pyrig

No Co

ht
t fo
rP

by N
ub

Q ui
Influence of Preparation Design on Failure Risks of lication
te ot

n
Ceramic Inlays: A Finite Element Analysis ss e n c e fo r

Masahiro Onaa/Chie Watanabeb/Yoshimasa Igarashic/Noriyuki Wakabayashid

Purpose: To assess the influence of preparation design on failure risk of ceramic inlays by means of finite ele-
ment (FE) analysis.
Materials and Methods: Both 2D and 3D models of the maxillary first premolar were constructed. MOD ceramic
inlays were designed with an isthmus width of 1.5 to 5.0 mm and a minimum thickness of 1 to 3 mm. Occlusal
contact motion and debonding at the tooth/ceramic adhesive interface were simulated in 2D models by down-
ward displacement of a ceramic ball (1.5 mm in diameter) onto the central groove. The stress distributions
within the inlays and at the interface were analyzed using 3D models. A maximum occlusal load of 250 N was
assumed.
Results: The maximum principal stress at the base of the inlays was constant among models of different
isthmus widths, although it decreased considerably as inlay thickness increased from 1 mm to 2 mm. The
maximum shear stress at the adhesive interface increased as the width of the inlay decreased and the distance
between the occlusal contact and the margin decreased to 0.3 mm or less. The maximum principal stress on
the occlusal surface was relatively low and insensitive to inlay design; however, it was increased by a simulated
adhesive failure.
Conclusion: The failure risk at the base of the inlay is minimized by increasing the minimum inlay thickness. As
the occlusal contact becomes close to the margin, the adhesive failure risk is increased, potentially leading to
an increase in risk of fracture on the occlusal surface.
Keywords: ceramic inlay, preparation design, failure risk, finite element analysis.
J Adhes Dent 2011; 13: 367–373 Submitted for publication: 25.12.2009; accepted for publication: 09.04.2010.
doi: 10.3290/j.jad.a19473

A dhesively bonded computer-aided design/com-

puter-aided manufacturing (CAD/CAM)-generated
ceramic inlays and onlays have grown in popularity as
the technique has evolved and improved. Long-term
a Clinical Resident, Prosthodontics, Dental Hospital, Tokyo Medical and Dental
University, Tokyo, Japan. Experimental design, performed experiments, data
analysis.
b Dental Student, Tokyo Medical and Dental University, Tokyo, Japan. Per-
formed experiments.
c Professor, Department of Removable Partial Prosthodontics, Masticatory
Function Rehabilitation, Graduate School, Tokyo Medical and Dental Univer-
sity, Tokyo, Japan. Contributed substantially to discussion.
d
Associate Professor, Department of Removable Partial Prosthodontics, Mas-
ticatory Function Rehabilitation, Graduate School, Tokyo Medical and Dental
University, Tokyo, Japan. Idea, hypothesis, wrote manuscript.
Correspondence: Dr. Noriyuki Wakabayashi, Department of Remov-
able Partial Prosthodontics, Masticatory Function Rehabilitation, Gradu-
ate School, Tokyo Medical and Dental University, Tokyo, Japan. 1-5-45,
Yushima, Bunkyo, Tokyo, Japan 113-8549. Tel/Fax: +81-3-5803-4935.
e-mail: wakabayashi.rpro@tmd.ac.jp
prospective studies on CAD/CAM ceramic inlays and
onlays reported survival probabilities of over 89% at
10+ years,19,20,25 and clinical case series indicated
that the restorations were fairly predictable. 14,22,27
These studies demonstrated that ceramic fracture was
the primary mode of clinical failure of the restorations.
Despite the encouraging results, relatively low resis-
tance to fracture remains the major clinical reason for
complications of the ceramic inlays, especially in the
posterior region.10
Only limited information is available regarding the influ-
ence of preparation design on the fracture resistance of
CAD/CAM inlays made from feldspathic ceramic blocks.
Either the conventional retention form3,9 is used, or the
alternative form that is largely dependent on the adhesive
luting cement.28 For partial ceramic crowns, previous in
vitro studies found that a minimum ceramic thickness of
1.5 to 2.0 mm in stress-bearing areas decreased the risk
of ceramic fracture,7 and that coverage of thin cusp walls
decreased the likelihood of enamel cracks and marginal
deficiency.15 However, a standard preparation has not

Vol 13, No 4, 2011 367


Ona et al
Fig 1 Meshed 2D model of the maxillary first premolar and
loading ball. Simplified outline of the enamel and dentin areas
were derived from an image of a buccolingual section at the
mesiodistal center of the tooth. The model has an inlay with
isthmus width of 2 mm.
yet been established because fracture mechanisms are
not well understood in relation to actual modes of clinical
failure.
Without an understanding of the mechanisms of
fracture types, the results of in vitro testing of mate-
rial strength lack clinical validity.12 In general, ceramic
restorations fail due to material flexure and subsequent
propagation of cracks under cyclic loading.24,26 To as-
sess the effects of preparation design on fracture resist-
ance, static or fatigue load-to-fracture tests should be
conducted on adhesively bonded tooth-restoration speci-
mens with different preparation geometries. The predic-
tive power of such in vitro data could be increased by finite
element (FE) stress analysis, which would estimate the
stress distributions under loading.2,17 Understanding the
stress distribution in the ceramic and tooth structures
under simulated functional loading helps predict both the
likely location of initial failure and overall probability of
failure.18 Therefore, the objective of this preliminary study
is (1) to assess the stress distributions of bonded inlays,
and (2) to determine which aspects of preparation de-
sign influence the risk of fracture. Based on the intensity
and location of stress concentrations, the relative prob-
abilities of bulk failures, contact fractures, and adhesive
failures were investigated. The null hypothesis was that
the failure mode of an adhesively bonded ceramic inlay is
determined by its preparation design.
MATERIALS AND METHODS
Nonlinear Contact Analysis by Two-dimensional (2D) FE
Models
To evaluate the impact of occlusal contact on stress
distribution, a nonlinear contact analysis was conducted
using 2D models of the maxillary first premolar crown.
368
pyrig
No Co
ht
t fo
2D outline of the enamel and dentin areas rwas
Pu traced
by N
from an image of an anatomically average buccolingual
bli
Q ui
cat
section at the mesiodistal center of the tooth (Dental
Anatomy & Interactive 3-D Tooth Atlas, tBrown oti
& Her- on
n
e
nc e
ss edecision
branson Imaging; Portola Valley, CA, USA). The
fo r
to exclude the pulp chamber and root from models
was based on the preliminary test, in which the effects
of these factors were found to be negligible upon the
comparative evaluations of the stress distributions
within the inlay and at the adhesive interface. Cavity de-
signs were made with one of five buccolingual isthmus
widths: 1.5 mm, 2 mm, 3 mm, 4 mm, and 5 mm. Cavity
walls were assumed parallel, and minimum cavity depth
at the center of the groove was assumed to be 1.0 mm.
In addition, a model having a thin ceramic layer near the
margin was produced by modifying the 3.0-mm-width
model. An intact tooth model was also reserved for
stress calculation.
Occlusal contact motion was simulated by the down-
ward displacement of a ceramic ball, 1.5 mm in diam-
eter, onto the central groove of each model. The ball
was placed with a slight space between the ball and the
occlusal surfaces of the buccal and lingual cusps. Each
model was meshed by approximately 1200 plain stress
elements defined by 4 nodes having two degrees of free-
dom (Fig 1) (ANSYS 11.0, ANSYS; Canonsburg, PA, USA).
Surface-to-surface contact elements were applied to the
lower surface of the ball, and the target segment was
used on the lines of occlusal surface. Preliminary con-
vergence tests determined mesh size and arrangement
for solution consistency.1 All materials were considered
homogeneous, isotropic, and linearly elastic. The Young’s
moduli and the Poisson’s ratios, resp., were 84.1 GPa5
and 0.336 for enamel, 14.7 GPa23 and 0.316 for dentin,
and 70.0 GPa11 and 0.1911 for ceramic inlays. A perfect
bond was assumed at the tooth/ceramic interface. A fric-
tion coefficient of 0.5 was assumed and entered into the
program to simulate the contact between ball and inlays.
Downward force was applied to the ball surface until the
force reached 250 N, which has been reported as the
approximate maximum bite force seen in the maxillary
premolar tooth of the average adult.8 In each model, the
movement of the bottom line was restricted.
The principal stress and shear stress distributions
were calculated for all simulations. The following maxi-
mum stresses were coded: maximum principal stresses
on the occlusal surface of inlay (Ss) and on the bottom
of the inlay (Sb), and the maximum shear stress at the
interface between the inlay and the tooth structures (Si).
Next, contact elements were applied to the enamel/
ceramic interface on the lingual side of the model with
1.5-mm width to assess the effects of adhesion failure
on the stress distribution. The contact analysis was ex-
ecuted with this model in the same loading condition as
above. The least-width model was used for this simulation
of debonding at the interface because it revealed highest
Si in all the models with perfect bonding. The coefficient
of friction of 0.15 was assumed at the failed interface.
The Journal of Adhesive Dentistry
 pyri
No CoOna et gal

ht
t fo
rP

by N
Fig 2 Construction of the 3D FE model. ub

Q ui
The construction process: creation of the
lica
tio
cross-sectional areas (A), connection of the te ot n

n
dentin areas for volume (B), connection of ss e n c e
fo r
the enamel area for volume (C), and mesh-
ing (D). E: Volume model with the MOD inlay
A B C D E
(gray).

Principle
Fig 3 Principal stress and shear stress
tensile stress
distributions of inlays and adjacent tooth
structures in 2D models. The models with
isthmus widths of 2 mm and 4 mm, both
with minimum thickness of 1 mm, are
shown. Principal contours are shown in the
upper row, and shear contours are in the
lower row. The contours were divided into
nine parts with different colors according Shear stress
to the stress levels. Red areas represent
the highest principal or shear stress, as
indicated by the scale bars. In the shear
contours, the maximum (red) and minimum
(blue) shear stresses are symmetrically
indicated as they are created by the shear
forces of clockwise and counterclockwise 2mm-width 4mm-width
directions.

Linear Structural Analysis by Three-dimensional (3D) FE
Models
3D models of the maxillary first premolar, with realistic
configurations capable of representing a MOD inlay with
mesial and distal box forms, were created to conduct a
structural stress analysis. Based on the same anatomi-
cal data source as we used for the 2D models, 2D out-
lines of the enamel and dentin structures of the tooth
crown were created using eight horizontal sections. The
outlines in all sections were connected and further ex-
tended axially to create the 3D model (Fig 2). The isth-
mus widths were the same as for the 2D models. For
each width, models were created with minimum depths
of 1 mm, 2 mm, and 3 mm. The width of the gingival
floor of the proximal boxes was 0.8 mm. The height of
the gingival floor from the occlusal cavity floor was 1.5
mm for all models.
Each model was meshed by approximately 58,000
elements defined by 20 nodes having three degrees of
freedom in hexahedral bodies. Material properties were
the same as those for the 2D models. Contact simulation
for the occlusal stress was feasible with the 2D models;
however, it is highly nonlinear and difficult to perform in
an accurate and reliable manner in 3D models. The lin-
ear analyses of the 3D models required the application
of a loading force on a fixed circular loading area. The
disadvantage of this analysis is that setting a fixed area
unrealistically concentrates stress around the edge of the
loading circle, therefore, the stress distribution on the oc-
clusal surface (Ss) of the 3D models was not calculated.
A 250-N load perpendicular to the surface of the inlay was
evenly distributed over a circular area with a diameter of
0.5 mm adjacent to the central groove on the internal buc-
cal cusp at the mesial-distal center, while the movement
of the bottom surface was restricted.
Failure risks, Rb and Ri, were defined based on the
maximum stresses as follows; Rb = Sb/Flexural strength
of the ceramic, and Ri = Si/Shear bond strength of the ce-
ramic to enamel, where the flexural strength of 94 MPa4
and the shear bond strength, with a dual-curing resin
luting agent together with a silane-coupling agent, of 37
MPa30 for a feldspathic ceramic for CAD/CAM processing
(Vitablocs Mark II, Vita Zahnfabrik; Bad Säckingen, Ger-
many) were used for calculations.
RESULTS
2D Models
The principal and shear stress distributions in 2D mod-
els with inlay widths of 2 mm and 4 mm are shown in
Fig 3. In the 2-mm-wide model, the maximum principal
and shear stresses appeared at the interface between
the enamel and the ceramic. However, in the 4-mm-wide
model, the maximum principal stress was shown at the
base of the inlay (Sb). The maximum shear stress at

Vol 13, No 4, 2011 369

 pyrig
Ona et al
No Co

ht
t fo
rP

by N
(MPa) ub

Q ui
150 lica
Sb tio
te ot n

n
ss e n c e
fo r
100

50
Ss
Fig 4 Maximum stresses (Ss, Sb, and Si)
Si in 2D models. The sites of interest are the
0
maximum principal stresses on the occlusal
1.5mm 2mm 3mm 4mm 5mm surface of inlay (Ss) and on the bottom
of the inlay (Sb), and the maximum shear
Inlay width stress at the interface between the inlay
and the tooth structures (Si).

Principle
tensile stress

Fig 5 Principal stress distributions of the

2D thin ceramic layer model (A) and the
intact tooth model (B). Both models were
based on the model of 3-mm width and
1-mm depth. The scale bar shown is the
A B
same as in Fig 3.

Fig 6 Principal stress distributions of the

2D models before and after adhesive fail-
ure. The figures represent the lingual half
of the inlay (square area in left figure). (A)
Perfect bonding was assumed at the inter-
face between the ceramic and the tooth
structures (red line in left figure), and (B)
debonding was assumed at the same in-
A B terface. Both models used a 1.5-mm width
and 1-mm depth. The maximum stress on
the occlusal surface (Ss, red arrows) was
23.1 MPa for (A) and 53.3 MPa for (B).

the adhesive interface (Si) was considerably higher in
the 2-mm-wide than in the 4-mm-wide model. Figure 4
shows Ss, Sb, and Si as a function of the inlay width.
As the width became larger, Sb increased sharply while
Si decreased. Ss results were relatively constant, with
intermediate stress magnitudes across all widths. The
model with a thin ceramic layer near the margin (Fig 5A)
and the intact tooth model without cavity (Fig 5B) show
relatively low maximum principal stress. The adhesion
failure simulation showed the maximum principal stress
on the occlusal groove, with an Ss of 53.3 MPa (Fig
6B), as compared to the same model with the perfect
adhesive bonding with an Ss of 23.1 MPa (Fig 6A).
3D Models
The principal stress distributions at the base of the in-
lays of different minimum thicknesses are shown in Fig
7A. The maximum principal stress at the base of each
inlay (Sb) was found to be under the loading site. The
maximum shear stress at the interface (Si) was found
to be at the enamel region on the adhesive interface
(Fig 7B). Si exhibited high stresses with widths of 2 mm

370 The Journal of Adhesive Dentistry


A ss e n c e
B
Fig 7 Stress distributions of 3D models.
(A) The principal stress distributions at the
base of the inlays (width of 3 mm, thick-
ness of 1 mm to 3 mm, from left to right).
Red areas represent the highest principal
stress, as indicated by the scale bar of
each graphic. (B) The shear stress distribu-
tion of the tooth with the 1.5-mm-wide cav-
ity. The inlay was removed from the graphic
for viewing. The maximum (red) and mini-
mum (blue) shear stresses are symmetri-
cally indicated as they are created by shear
forces clockwise and counterclockwise be-
neath the loading location.
or less, and showed sharp decreases, close to zero, as
the width increased (Fig 8). The Sb was relatively con-
stant among models of different widths. High maximum
stresses, over 80 MPa, were calculated for Sb with an
inlay thickness of 1 mm, while stresses were reduced
considerably with thicknesses of 2 mm and 3 mm.
Failure risks Rb and Ri for all widths and thicknesses
are shown in Fig 9. Ri exhibited high scores, over 1.6, with
the smallest width of 1.5 mm, and it became lower than
0.3 with a width over 3 mm. Rb was insensitive to the
width, although it was higher with a 1-mm-thickness than
it was with larger thicknesses. Ri was higher at 1.5-mm
and 2-mm width than Rb was at any thickness. However,
if the width was 3 mm or over, Rb was higher than Ri with
1-mm thickness, and the two stresses were almost equal,
and very low, when the thickness was 2 mm and 3 mm.
DISCUSSION
The Sb is responsible for crack initiation at the base,
which may lead to bulk fracture of the inlay. The result
of the 3D models indicates that the Sb was reached
beneath the loading location and that an increase of
the minimum inlay thickness from 1 mm to 2 mm con-
siderably reduced the Sb. The results of the present
study coincide with the findings of previous in vitro stud-
ies which recommended that the preparations have a
minimum thickness of more than 1 mm in the region be-
neath the potential occlusal contact area.7,16 The result
of the 2D models indicates that the Sb increases as a
function of isthmus width. This was not consistent with
the 3D models, in which the Sb was relatively insensi-
Vol 13, No 4, 2011
pyri
No CoOna et gal
ht
t fo
rP
by N
ub
Q ui
lica
tio
te ot n
n
fo r
tive to the width. In the 3D models, the support offered
by the rigid enamel of the mesial and distal proximal
parts of the tooth absorbed the high stress within the
inlay. However, in the 2D models, the mesial and distal
proximal enamel structures were absent, as they were
created based on the section view through the mesial-
distal center of the tooth. The increased inlay width
considerably reduced the amount of enamel in the 2D
models, resulting in pronounced stress on the inlays.
Because of this limitation of the 2D models, the result
of the 3D models was mainly used to evaluate the fail-
ure risk analysis. Since the 2D models were still useful
for nonlinear contact analysis, they were employed spe-
cifically to assess the impact of occlusal contact on the
stress at the occlusal contact.
The initiation of cone cracks, due to the Hertzian inden-
tation of the ceramic compacts, were induced largely as
a result of the circular symmetry of the principal stress
distributions, and is associated with the maximum princi-
pal stress magnitude.31 The Ss values were relatively con-
stant with different widths, suggesting that the potential
risk of the contacting ceramic fracture was independent of
the preparation design. In addition, it was indicated that
the loading force could be transferred from the occlusal
surface to the inner enamel and dentin structures if the
ceramic was adhesively bonded to the structures.13 The
thin ceramic layer model demonstrated this adhesion
mechanism with its low stress concentration (Fig 5A).
The contact stress is largely dependent on the material
properties and the geometry of the cusps,21,29 and the
ceramics fail due to the subcritical fatigue damage at load
levels that are considerably lower than those required for
damage under single loading. However, the results of the
371

Ona et al
(MPa)
140
120
Sb T=1mm
100
Sb T=2mm
80 Sb T=3mm
60 Si T=1mm
Si T=2mm
40
Si T=3mm
20
0 0.0
1.5mm 2mm 3mm 4mm 5mm
Inlay width
Fig 8 Maximum stresses (Sb and Si) in 3D models. Sb: the
maximum principal stress at the base of the inlays. Si: the
maximum shear stress at the interface between the ceramic
and the tooth structures. T: thickness.
monotonic loading assumed in this study suggests that,
if the inlays are perfectly bonded to the tooth structures,
fracture at the contact surface may be a lesser failure risk
than the bulk fracture and the adhesive failures.
The maximum shear stress at the adhesive interface
increased as the inlay width decreased. The result was
associated with the reduced distance between the occlu-
sal loading contact and the inlay-tooth margin. The result
of the 2D contact simulation indicated that an isthmus
width of 2 mm or less resulted in a higher Si compared
to isthmus widths of 3 mm or more. The measured dis-
tance between the contact and the margin in the 3-mm-
width model was 0.8 mm, which may be the minimum
contact-to-margin distance that should be considered in
the preparation of ceramic inlays. The Si was greater
than the Rb at the smallest inlay width, whereas the Rb
was greater than the Si at inlay widths > 3 mm for cor-
responding inlay thicknesses (Fig 9). This suggests that
the highest failure risk was associated with the Ri when
the contact-to-margin distance was small, but it was as-
sociated with Rb when the distance was large. Therefore,
the results support the hypothesis that the failure mode
of an adhesively bonded ceramic inlay changes as the
preparation design changes.
Debonding, or adhesive failure, has been excluded
from the criteria for evaluation in case-series and long-
term prospective studies of CAD/CAM inlays, probably
because it is not possible to detect debonding failure
as a single factor without other related symptoms, such
as marginal discoloration and adaptation. However, our
contact analysis demonstrated an alarming example of
how debonding may cause fracture of the inlay (Fig 6).
The simulation of interfacial debonding resulted in a
dramatically increased maximum principal stress on the
occlusal surface near the central groove, unseen in the
372
pyrig
No Co
ht
t fo
4.0 rP
by N
ub
Q ui
lica
3.5
tio
te ot n
n
3.0
sRb
s eT=1mm
fo r
n c e
Failure Risks
2.5
Rb T=2mm
2.0 Rb T=3mm
1.5 Ri T=1mm
Ri T=2mm
1.0
Ri T=3mm
0.5
1.5mm 2mm 3mm 4mm 5mm
Inlay width
Fig 9 Failure risks (Rb and Ri) by inlay widths and thick-
nesses. Rb = Sb/Flexural strength of the ceramic, and Ri =
Si/Shear bond strength of the ceramic to enamel. T: thick-
ness.
perfectly bonded model. The failure at the interface may
lead to the separation of a ceramic layer that was pre-
viously part of the ceramic-enamel complex, derailing
the mechanism that transfers stress from ceramic to
enamel. The maximum principal stress at the central
fissure increased because of the wedging effect caused
by the loads on the right and left side of the V-shaped
ceramic surfaces that were no longer supported by the
bonded tooth structure, potentially leading to fracture
into fragments, or ceramic chipping. One long-term pro-
spective study20 of CAD/CAM inlays revealed that after
17 years, approximately 5% of inlays had fractured and
were replaced by a new restoration, while similar num-
bers of other failures were repaired simply with compos-
ite material or ceramic. Although fracture mode rates
were not documented in this study, it is reasonable to
assume that the former cases were comprised of bulk
fracture, while the latter were comprised of marginal
chipping or fracture into small fragments. Based on the
results of the current analysis, it is further hypothesized
that adhesive failure increases the probability of fracture
on the occlusal ceramic surface.
An intact tooth model without cavity preparation dem-
onstrated more evenly distributed stress fields than any
restored models; there were only slight increases of
stress at the enamel/dentin interface region (Fig 5B).
An intact natural tooth may possess a more conservative
structure that can resist greater functional occlusal load
than a restored tooth. The limitations of the FE models
used in this study should be considered, especially if the
quantitative assessment of the stresses is to be empha-
sized. Nevertheless, the results of this FE study serve
to clarify the critical elements of preparation design for
specimen constructions of future fracture tests of bonded
ceramic restorations and fixed partial dentures.
The Journal of Adhesive Dentistry

CONCLUSION
The results of this FE study suggest that the maximum
shear stress at the ceramic/tooth interface increases
as the occlusal contact becomes close to the inlay-
tooth margin. This increases the risk of debonding at
the adhesive interface and increases the fracture risk
at the contact surface of the ceramic. The risk of failure
at the base of the inlay may be minimized by increasing
the minimum inlay thickness.
ACKNOWLEDGMENTS
This study was supported by KAKENHI 20592307 (to N.W.)
from JSPS/MEXT.
REFERENCES
1. Al-Sukhun J, Kelleway J, Helenius M. Development of a three-dimen-
sional finite element model of a human mandible containing endos-
seous dental implants. I. Mathematical validation and experimental
verification. J Biomed Mater Res A 2007;80:234-246.
2. Anusavice KJ, Kakar K, Ferree N. Which mechanical and physical testing
methods are relevant for predicting the clinical performance of ceramic-
based dental prostheses? Clin Oral Implants Res 2007;18(suppl 3):
218-231.
3. Banks RG. Conservative posterior ceramic restorations: a literature
review. J Prosthet Dent 1990;63:619-626.
4. Charlton DG, Roberts HW, Tiba A. Measurement of select physical and
mechanical properties of 3 machinable ceramic materials. Quintes-
sence Int 2008;39:573-579.
5. Craig RG, Peyton FA, Johnson DW. Compressive Properties of Enamel,
Dental Cements, and Gold. J Dent Res 1961;40:936-945.
6. Farah JW, Craig RG, Meroueh KA. Finite element analysis of three- and
four-unit bridges. J Oral Rehabil 1989;16:603-611.
7. Federlin M, Krifka S, Herpich M, Hiller KA, Schmalz G. Partial ceramic
crowns: influence of ceramic thickness, preparation design and luting
material on fracture resistance and marginal integrity in vitro. Oper
Dent 2007;32:251-260.
8. Ferrario VF, Sforza C, Serrao G, Dellavia C, Tartaglia GM. Single tooth
bite forces in healthy young adults. J Oral Rehabil 2004;31:18-22.
9. Hansen S. Preparations for Cerec inlays, onlays, and partial crowns. Int
J Comput Dent 1999;2:299-309.
10. Hernandez AI, Roongruangphol T, Katsube N, Seghi RR. Residual inter-
face tensile strength of ceramic bonded to dentin after cyclic loading
and aging. J Prosthet Dent 2008;99:209-217.
11. Käse HR, Tesk JA, Case ED. Elastic constants of two dental porcelains.
J Mater Sci 1985;20:524-531.
12. Kelly JR. Perspectives on strength. Dent Mater 1995;11:103-110.
13. Kim JW, Bhowmick S, Chai H, Lawn BR. Role of substrate material in
failure of crown-like layer structures. J Biomed Mater Res B Appl Bio-
mater 2007;81:305-311.
14. Krämer N, Frankenberger R. Clinical performance of bonded leucite-re-
inforced glass ceramic inlays and onlays after eight years. Dent Mater
2005;21:262-271.
Vol 13, No 4, 2011
pyri
No CoOna et gal
ht
t Federlin M. Ce-
15. Krifka S, Anthofer T, Fritzsch M, Hiller KA, Schmalz G,fo
r P cusp
by N
ramic inlays and partial ceramic crowns: influence of remaining
ub in
Q ui
wall thickness on the marginal integrity and enamel crack formation
lica
vitro. Oper Dent 2009;34:32-42.
ti
16. Krifka S, Stangl M, Wiesbauer S, Hiller KA, Schmalz G, Federlin M. on
te ot
n
Influence of different cusp coverage methods for the extension of ce-
ce
ss einnvitro.
fo r
ramic inlays on marginal integrity and enamel crack formation
Clin Oral Investig 2009;13:333-341.
17. Lang LA, Wang RF, Kang B, White SN. Validation of finite element analy-
sis in dental ceramics research. J Prosthet Dent 2001;86:650-654.
18. Oh W, Götzen N, Anusavice KJ. Influence of connector design on
fracture probability of ceramic fixed-partial dentures. J Dent Res
2002;81:623-627.
19. Otto T, De Nisco S. Computer-aided direct ceramic restorations: a 10-
year prospective clinical study of Cerec CAD/CAM inlays and onlays. Int
J Prosthodont 2002;15:122-128.
20. Otto T, Schneider D. Long-term clinical results of chairside Cerec CAD/
CAM inlays and onlays: a case series. Int J Prosthodont 2008;21:
53-59.
21. Qasim T, Bush MB, Hu X, Lawn BR. Contact damage in brittle coating
layers: influence of surface curvature. J Biomed Mater Res B Appl Bio-
mater 2005;73:179-185.
22. Reiss B. Clinical results of Cerec inlays in a dental practice over a pe-
riod of 18 years. Int J Comput Dent 2006;9:11-22.
23. Sano H, Ciucchi B, Matthews WG, Pashley DH. Tensile Properties of
Mineralized and Demineralized Human and Bovine Dentin. J Dent Res
1994;73:1205-1211.
24. Seghi RR, Sorensen JA. Relative flexural strength of six new ceramic
materials. Int J Prosthodont 1995;8:239-246.
25. Sjögren G, Molin M, van Dijken JW. A 10-year prospective evaluation
of CAD/CAM-manufactured (Cerec) ceramic inlays cemented with a
chemically cured or dual-cured resin composite. Int J Prosthodont
2004;17:241-246.
26. Thompson JY, Bayne SC, Heymann HO. Mechanical properties of a new
mica-based machinable glass ceramic for CAD/CAM restorations. J
Prosthet Dent 1996;76:619-623.
27. Thordrup M, Isidor F, Hörsted-Bindslev P. A prospective clinical study of
indirect and direct composite and ceramic inlays: ten-year results. Quin-
tessence Int 2006;37:139-144.
28. van Dijken JW, Hasselrot L, Ormin A, Olofsson AL. Restorations with
extensive dentin/enamel-bonded ceramic coverage. A 5-year follow-up.
Eur J Oral Sci 2001;109:222-229.
29. Yi YJ, Kelly JR. Effect of occlusal contact size on interfacial stresses
and failure of a bonded ceramic: FEA and monotonic loading analyses.
Dent Mater 2008;24:403-409.
30. Yoshida K, Meng X, Kamada K, Atsuta M. Influence of surface charac-
teristics of four silica-based machinable ceramics on flexural strength
and bond strength of a dual-curing resin luting agent. J Adhes Dent
2007;9:407-413.
31. Zhou J, Mah J, Shrotriya P, Mercer C, Soboyejo WO. Contact damage
in an yttria stabilized zirconia: implications. J Mater Sci Mater Med
2007;18:71-78.
Clinical relevance: This finite element contact stress
analysis indicates that risk of debonding at the
adhesive interface of bonded ceramic inlays is in-
creased if the occlusal contact becomes close to the
inlay-tooth margin, and that the debonding potentially
causes ceramic fracture on the occlusal surface.
373