increasehe coefficient of friction and hence frictional shear stresses. Thencreased contact stresses along with the increased shear stressesave the potential to induce cartilage fibrillation thus compro-

ising the integrity of the hip joint in general and cartilage in

articular. In OA joints in which the cartilage structure has alreadyeen compromised and diminished, higher local contact stressesight enhance this effect. However, it should be noted, that the

ong term survivorship has been shown in both unipolar [59] and

nd (e) total fluid load support during first cycle of going down stairs and climbing

bipolar [60] hemiarthroplasties which may be due to higher fluidload support and lower contact stresses.

Contact area and peak contact pressure depended not only onthe magnitude of the load but also on the location of the contact.For example, in standing on one leg, contact stresses were 3.65 MPaand 2.88 MPa for similar loads of 1937 N and 1935 N respectivelyduring two different stages of the activity. However, in the firstcase the contact was near the lateral roof (contact area – 42.45%)whereas in the second case it was in the superior dome (contact area– 48.43%) where larger area was available for contact. The contactareas and their location also varied depending upon the activity. Inmost of the activities the contact was found in the superior domeof the acetabular cartilage as can be seen from Table 3. McGibbonet al. observed this frequent loading of the superior dome in theirclinical hemiarthroplasty study which may explain cartilage degra-

dation in this area [14]. Although the present study focused onhemiarthroplasty, these findings may also be related to the naturaljoint. The thickness of the cartilage has been hypothesized to varywith contact stresses [14], the thickest cartilage is found antero-superiorly near lateral roof [61,62]. The concentration of stresses

S.S. Pawaskar et al. / Medical Engineering & Physics 33 (2011) 96–105 103

F t areao

ih

tstp[ahemwemchrpb

ig. 8. (a) Peak contact pressure, (b) peak fluid pressure, (c), (d) acetabular contacne leg.

n the superior dome and near lateral roof seemed to support thisypothesis.

The values of contact pressure in general were however lowerhan those measured using endoprosthesis in hemiarthroplastytudies [12,14,15,63]. For example, in normal walking, peak con-act pressure in the present study was 2.78 MPa during the stancehase as compared to clinical values of 5.5 MPa [13] and 3.69 MPa63]. Hodge et al. also reported a reduced peak pressure of 4.0 MPafter 36 months of surgery [13] during walking. Bachtar et al.ave reported 5.5 MPa in their finite element study [17]. How-ver, predictions in the current study were more in line with theathematical/numerical models in which a spherical geometryas assumed for the acetabulum and femoral head [64–66]. Ipavec

t al. found a peak stress of 3 MPa during the stance phase of nor-al walking. In a discrete element model with spherical acetabular

artilage of uniform thickness, peak contact pressure of 3.26 MPaas been reported for normal walking [18]. It has been shownecently that both conchoid and spherical shapes underestimateeak contact stresses by nearly 50% and overestimate contact areasy around 25% when compared to subject-specific models [67]. It

and (e) total fluid load support during first cycle of knee bending and standing on

should be noted that the study of failure is associated with materialproperties as well as level of stresses.

The use of void ratio dependent permeability did not changethe predictions substantially in spite of large cartilage deformation(around 44% when maximum load of 2.6 times BW was appliedduring walking down the stairs). This may be due to a very lowpermeability of cartilage and a small variation in it due to corre-sponding small change in void ratio (∼0.06% of initial value).

One of the limitations of the current study was that the predic-tions were totally dependent on only one set of kinematic and forcedata [16]. Thus some of the inferences may be purely due to thedata being used and might not be a general trend, for example, thecontact moving away from the medial edge in normal walking vis-à-vis slow and fast walking (Fig. 5). Moreover, as mentioned above,a more realistic geometry of the acetabular cartilage needs to be

considered to take into account the effect of variable clearance.

The contact stresses and fluid load support have been predictedfor the hemiarthroplasty in the hip. Although the stresses werelower than in total joint replacement [68], currently it is not knownhow natural cartilage will respond to this level of cyclic stress over

104 S.S. Pawaskar et al. / Medical Engineeri

Fb

pstifn

atclpta

A

S(wbfF

A

LCCCC

EGK�V

C

tP

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

ig. 9. Comparison of total fluid load support during first cycle of normal walkingetween model using void ratio dependent permeability and the one without.

rolonged periods. Further experimental work is needed to under-tand the response of articular cartilage in the hip to this type ofribological and biomechanical demand. We have recently shownn the knee that at higher levels of contact stress and shear stress,ailure of cartilage can occur [69]. Further experimental studies areeeded in the hemiarthroplasty in the hip.

In conclusion, the present study showed that mean contactreas were generally only around 40% of the total potential con-act area despite the surfaces being conforming. Certain activitiesould result in an increase in contact stresses and decrease in fluidoad support consistent with the load. However, the fluid load sup-ort was high in most of the activities aiding in stress shielding ofhe cartilage matrix. This may explain the remarkable survival ofrticular cartilage in the hemiarthroplasty.

cknowledgements

Sainath Shrikant Pawaskar was supported by Overseas Researchtudents Awards Scheme. This work was supported by the NIHRNational Institute for Health Research) as a part of collaborationith the LMBRU (Leeds Musculoskeletal Biomedical Research Unit),

y EPSRC, by the Leeds Centre of Excellence in Medical Engineeringunded by the Wellcome Trust and EPSRC, WT088908/z/09/z. Johnisher is an NIHR senior investigator.

ppendix A

ist of notations3D4 four-node linear tetrahedral elements3D6 six-node linear triangular prism3D8 eight-node linear brick3D8RP eight-node trilinear displacement and pore pressure,

reduced integrationYoung’s modulusShear modulusBulk modulusPoisson’s ratio

RDP void ratio dependent permeability

onflict of interest

The authors do not have any conflict of interests pertaining tohe study submitted for publication in the Medical Engineering andhysics.

[[

[

ng & Physics 33 (2011) 96–105

References

[1] Crowninshield RD, Johnston RC, Andrews JG, Brand RA. A biomechanical inves-tigation of the human hip. Journal of Biomechanics 1978;11(1–2):75–85.

[2] Bergmann G, Graichen F, Rohlmann A. Hip joint loading during walking and run-ning, measured in two patients. Journal of Biomechanics 1993;26(8):969–90.

[3] Meachim G. Cartilage breakdown. In: Owen R, Goodfellow J, Bullough P, edi-tors. Scientific foundation of orthopaedics and traumatology. London: WilliamHeinemann Medical Books Limited; 1980. p. 290–6.

[4] Radin ELMD, Ehrlich MGMD, Chernack RBS, Abernethy PMS, Paul ILSD, RoseRMSD. Effect of repetitive impulsive loading on the knee joints of rabbits.Clinical Orthopaedics and Related Research 1978;131:288–93.

[5] Muehleman C, Arsenis CH. Articular cartilage. Part II: the osteoarthritic joint.Journal of the American Podiatry Medical Association 1995;85(5):282–6.

[6] Aigner T, McKenna L. Molecular pathology and pathobiology of osteoarthriticcartilage. Cellular and Molecular Life Sciences (CMLS) 2002;59(1):5–18.

[7] Rydell N. Forces in the hip-joint. Part (II) intravital measurements. In: KenediRM, editor. Proceedings of a symposium on biomechanics and related bio-engineering topics. Oxford: Symposium Publications Division, Pergamon Press;1965. p. 351–7.

[8] English TA, Kilvington M. In vivo records of hip loads using a femoral implantwith telemetric output (a preliminary report). Journal of Biomedical Engineer-ing 1979;1(2):111–5.

[9] Davy D, Kotzar G, Brown R, Heiple K, Goldberg V, Berilla J, et al. Telemetric forcemeasurements across the hip after total arthroplasty. Journal of Bone and JointSurgery American Volume 1988;70(1):45–50.

10] Bergmann G, Graichen F, Rohlmann A, Linke H. Hip joint forces during loadcarrying. Clinical Orthopaedics and Related Research 1997;335:190–201.

11] Rushfeldt PD, Mann RW, Harris WH. Improved techniques for measuring invitro the geometry and pressure distribution in the human acetabulum – i.Ultrasonic measurement of acetabular surfaces, sphericity and cartilage thick-ness. Journal of Biomechanics 1981;14(4):253–5.

12] Hodge WA, Fijan RS, Carlson KL, Burgess RG, Harris WH, Mann RW. Con-tact pressures in the human hip joint measured in vivo. Proceedings of theNational Academy of Sciences of the United States of America 1986;83(9):2879–83.

13] Hodge W, Carlson K, Fijan R, Burgess R, Riley P, Harris W, et al. Contact pressuresfrom an instrumented hip endoprosthesis. Journal of Bone and Joint SurgeryAmerican Volume 1989;71(9):1378–86.

14] McGibbon CA, Krebs DE, Trahan CA, Trippel SB, Mann RW. Cartilage degen-eration in relation to repetitive pressure: case study of a unilateral hiphemiarthroplasty patient. The Journal of Arthroplasty 1999;14(1):52–8.

15] Park S, Krebs DE, Mann RW. Hip muscle co-contraction: evidence from con-current in vivo pressure measurement and force estimation. Gait and Posture1999;10(3):211–22.

16] Bergmann G. Hip98. Berlin: Free University; 2001;Bergmann G, Deuretzbacher G, Heller M, Graichen F, Rohlmann A, Strauss J,Duda GN. Hip contact forces and gait patterns from routine activities. Journalof Biomechanics 2001;34(7):859–71.

17] Bachtar F, Chen X, Hisada T. Finite element contact analysis of the hip joint.Medical and Biological Engineering and Computing 2006;44(8):643–51.

18] Yoshida H, Faust A, Wilckens J, Kitagawa M, Fetto J, Chao EY-S. Three-dimensional dynamic hip contact area and pressure distribution duringactivities of daily living. Journal of Biomechanics 2006;39(11):1996–2004.

19] Katta J, Jin Z, Ingham E, Fisher J. Effect of nominal stress on the long term fric-tion, deformation and wear of native and glycosaminoglycan deficient articularcartilage. Osteoarthritis and Cartilage 2009;17(5):662–8.

20] Hadley NA, Brown TD, Weinstein SL. The effects of contact pressure elevationsand aseptic necrosis on the long-term outcome of congenital hip dislocation.Journal of Orthopaedic Research 1990;8(4):504–13.

21] Maxian TA, Brown TD, Weinstein SL. Chronic stress tolerance levels for humanarticular cartilage: two nonuniform contact models applied to long-termfollow-up of cdh. Journal of Biomechanics 1995;28(2):159–66.

22] Mavcic B, Pompe B, Antolic V, Daniel M, Iglic A, Kralj-Iglic V. Mathematicalestimation of stress distribution in normal and dysplastic human hips. Journalof Orthopaedic Research 2002;20(5):1025–30.

23] Anderson AE, Ellis BJ, Maas SA, Peters CL, Weiss JA. Validation of finite ele-ment predictions of cartilage contact pressure in the human hip joint. Journalof Biomechanical Engineering 2008;130(5):051008–10.

24] Mow VC, Lai WM. Recent developments in synovial joint biomechanics. SIAMReview 1980;22(3):275–317.

25] Forster H, Fisher J, Dowson D, Wright V. The effect of stationary loading on thefriction and boundary lubrication of articular cartilage in the mixed lubricationregime. In: Dowson D, Taylor CM, Childs THC, Dalmaz G, editors. Proceedingsof the 21st Leeds/Lyon Symposium on Tribology, Lubrication and Lubricants,30. Amsterdam: Elsevier Science; 1995. p. 71–84.

26] Ateshian GA, Wang HQ, Lai WM. The role of interstitial fluid pressurizationand surface porosities on the boundary friction of articular cartilage. Journal ofTribology-Transactions of the ASME 1998;120(2):241–8.

27] Macirowski T, Tepic S, Mann RW. Cartilage stresses in the human hip joint. Jour-

nal of Biomechanical Engineering-Transactions of the ASME 1994;116(1):10–8.

28] Persson BNJ. Sliding friction. second ed. Berlin: Springer-Verlag; 2000. p. 515.29] McCutchen CW. The frictional properties of animal joints. Wear

1962;5(1):1–17.30] Mow VC, Ratcliffe A, Poole AR. Cartilage and diarthrodial joints as paradigms for

hierarchical materials and structures (review). Biomaterials 1992;13(2):67–97.

gineer

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

S.S. Pawaskar et al. / Medical En

31] Forster H, Fisher J. The influence of loading time and lubricant on the frictionof articular cartilage. Proceedings of Institution of Mechanical Engineers, PartH: Journal of Engineering in Medicine 1996;210(2):109–19.

32] Ateshian GA. A theoretical formulation for boundary friction in articular carti-lage. Journal of Biomechanical Engineering 1997;119(1):81–6.

33] Ateshian GA, Wang H. Rolling resistance of articular cartilage due to intersti-tial fluid flow. Proceedings of the Institution of Mechanical Engineers, Part H:Journal of Engineering in Medicine 1997;211:419–24.

34] Forster H, Fisher J. The influence of continuous sliding and subsequent sur-face wear on the friction of articular cartilage. Proceedings of Institution ofMechanical Engineers, Part H 1999;213:329–45.

35] Katta J, Pawaskar SS, Jin ZM, Ingham E, Fisher J. Effect of load variationon the friction properties of articular cartilage. Proceedings of Institution ofMechanical Engineers: Part J: Journal of Engineering Tribology 2007;221(J3):175–81.

36] Ateshian GA. The role of interstitial fluid pressurization in articular cartilagelubrication. Journal of Biomechanics 2009;42(9):1163–76.

37] Soltz MA, Ateshian GA. Experimental verification and theoretical prediction ofcartilage interstitial fluid pressurization at an impermeable contact interfacein confined compression. Journal of Biomechanics 1998;31(10):927–34.

38] Park S, Krishnan R, Nicoll SB, Ateshian GA. Cartilage interstitial fluid load sup-port in unconfined compression. Journal of Biomechanics 2003;36:1785–96.

39] Krishnan R, Kopacz M, Ateshian GA. Experimental verification of the role ofinterstitial fluid pressurization in cartilage lubrication. Journal of OrthopaedicResearch 2004;22(3):565–70.

40] Visible Human Project Full pelvis. National Library of Medicine–National Insti-tutes of Health. http://www.nlm.nih.gov/research/visible/visible human.html(accessed on February 2005).

41] Xi J, Hu X, Jin Y. Shape analysis and parameterized modeling of hip joint. Journalof Computing and Information Science in Engineering 2003;3(3):260–5.

42] Quinonez AF, Fisher J, Jin ZM. A steady-state elastohydrodynamic lubricationmodel aimed at natural hip joints with physiological loading and anatomicalposition. Proceedings of the Institution of Mechanical Engineers, Part J: Journalof Engineering Tribology 2008;222(3):503–12.

43] Anderson AE, Peters CL, Tuttle BD, Weiss JA. Subject-specific finite elementmodel of the pelvis: development, validation and sensitivity studies. Journal ofBiomechanical Engineering 2005;127(3):364–73.

44] Athanasiou KA, Agarwal A, Dzida FJ. Comparative study of the intrinsic mechan-ical properties of the human acetabular and femoral head cartilage. Journal ofOrthopaedic Research 1994;12(3):340–9.

45] Armstrong CG, Mow VC. Variations in the intrinsic mechanical properties ofhuman articular cartilage with age, degeneration, and water content. Journalof Bone and Joint Surgery American Volume 1982;64(1):88–94.

46] ABAQUS. Manuals, Version 6.7-1. Dassault Systemes., 2007, Suresnes Cedex,France.

47] Pawaskar SS, Fisher J, Jin Z. Robust and general method for determining sur-face fluid flow boundary conditions in articular cartilage contact mechanicsmodeling. Journal of Biomechanical Engineering 2010;132(3). p. 8 (031001).

48] Holmes MH. Finite deformation of soft-tissue - analysis of a mixture model inuniaxial compression. Journal of Biomechanical Engineering-Transactions ofthe ASME 1986;108(4):372–81.

49] Wu JZ, Herzog W. Finite element simulation of location- and time-dependentmechanical behavior of chondrocytes in unconfined compression tests. Annalsof Biomedical Engineering 2000;28(3):318–30.

50] Holmes MH, Mow VC. The nonlinear characteristics of soft gels andhydrated connective tissues in ultrafiltration. Journal of Biomechanics1990;23(11):1145–56.

51] Daniel M, Iglic A, Kralj-Iglic V. Hip contact stress during normal and staircasewalking: the influence of acetabular anteversion angle and lateral coverage of

the acetabulum. Journal of Applied Biomechanics 2008;24(1):88–93.

52] Pawaskar SS, Jin ZM, Fisher J. Modelling of fluid support inside articular carti-lage during sliding. Proceedings of Institution of Mechanical Engineers: Part J:Journal of Engineering Tribology 2007;221(J3):165–74.

53] Caligaris M, Ateshian GA. Effects of sustained interstitial fluid pres-surization under migrating contact area, and boundary lubrication by

[

ing & Physics 33 (2011) 96–105 105

synovial fluid, on cartilage friction. Osteoarthritis and Cartilage 2008;16(10):1220–7.

54] Ferguson SJ, Bryant JT, Ganz R, Ito K. The influence of the acetabular labrum onhip joint cartilage consolidation: a poroelastic finite element model. Journal ofBiomechanics 2000;33(8):953–60.

55] Harris WH, Rushfeldt PD, Carlson DE, Scholler J-M, Mann RW. Pressure distri-bution in the hip and selection of hemiarthroplasty. In: The Hip: Proceedingsof the Third Open Scientific Meeting of the Hip Society. 1975, , C. V. Mosby, St.Louis, MO, USA, 3.

56] Yamagata M, Chao EY, Ilstrup DM, Melton III LJ, Coventry MB, StaufferRN. Fixed-head and bipolar hip endoprostheses: a retrospective clinical androentgenographic study. The Journal of Arthroplasty 1987;2(4):327–41.

57] Kosashvili Y, Backstein D, Safir O, Ran Y, Loebenberg MI, Ziv YB. Hemiarthro-plasty of the hip for fracture – what is the appropriate sized femoral head.Injury 2008;39(2):232–7.

58] McDevitt C, Muir H. Biochemical changes in the cartilage of the knee in experi-mental and natural osteoarthritis in the dog. Journal of Bone and Joint SurgeryBritish Volume 1976;58-B(1):94–101.

59] Wachtl SW, Jakob RP, Gautier E. Ten-year patient and prosthesis survival afterunipolar hip hemiarthroplasty in female patients over 70 years old. The Journalof Arthroplasty 2003;18(5):587–91.

60] Haidukewych GJ, Israel TA, Berry DJ. Long-term survivorship of cemented bipo-lar hemiarthroplasty for fracture of the femoral neck. Clinical Orthopaedics andRelated Research 2002;(403):118–26.

61] Oberlander W. The stress of the human hip joint. Vii. The distribution of carti-lage thickness in the acetabulum and its functional explanation. Anatomy andEmbryology 1977;150(2):141–53.

62] von Eisenhart R, Adam C, Steinlechner M, Müller-Gerbl M, Eckstein F. Quan-titative determination of joint incongruity and pressure distribution duringsimulated gait and cartilage thickness in the human hip joint. Journal ofOrthopaedic Research 1999;17(4):532–9.

63] Krebs DE, Elbaum L, Riley PO, Hodge WA, Mann RW. Exercise and gait effectson in vivo hip contact pressures. Physical Therapy 1991;71(4):301–9.

64] Genda E, Konishi N, Hasegawa Y, Miura T. A computer-simulation study ofnormal and abnormal hip-joint contact pressure. Archives of Orthopaedic andTrauma Surgery 1995;114(4):202–6.

65] Hipp JA, Sugano N, Millis MB, Murphy SB. Planning acetabular redirectionosteotomies based on joint contact pressures. Clinical Orthopaedics and RelatedResearch 1999;(364):134–43.

66] Ipavec M, Brand RA, Pedersen DR, Mavcic B, Kralj-Iglic V, Iglic A. Mathe-matical modelling of stress in the hip during gait. Journal of Biomechanics1999;32(11):1229–35.

67] Anderson AE, Ellis BJ, Maas SA, Weiss JA. Effects of idealized joint geometryon finite element predictions of cartilage contact stresses in the hip. Journal ofBiomechanics 2010;43(7):1351–7.

68] Udofia I, Liu F, Jin Z, Roberts P, Grigoris P. The initial stability and contactmechanics of a press-fit resurfacing arthroplasty of the hip. Journal of Boneand Joint Surgery British Volume 2007;89-B(4):549–56.

69] McCann L, Ingham E, Jin Z, Fisher J. An investigation of the effect of con-formity of knee hemiarthroplasty designs on contact stress, friction anddegeneration of articular cartilage: a tribological study. Journal of Biomechanics2009;42(9):1326–31.

70] Pawaskar, S. S. Joint contact modelling of articular cartilage in synovial joints.PhD Thesis. Institute of Biological and Medical Engineering, School of Mechan-ical Engineering, University of Leeds; 2010. p. 223.

71] Yang, T. A three-dimensional biphasic finite element formulation for hydratedsoft tissue. PhD Thesis, Mechanical, Aeronautical and Nuclear Engineering.Rensselaer Polytechnic Institute; 2003. p. 164.

72] Dalstra M, Huiskes R, Vanerning L. Development and validation of a three-

dimensional finite element model of the pelvic bone. Journal of BiomechanicalEngineering-Transactions of the ASME 1995;117(3):272–8.

73] Liu F, Udofia I, Jin Z, Hirt F, Rieker C, Roberts P, et al. Comparison of contactmechanics between a total hip replacement and a hip resurfacing with a metal-on-metal articulation. Proceedings of the Institution of Mechanical Engineers,Part C: Journal of Mechanical Engineering Science 2005;219(7):727–32.