231

Analysis of contact mechanics in ceramic-on-ceramic

hip joint replacements

M M Mak and Z M Jin*

Department of Mechanical and Medical Engineering, University of Bradford, Bradford, UK

Abstract: The contact mechanics in ceramic-on-ceramic hip implants has been analysed in this study
using the
nite element method. Only the ideal conditions where the contact occurs within the
acetabular cup were considered. It has been shown that the contact pressure distribution and the
contact area at the main articulating bearing surfaces depend largely on design parameters such as
the radial clearance between the femoral head and the acetabular cup, as well as the thickness of the
ceramic insert. For the ceramic-on-ceramic hip implants used in clinics today, with a minimum 5-mm-
thick ceramic insert, it has been shown that the radius of the contact area between the femoral head
and the acetabular cup is relatively small compared with that of the femoral head and the ceramic
insert thickness. Consequently, Hertz contact theory can be used to estimate the contact parameters
such as the maximum contact pressure and the contact area.

Keywords: contact mechanics, ceramic-on-ceramic hip implant,
nite element method

NOTATION being highly wettable, hard, smooth and inert [1]. The
alumina-on-alumina bearing couple was introduced in
a contact radius 1970 [2], while zirconia-on-zirconia or zirconia-on-
c radial clearance, R
R alumina combinations have recently received signi
cant
2 1 attention [3, 4]. The alumina-on-alumina combination
E modulus of elasticity for the ceramic bearing
surface has been shown to produce extremely low friction [5, 6 ]
p pressure and low wear in arti
cial hip joint replacements under
p maximum contact pressure normal conditions from both clinical [2] and simulator
0 studies [7, 8]. Furthermore, it has been shown that the
R equivalent radius
R femoral head radius ceramic wear particles can also cause osteolysis, but
1 cause less systemic eVects than the metallic wear particles
R acetabular socket radius
2 [9, 10]. Therefore, minimizing wear and the number of
w applied load in the vertical direction
x, y horizontal and vertical coordinates respectively ceramic wear debris particles in ceramic-on-ceramic hip
implants is critical. Extensive research and testing have
î Poisson’s ratio for the ceramic bearing surface been conducted in the literature to demonstrate the
intrinsically low wear and abrasion resistance of ceramic
materials.
1 INTRODUCTION In the meantime, a number of
nite element studies
have been carried out to examine the stresses with the
Long-term concerns over ultra-high molecular weight ceramic components in order to reduce the likelihood of
polyethylene ( UHMWPE ) wear debris and the resulting fracture failure, in particular the
xation of the ceramic
osteolysis have promoted renewed interest in alternative femoral head to the metallic stem employing a taper
bearing couples for total hip implants. Medical grade connection [11]. The importance of the design param-
ceramics such as alumina and zirconia have been used eters for the bearing surfaces such as the radial clearance
in ceramic-on-ceramic hip implants because of their between the femoral head and the acetabular cup, as
attractive tribological and biological features such as well as the bearing surface construct, has not been
adequately addressed. For example, matched pairs with
The MS was received on 19 October 2001 and was accepted after revision clearances of 7–10 ím were used from 1977 to 1993
for publication on 28 March 2002.
* Corresponding author: Department of Mechanical and Medical [12, 13]. However, relatively large clearances have been
Engineering, University of Bradford, Bradford BD7 1DP, UK. introduced recently in order to reduce manufacturing
H06201 © IMechE 2002 Proc Instn Mech Engrs Vol 216 Part H: J Engineering in Medicine

Downloaded from pih.sagepub.com at HOWARD UNIV UNDERGRAD LIBRARY on March 9, 2015

232 M M MAK AND Z M JIN

costs and to allow unmatched components to be used,

for example, radial clearances between 5 and 35 ím were
speci
ed by Kyocera [14] and a nominal radial clearance
of 40 ím by CeramTec (between 15 and 35 ím [7]). It
is well-known that these geometric parameters, together
with the mechanical properties of the ceramic materials,
also inuence the tribological conditions at the bearing
surfaces and hence wear. This is particularly important
if new ceramic materials are introduced in the future
with diVerent mechanical properties [15].
Contact mechanics also plays an important role in the
wear of the bearing components under adverse con-
ditions, such as the edge contact between the femoral
head and the cup, resulting from excessive tilting of the Fig. 1 A simple schematic model for contact mechanics
acetabular cup and microseparation between the femoral analysis of a ceramic-on-ceramic hip prosthesis
head and the acetabular cup [16 ]. The consequence of
these adverse conditions is the generation of the stripe
wear observed on both clinically retrieved and simulator-
tested femoral heads [16 ]. The purpose of this study was
to investigate the contact mechanics in ceramic-on-
ceramic bearing couples for hip joint replacements using
the
nite element method. The simpli
ed ball-on-plane
Hertzian contact theory was also used and the corre-
sponding results were compared with those obtained
from the
nite element method for a more realistic ball-
in-socket geometrical con
guration. Only ideal con-
ditions where the contact occurs within the acetabular
cup were considered; the eVect of adverse conditions will
be considered in a further study.

2 MATERIALS AND METHODS

The hip prosthesis considered in this study consisted of

an alumina ceramic femoral head, articulating with an
alumina ceramic bearing insert attached to a polyethyl-
ene backing. The polyethylene backing was then secured
in a metallic shell for the purpose of
xation. The radius
of the femoral head was 14 mm. The nominal radial
clearance between the femoral head and the acetabular
cup was assumed to be 0.04 mm. The total thickness of
the ceramic bearing insert and the polyethylene backing
was
xed at 10 mm, with nominal values of 5 mm for
each component. Both the radial clearance and the thick- Fig. 2 Axisymmetric
nite element model for contact
ness of the ceramic insert were varied for the contact mechanics analysis of a ceramic-on-ceramic hip
mechanics analysis. Figure 1 shows a schematic diagram prosthesis
of the ceramic-on-ceramic hip prosthesis considered in
this study. ling because of symmetry, and consequently all the nodes
A simple axisymmetric
nite element model was devel- on the axis of the symmetry were constrained in the
oped to simulate the contact between the two articulating horizontal direction (x). Both bonded and unbonded
surfaces as shown in Fig. 2. The metallic
xation shell conditions were assumed at the interface between the
was neglected in Fig. 1 and also in the
nite element ceramic insert and the polyethylene backing, as well as
model, since the elastic modulus for the metallic material between the ceramic femoral head and the metallic stem.
is at least two orders of magnitude of that of polyethyl- In the unbonded case, coeYcients of friction of 0.1 and
ene, and consequently the outside of the polyethylene 0.2 were assumed at the ceramic/UHMWPE and
backing was fully constrained. Only half a cross- ceramic/metal interfaces respectively, while the contact
sectional area was required for the
nite element model- at the main bearing surfaces was assumed to be fric-
Proc Instn Mech Engrs Vol 216 Part H: J Engineering in Medicine H06201 © IMechE 2002

Downloaded from pih.sagepub.com at HOWARD UNIV UNDERGRAD LIBRARY on March 9, 2015

 CONTACT MECHANICS IN CERAMIC-ON-CERAMIC HIP JOINT REPLACEMENTS
tionless to represent the well-lubricated condition.
Bilinear axisymmetric
nite elements with four nodes
were used to mesh all the components. A
xed load of
2500 N was applied through the pole of both the cup
and the head. The elastic modulus and Poisson’s ratio
were chosen to be 380 GPa and 0.26, 1 GPa and 0.4 and
210 GPa and 0.3 for alumina, the polyethylene backing
and the metallic stem respectively. The
nite element
model thus formulated was solved using ABAQUS
( Version 5.8–9). Both the contact area and the contact
stresses were readily obtained from the
nite element
model. The Hertz contact theory, based on the assump-
tion of semi-in
nite solids for the ceramic bearing sur-
faces, was also used to predict the contact pressure
distribution, p, and the contact radius, a. This was done
by means of an equivalent ball-on-plane model using the
equivalent radius R determined from the radii of the
femoral head and the cup, R and R , and the radial
1 2
clearance, c:
R R
R= 1 2
c parameters at the articulating surfaces. Therefore, only
This enabled the contact radius, a, and the maximum
contact pressure, p , to be determined under a given
0 Figure 4a shows the predicted contact pressure distri-
load, w:
[3wR(1
î2)] 1/3
G H
a=
2E
3w
p =
0 2ða2
The corresponding contact pressure distribution was
given as follows:
C ABD
x 2 1/2
p=p 1

0 a
For the purpose of comparison with the ball-in-socket
geometry, the horizontal axis, x, and the contact radius,
Fig. 3 De
nition of angular coordinates for the ball-in-socket con
guration
H06201 © IMechE 2002
Downloaded from pih.sagepub.com at HOWARD UNIV UNDERGRAD LIBRARY on March 9, 2015
233
a, were transformed into the angular coordinate and the
contact half-angle, using the femoral head radius R as
1
shown Fig. 3.
3 RESULTS
The eVect of mesh sensitivity was
rstly examined on
the predicted contact parameters at the bearing surface.
For the nominal case, with a ceramic insert and an
UHMWPE inlay, both of 5 mm thickness, it was found
that 36 and 8 elements, in the circumferential and the
radial directions respectively, for each component were
adequate. This also resulted in a satisfactory element
aspect ratio of about unity for the majority of the
elements, as shown in Fig. 2. The boundary conditions
at the interface between the UHMWPE inlay and the
ceramic insert, as well as at the taper connection between
the ceramic femoral head and the metallic stem, were
found to have a negligible eVect on the predicted contact
(1)
the results for the fully bonded condition at these
interfaces are presented in this study.
bution at the bearing surfaces for the nominal con-
ditions, with a ceramic insert and an UHMWPE inlay,
(2) both of 5 mm thickness, and a radial clearance of 40 ím.
The contact pressure distribution predicted from Hertz
contact theory is also superimposed for the purpose of
(3)
comparison. Figures 4b and c show the contour plot of
the von Mises and the larger principal stresses within
both the ceramic and the UHMWPE components.
The eVect of varying the radial clearance between 20
and 80 ím on the predicted contact pressure distribution
(4) is shown in Fig. 5 for a given ceramic insert and an
UHMWPE inlay, both of 5 mm thickness. Figure 6
shows the eVect of varying the thickness of the ceramic
insert for a
xed radial clearance of 40 ím, while the
Proc Instn Mech Engrs Vol 216 Part H: J Engineering in Medicine
234
Fig. 4a Comparison of the predicted contact pressure
distribution between the present
nite element model
and the Hertz contact theory (ceramic insert thick-
ness, 5 mm; UHMWPE inlay thickness, 5 mm; radial
clearance, 40 ím)
Fig. 4b Prediction of the von Mises stress distribution within
various components (ceramic insert thickness, 5 mm;
UHMWPE inlay thickness, 5 mm; radial clearance,
40 ím)
total thickness of the acetabular cup including the cer-
amic insert and the polyethylene backing was
xed at
10 mm. Table 1 compares the predicted maximum con-
tact pressure and the contact half-angle from the present

nite element model with the Hertz contact theory.
4 DISCUSSION
It is clear from the results presented in Section 3 that
the contact pressure distributions depend largely on the
Proc Instn Mech Engrs Vol 216 Part H: J Engineering in Medicine
Downloaded from pih.sagepub.com at HOWARD UNIV UNDERGRAD LIBRARY on March 9, 2015
M M MAK AND Z M JIN
Fig. 4c Prediction of the larger principal stress within various
components (ceramic insert thickness, 5 mm;
UHMWPE inlay thickness, 5 mm; radial clearance,
40 ím)
Fig. 5 EVect of the radial clearance on the predicted contact
pressure distribution (ceramic insert thickness, 5 mm;
UHMWPE inlay thickness, 5 mm)
design parameters such as the radial clearance between
the femoral head and the acetabular cup, and the thick-
ness of the acetabular insert. For the nominal conditions
considered, the predicted contact pressure distribution is
only slightly less than that from the corresponding Hertz
contact theory as shown in Fig. 4a; the diVerence in the
predicted maximum value being less than 15 per cent.
This is expected since the contact half-angle predicted
from the
nite element model is approximately 17.5°,
corresponding to a contact radius of 4 mm, which is less
than the thickness of the ceramic insert of 5 mm. This
H06201 © IMechE 2002
 CONTACT MECHANICS IN CERAMIC-ON-CERAMIC HIP JOINT REPLACEMENTS 235

signi
cant eVect on the predicted contact pressure distri-

bution, as shown in Fig. 6, particularly when the thick-
ness of the ceramic insert is reduced to below 5 mm. For
example, for a ceramic insert thickness of 2.5 mm, a large
diVerence is observed between the present
nite element
model and the Hertz contact theory, as shown in Fig. 6.
Although the contact stress is signi
cantly reduced at
the main bearing surfaces, the tensile stress on the out-
side of the ceramic insert is signi
cantly increased to
96.1 MPa; this may increase the likelihood of the ceramic
insert fracturing.
It is clear from the present
nite element model
that under nominal conditions currently employed in
ceramic-on-ceramic hip implants, the ceramic insert
Fig. 6 EVect of the ceramic insert thickness on the predicted essentially behaves like a semi-in
nite solid and the
contact pressure distribution (total thickness of both thickness of the insert above 5 mm has a small eVect on
ceramic insert and UHMWPE inlay, 10 mm; radial the predicted contact parameters, such as the maximum
clearance, 40 ím) contact pressure and the contact area, as shown in
Table 1. Consequently, Hertz contact theory can be used
to predict these contact parameters. This also implies
Table 1 Comparison of the predicted maximum contact press- that the uid
lm lubrication, which is not considered
ure and the contact half-angle between the present
in this study, can be analysed by the use of the Hamrock

nite element study and the Hertz contact theory
and Dowson formula [18], as carried out by Jin et al.
Contact half-angle Maximum contact [19] and Scholes and Unsworth [5].
Ceramic (deg) pressure (MPa) Only the sandwiched ceramic insert has been con-
Radial insert
clearance thickness Finite Finite
sidered in this study. However, the conclusions obtained
(ím) (mm) element Hertz element Hertz can also be expected to be applicable to the taper-
connected ceramic insert provided that the insert thick-
40 2.5 20 14.6 40 94.1
40 5.0 17.5 14.6 80 94.1
ness is above 5 mm. Contact mechanics should be a key
40 7.5 15 14.6 90 94.1 design consideration when new ceramic materials are
20 5 22.5 18.4 50 59.3 introduced for ceramic-on-ceramic hip implants [15].
80 5 12.5 11.6 140 149.0
The diVerence in the mechanical properties of these cer-
amic materials from alumina means that the optimum
geometrical parameters designed for the alumina-on-
is also reected in Fig. 4b where the maximum von Mises alumina, such as the radial clearance, may have to be
stresses can be seen to be located below the surface and modi
ed.
within the ceramic components. It is also interesting to The main assumption adopted in the present study is
note that a small tensile principal stress is developed on the two-dimensional axisymmetric model where the load
the outside of the ceramic insert mainly due to bending, goes through the pole in the vertical direction. In reality,
but the magnitude of 49.3 MPa is very small. Meanwhile the acetabular cup is generally inserted with a 45° incli-
on the bearing surfaces, the principal stress is mainly nation, under a load approximately 10–16° medially
compressive, with a magnitude of 70.9 MPa. A small from the vertical direction [20]. The contact half-angle
tensile stress can be developed on the bearing surface predicted from the present
nite element model is less
near the edge of the contact area in the classical Hertz than 25° for all the cases considered, as compared with
contact [17] and its magnitude can be estimated as at least 55° available for contact within the acetabular
1/3(1
2î)p =15 MPa for this particular example. cup. Therefore the assumption of the axisymmetric
0
It is generally noted from Fig. 5 that a decrease in the model is reasonable, since the contact occurs well within
radial clearance results in an increase in the predicted the acetabular cup. In addition to these ideal contact
contact area, and consequently a decrease in the maxi- conditions, contact mechanics can be expected to play a
mum contact pressure. It is interesting to note that as more important role under adverse conditions such as
the radial clearance is decreased to 20 ím, the contact the edge contact between the acetabular cup and the
half-angle extends to beyond 20°, and yet the diVerence femoral head. To do this a three-dimensional
nite
in the predicted maximum contact pressure between the element model will be required, rather than the two-
present
nite element model and the Hertz contact dimensional axisymmetric model considered in this
theory is still about 15 per cent, as shown in Table 1. study. The eVect of microseparation and the edge con-
Variation of the thickness of the ceramic insert, while tact, as well as validating the simple axisymmetric model,
keeping the total thickness
xed at 10 mm, also has a will be considered in a further study.
H06201 © IMechE 2002 Proc Instn Mech Engrs Vol 216 Part H: J Engineering in Medicine

Downloaded from pih.sagepub.com at HOWARD UNIV UNDERGRAD LIBRARY on March 9, 2015

236
5 CONCLUSIONS
The contact mechanics under ideal conditions where the
contact occurs within the acetabular cup has been ana-
lysed for ceramic-on-ceramic hip prostheses using the

nite element method. It has been shown that both the
radial clearance between the femoral head and the acet-
abular cup and the thickness of the acetabular ceramic
insert can have large eVects on predicted contact param-
eters such as the contact stresses and the contact area.
For ceramic-on-ceramic hip implants currently used in
clinical practice, Hertz contact theory can be used to
estimate these contact parameters. Optimizing the radial
clearance and the ceramic insert thickness in terms of
contact mechanics may oVer further opportunities for
improving the tribological performance of these hard-
on-hard hip implants employing new ceramic materials.
ACKNOWLEDGEMENT
This work was supported by the Engineering and
Physical Sciences Research Council ( EPSRC ), UK.
REFERENCES
1 Garino, J. P. and Sedel, L. Editorial comment. Clin.
Orthop. Related Res., 2000, 379, 2.
2 Sedel, L., Kerboull, L., Christel, P., Meunier, A. and
Witvoet, J. Alumina-on-alumina hip replacement. Results
and survivorship in young patients. J. Bone Jt Surg., 1990,
72-4(4), 658–663.
3 Cales, B., Blaise, L. and Villermaux, F. Improved bearing
systems for THP using zirconia ceramic. In Biomaterials
and Bioengineering Handbook (Ed. D. L. Wise), 2000
(Marcel Dekker, New York), pp. 483–508.
4 Clarke, I. C., Good, V., Williams, P., Schroeder, D.,
Anissian, L., Stark, A., Oonishi, H., Schuldies, J. and
Gustafson, G. Ultra-low wear rates for rigid-on-rigid bear-
ings in total hip replacements. Proc. Instn Mech. Engrs,
Part H, Journal of Engineering in Medicine, 2000,
214(H4), 331–347.
5 Scholes, S. C. and Unsworth, A. A comparison of friction
and lubrication of diVerent hip prostheses. Proc. Instn
Mech. Engrs, Part H, Journal of Engineering in Medicine,
2000, 214(H1), 49–57.
6 Scholes, S. C., Unsworth, A. and Goldsmith, A. A. A fric-
tional study of total hip joint replacements. Physics Med.
Biol., 2000, 45(12), 3721–3735.
Proc Instn Mech Engrs Vol 216 Part H: J Engineering in Medicine
Downloaded from pih.sagepub.com at HOWARD UNIV UNDERGRAD LIBRARY on March 9, 2015
M M MAK AND Z M JIN
7 Saikko, V. and PfaV, H. G. Low wear friction in alumina/
alumina total hip joints: a hip simulator study. Acta
Orthopaedica Scand., 1998, 69(5), 443–448.
8 Nevelos, J. E., Ingham, E., Doyle, C., Nevelos, A. B. and
Fisher, J. Wear of HIPed and non-HIPed alumina-alumina
hip joints under standard and severe simulator testing
conditions. Biomaterials, 2001, 22(16), 2191–2197.
9 Fisher, J., Stone, M. H. and Ingham, E. Functional biologi-
cal activity of wear debris generated in arti
cial hip joints.
In 2nd World Tribology Congress, Vienna, 2001, p. 740.
10 Ingham, E. and Fisher, J. Biological reactions to wear debris
in total joint replacement. Proc. Instn Mech. Engrs, Part H,
Journal of Engineering in Medicine, 2000, 214(H1), 21–37.
11 Middleton, J., Pande, G. N., Richter, H. G. and
Willmann, G. Investigations into the mechanical reliability
of ceramic femoral heads for hip joints. J. Mater. Sci.:
Mater. Medicine, 1994, 5, 503–506.
12 Nizard, R. S., Sedel, L., Christel, P., Meunier, A.,
Soudry, M. and Witvoet, J. Ten-year survivorship of
cemented ceramic–ceramic total hip prosthesis. Clin.
Orthop. Related Res., 1992, 282, 53–63.
13 Sedel, L. Evolution of alumina-on-alumina implants: a
review. Clin. Orthop. Related Res., 2000, 379, 48–54.
14 Ueno, A., Amino, H., Okimatu, H. and Oonishi, H. Wear,
friction, and mechanical investigation and development of
alumina-to-alumina combination total hip joint. In Joint
Arthroplasty (Eds S. Imura, M. Wada and H. Omori), 1999,
pp. 119–131 (Springer-Verlag, Tokyo).
15 AVatato, S., Goldoni, M., Testoni, M. and Toni, A. Mixed
oxides prosthetic ceramic ball heads. Part 3: eVect of the
ZrO fraction on the wear of ceramic on ceramic hip joint
2
prostheses. A long-term in vitro wear study. Biomaterials,
2001, 22, 717–723.
16 Nevelos, J., Ingham, E., Doyle, C., Streicher, R., Nevelos, A.
and Fisher, J. Micro-separation in vitro produces clinically
relevant wear of ceramic–ceramic total hip replacements.
Key Engng Mater., 2001, 192–195, 529–532.
17 Johnson, K. L. Contact Mechanics, 1987 (Cambridge
University Press).
18 Hamrock, B. J. and Dowson, D. Elastohydrodynamic lubri-
cation of elliptical contacts for materials of low elastic
modulus. I: fully ooded conjunction. Trans. ASME,
J. Lubric. Technol., 1978, 100(2), 236–245.
19 Jin, Z. M., Downson, D. and Fisher, J. Analysis of uid

lm lubrication in arti
cial hip joint replacements with
surfaces of high elastic modulus. Proc. Instn Mech. Engrs,
Part H, Journal of Engineering in Medicine, 1997,
211(H3), 247–256.
20 Bergmann, G., Deuretzbacher, G., Heller, M., Graichen, F.,
Rohlmann, A., Strauss, J. and Duda, G. N. Hip contact
forces and gait patterns from routine activities.
J. Biomechanics, 2001, 34(7), 859–871.
H06201 © IMechE 2002