An in vitro study of the compressive load at fracture of Procera AllCeram

crowns with varying thickness of veneer porcelain

Briony Webber, BDS, MSc,a Ailbhe McDonald, BDentSc, MSc, PhD,b and
Jonathon Knowles, BSc, PhD, MIM, Cengc
Eastman Dental Institute for Oral Health Care Sciences, University College London, London,
United Kingdom
Statement of problem. When designing and cutting a crown preparation, a balance must be achieved
between the amount of tooth structure conserved and the space created for the restoration. All-ceramic crowns,
which contain a core, are usually veneered with porcelain to produce the desired aesthetics and contour. The
effect of veneer porcelain on the strength of the restoration has been a cause of debate.
Purpose. The purpose of this study was to investigate the effect of different thickness of veneer porcelain on the
compressive load at fracture of Procera AllCeram crowns.
Material and Methods. Sixty brass dies were fabricated with a crown-like preparation and a chamfer margin.
Sixty crowns were fabricated with a 0.6-mm-thick core: Procera crowns with either a 0.4-mm- or 0.9-mm-thick
veneer of AllCeram (Groups 1 and 2 respectively) and In-Ceram crowns with a 0.9-mm-thick veneer of Vitadur
Alpha porcelain (Group 3). Each group consisted of 20 crowns. In-Ceram crowns were used as the control group.
All crowns were measured at 4 axial and 1 occlusal random locations before autoglazing. They were air abraded
and adhesively bonded onto the appropriate brass die using Clearfil Newbond Bonding Agent, Clearfil Porcelain
Bond Activator, and Panavia 21 TC Dental Adhesive as the luting agent. After storage in distilled water at 37°C
for 24 hours, the specimens were placed in a compressive test rig within an Instron universal testing machine and
loaded in the center of the occlusal surface with a 4-mm diameter stainless steel ball. An axial preload of 20 N was
applied before compressive testing at a crosshead speed of 0.1 mm/minute until fracture occurred. An analysis of
variance revealed no significant difference in the load at fracture between the 3 groups (P⫽.3). Using a Bonferroni
correction, independent t tests revealed no significant difference in load at fracture for any of the pair-wise
comparisons between the 3 test groups (P ⬍ .05).
Results. The mean load at fracture for Group 1 was 2197.6 N (SD ⫽ 776.4); Group 2, 2401.4 N (SD ⫽ 699.1);
and Group 3, 2581.0 N (SD ⫽ 715.6).
Conclusion. Within the parameters of this study, the axial thickness of veneer porcelain did not have a
significant effect on the compressive load at fracture of Procera AllCeram crowns. (J Prosthet Dent 2003;89:
154-60.)

CLINICAL IMPLICATIONS
In this study, changing the veneer porcelain thickness did not alter the load to fracture of the
crown specimens. Further studies are required to assess how these restorations may behave under
in vivo conditions.

T he design of a crown preparation should achieve a

balance between preserving tooth structure and provid-
Ceramics have the capacity to replicate the aestheti-
cally pleasing characteristics and vitality of natural teeth.
ing sufficient retention and resistance-form. Full veneer However, although strong in compression, ceramics can
crown preparations intended for metal castings are gen- be inherently brittle and weak when placed under tensile
erally more conservative of tooth structure than those and torsional stresses.2 Until now, the provision of metal
for either metal ceramic or all-ceramic restorations. substructures to support ceramics has enabled the use of
However, there is an ever-increasing demand for non- tooth-colored crowns in load-bearing areas. Aesthetic
metallic restorations.1 demands have now resulted in the development of all-
ceramic restorations suitable for posterior teeth.3
Submitted in partial fulfillment of the requirements for the degree of In order to provide satisfactory posterior all-ceramic
Master of Science in Conservative Dentistry, Eastman Dental restorations, strong alumina cores have been produced.
Institute for Oral Health Care Sciences, 1999. Presented as an Procera AllCeram (Nobel Biocare, Goteborg, Sweden)
IADR 2000 abstract, Washington, D.C.
a
Postgraduate, Department of Conservation.
and In-Ceram (Vita Zahnfabrik, Bad Sackingen, Ger-
b
Consultant, Department of Conservation. many) are 2 ceramic systems that incorporate a high
c
Reader and Head of Unit, Department of Biomaterials. alumina core. These cores differ in their manufacturing

154 THE JOURNAL OF PROSTHETIC DENTISTRY VOLUME 89 NUMBER 2

WEBBER, McDONALD, AND KNOWLES THE JOURNAL OF PROSTHETIC DENTISTRY

process and are also intrinsically different in that the

Procera AllCeram core contains a densely sintered alu-
mina core, and In-Ceram is made of a high alumina core
(approximately 70%) which is subsequently glass infil-
trated.4 However, alumina cores tend to be opaque and
require the use of veneer porcelain to mask the core and
provide the desired contours.
It has been suggested that the veneer porcelain may
be substantially weaker than the core5 and may be im-
portant in determining the load bearing capacity of the
restoration.6 Wagner and Chu7 found that the flexural
strength of In-Ceram core material varied greatly,
whereas the Procera core material was found to be more
uniform. Similarly, previous studies have indicated that
the Procera core material has a higher failure stress than Fig. 1. External contouring template used to shape veneer
In-Ceram core material.4,8 However Neiva et al9 found porcelain prior to firing. Core contouring template also used.
no difference in load to fracture between Procera and Both core and external contouring templates positioned
In-Ceram crowns that were resin-cemented, which sug- around central ball race jig.
gests that the choice of luting agent may affect the
strength of the ceramic restoration. The presence of
flaws on the internal fit surface are often the site of (Group 1) or 9 mm (Groups 2 and 3). Group 1 repre-
fracture initiation,10 and a resin luting agent may reduce sented Procera crowns with an axial reduction of 1.0
the tendency for fracture initiation. Thompson et al11 mm. Group 2 (Procera crowns) and Group 3 (In-Ceram
cited that the clinical failure rate for acid-etched, resin- crowns) had an axial reduction of 1.5 mm. The brass
bonded ceramic restorations had been found to be lower dies were circular in the horizontal plane.
than when traditional cements were used. Ceramic res- A vinyl polysiloxane impression was made of each
torations are usually etched internally prior to cementa- individual brass preparation (Extrude, light and heavy
tion. However, it has been demonstrated that acid-etch body; Kerr Corp, Romulus, Mich.). After 24 hours a
has little effect on high alumina cores, and air abrasion Silky Rock die stone (WhipMix Corp, Louisville, Ky.)
has been advised as a surface preparation.12 was poured, trimmed, labeled, and assigned to the ap-
Many factors may be associated with crack initiation propriate group (Procera or In-Ceram).
and propagation within a dental ceramic. These include The stone dies intended for the Procera crown group
the shape and thickness of the restoration, microstruc- were coated with a small amount of air-thinned cyano-
tural inhomogeneities, size and distribution of surface acrylate (Bracon Ltd, East Sussex, UK). One operator
flaws, residual processing stresses, the magnitude, direc- scanned these dies, and the data were transmitted to a
tion, and frequency of the applied load,13,14 the restora- remote manufacturing site (Procera Sandvik, Stock-
tion-cement interfacial defects, the elastic modulus of holm, Sweden) where the cores were fabricated and re-
the restoration components, and environmental ef- turned to the laboratory. A nickel chromium template
fects.13 was fabricated to allow the Procera core dimensions to
The objective of this in vitro study was to investigate be replicated for the In-Ceram group.
the effect on load at fracture of 2 different thickness of The Silky Rock stone dies for the In-Ceram speci-
veneer porcelain when supported by a constant thick- mens were coated with spacer (Blue cement spacer
ness of Procera AllCeram core. Crown-shaped restora- [Batch 747]; Belle de St Claire, Chatsworth, Calif.) du-
tions were used because it has been reported that these plicated in In-Ceram Special Plaster (Vita Zahnfabrik,
may allow the restorations to behave in a manner that Bad Sackingen, Germany), trimmed, and labeled.
potentially represents the clinical situation more closely An In-Ceram slip was made by adding 38 g of In-
than ceramic discs.15 Ceram Alumina Powder to 5 mL of In-Ceram Mixing
Liquid and 1 drop of Additive. The slip was applied to
the plaster dies and formed to correspond with the core
MATERIAL AND METHODS contouring template (Bs 970; RS Components Ltd,
Sixty brass dies (BS 2874; Rotary Engineering Ltd, Corby, UK) (Fig. 1). After 30 minutes, 1 coat of Stabi-
Reading, UK) were fabricated with a height of 47 mm liser (Vita Zahnfabrik) was applied, and the cores were
and base diameter of 10 mm. Each die was milled to fired in the Vita-Inceramat furnace (Vita Zahnfabrik) at
produce a crown preparation shape with an 8-degree 1120°C.
taper, 7-mm height, 4.2-mm occlusal diameter, and a The sintered specimens were then infiltrated with
diameter between the opposing finish lines of 8 mm molten glass and fired at 1100°C. The sintered substruc-

FEBRUARY 2003 155

THE JOURNAL OF PROSTHETIC DENTISTRY WEBBER, McDONALD, AND KNOWLES

Fig. 2. Split brass mold and 2 different sizes of completed Fig. 3. Compressive test rig. A, Ceramic crown. B, Test rig
crowns (0.4-mm and 0.9-mm veneer thickness) in respective holding samples in place.
split brass molds. Space between mold and crown indicates
degree of porcelain shrinkage on firing, and final shape
provided by external contouring template.

Table I. The mean and median load at fracture, the standard deviation and the 95% confidence interval for each
experimental group

Crown type and Mean load 95% CI Median load

veneer thickness at fracture SD at fracture
Group (mm) (N) (N) Lower Upper (N)

1 Procera 0.4 2197.6 776.4 1834.2 2561.0 2230.0

2 Procera 0.9 2401.4 699.1 2074.2 2728.5 2371.5
3 In-Ceram 0.9 2581.0 715.6 2246.0 2915.9 2391.0
N, Newton.

ture was then checked for microcracks. Excess glass shrinkage of the dentine porcelain allowed enamel por-
was removed, and the occlusal and axial walls were celain build-up to full contour.
reduced where necessary with a 2 striper diamond bur Each crown was viewed under magnification (⫻12)
(709.9M, 100 ␮m grit; Austenal Ltd, Middlesex, on the corresponding brass die to ensure it was fully
UK). The cores were steam cleaned and air dried, and seated. The crowns were measured with calipers (Calip-
their fit was verified on the brass dies using a micro- rett “S” calipers, Nr 1122-1000; Renfert GmbH Indus-
scope at ⫻12 magnification. The external core sur- triegebiet, Hilzingen, Germany) at 4 random locations
faces were air abraded with 50-␮m aluminum oxide axially and 1 occlusally to ensure that the final occlusal
and steam-cleaned. and axial dimensions were appropriate. They were then
Each veneer porcelain powder was mixed with the autoglazed.
appropriate liquid and applied to the core with a porce- The crowns were internally air abraded with 50-␮m
lain brush. The veneer porcelain for the Procera crown aluminium oxide. All crowns and brass dies were steam
groups was AllCeram porcelain (Duceram Dental cleaned and air dried. ED primer (Kuraray Co Ltd, Dus-
GmbH, Rosbach, Germany). Vitadur Alpha porcelain seldorf, Germany) was applied to the brass die before air
(Vita Zahnfabrik) was used for veneering the In-Ceram drying, and the fit surfaces of all crowns were silanated
crowns. The brass die and crown were placed in a split with Clearfil Porcelain Bond Activator in conjunction
brass mold (Fig. 2) to allow porcelain build-up. The with Clearfil Newbond Bonding Agent (Kuraray Co
brass die with crown in situ was transferred to a ball race Ltd) before air-drying. Two complete revolutions of the
jig (a jig containing a circular arrangement of ball bear- Panavia 21 TC (Kuraray Co Ltd) dispensing cartridge
ings16,17), where an external contouring template was were dispensed, mixed, and applied to the internal sur-
used to shape the porcelain (Fig. 1). The crowns were face of each crown. Manual pressure was initially applied
fired (Multimat MC II furnace; Degussa, Macclesfield, to seat each crown on the appropriate brass die, and each
UK) according to the manufacturers recommended fir- crown was held in place while excess cement was re-
ing cycles for each type of porcelain. Upon firing, the moved and Oxyguard II (Kuraray Co Ltd) applied to the

156 VOLUME 89 NUMBER 2

WEBBER, McDONALD, AND KNOWLES THE JOURNAL OF PROSTHETIC DENTISTRY

Table II. One-way analysis of variance (ANOVA)

Sum of Variance ratio

Load at fracture Squares df Mean square (F ratio) Significance

Comparison between groups 1471516 2 735758.1 1.4 0.3

Comparison within groups 3.0 ⫻ 107 57 534539.2 – –
Total 3.2 ⫻ 107 59 – – –

Table III. The frequency, the mean load at fracture and the standard deviation (SD) for each mode of fracture are indicated

Fracture frequency per group

Mean Median
Frequency load at load at Group 1 Group 2 Group 3
of fracture fracture fracture Procera Procera In-Ceram
Fracture mode mode (%) (N) SD (N) 0.4 mm 0.9 mm 0.9 mm

Sheer veneer 5 (8.3) 1536.6 102.2 1494.0 2 1 2

Delam. Veneer 23 (38.3) 2109.2 709.8 2103.0 11 12 –
Shear core 8 (13.3) 2788.5 691.4 2624.0 2 2 4
Brass exposed 24 (40.0) 2712.3 598.9 2584.5 5 5 14
N, Newtons; Sheer veneer, Shear fracture occurred through the veneer porcelain only; Delam. Veneer, delamination of the veneer porcelain occurred, exposing
the core; Shear core, shear fracture occurred through both the veneer porcelain and part of the core. There was no brass die exposed; Brass exposed, the brass die
was exposed.

Fig. 4. Frequency versus load at fracture column chart for Procera 0.4-mm specimens

external margins. The specimens were placed in a verti- ing force in the molar region has been reported to be
cal static load apparatus (constructed within the Depart- 847 N for men and 597 for women.18 The load at
ment of Biomaterials, Eastman Dental Institute) for 10 fracture of each specimen was recorded in Newtons
minutes under a 5-kg load. They were stored in dis- (N), entered into a Microsoft Excel (Version 4)
tilled water for 24 hours at 37°C, after which each spreadsheet program (Microsoft, Redmond, Va.), and
specimen was placed in a compressive test apparatus in imported into an SPSS statistical program (Release 8;
a universal testing machine (Instron, High Wycombe, SPSS, Chicago, Ill.) for statistical analysis. The results
UK) (Fig. 3). The apparatus included a 4-mm– diam- were analyzed with a 1-way analysis of variance
eter stainless steel ball, which loaded the center of the (ANOVA). Independent t tests with a Bonferroni cor-
occlusal surface of the crown axially. An axial preload rection were also applied (P⬍.05).
of 20 N was applied, followed by the compressive test The fractured specimens and fragments were col-
at a crosshead speed of 0.1 mm/minute (use 50 kN lected, and 4 specimens from each group were selected
load cell) until fracture occurred. The maximum bit- for SEM examination. A polyvinyl siloxane impression

FEBRUARY 2003 157

 THE JOURNAL OF PROSTHETIC DENTISTRY WEBBER, McDONALD, AND KNOWLES

Fig. 5. Frequency versus load at fracture column chart for Procera 0.9-mm specimens.

Fig. 6. Frequency versus load at fracture column chart for In-Ceram 0.9-mm specimens.

(Doric CLD; Schottlander and Davis Ltd, Herts, UK) dies with a high modulus of elasticity.21,22 This factor
was made of the fractured surfaces of the specimens, will be considered when interpreting the results.
poured in Epoxy Coating Resin (Alec Tiranti Ltd, Lon-
don, UK), desiccated, and sputter coated (Edwards In- RESULTS
struments, Sussex, UK). They were then viewed for frac-
tography with an SEM (Cambridge Instruments Ltd, The mean and median load at fracture for all groups is
Cambridge, UK) at an accelerating voltage of 15 kV. recorded in Table I. The data are graphically presented
The experimental method has been described in pre- in Figures 4 through 6. The results were analyzed with a
vious studies16,17 and uses metal dies to support the 1-way analysis of variance (ANOVA), which demon-
ceramic restorations. Numerous studies have reported strated that there was no significant difference in the
the use of metal dies17,19,20 although the fracture load load at fracture between the groups (P⫽.3) (Table II).
of ceramic may be greater if crowns are supported by Independent t tests with a Bonferroni correction were

158 VOLUME 89 NUMBER 2

WEBBER, McDONALD, AND KNOWLES THE JOURNAL OF PROSTHETIC DENTISTRY

crowns has been recommended (Procera clinical manual

1998).
The brass dies used in this study, although not repli-
cating the elastic modulus of teeth, were homogenous in
composition and provided even, void-free support for
the ceramic restorations. In this study, crowns were sup-
ported by and bonded onto the brass dies, and the de-
gree to which the brass-resin bond influenced the load at
fracture of the restorations is not known. However, it
can be estimated that the fracture load was greater than
would have been achieved if the crowns were supported
by dies with a lower modulus of elasticity.21,22 However,
as the substrate was consistent for all groups, this was
not considered a confounding factor.
Surface preparation of all of the specimens prior to
cementation included air abrasion and silanation. It was
anticipated that silanation would have a minimal effect
on the bond between the resin and ceramic12 due to the
high alumina content of the cores. Its action was there-
fore most likely to be as a wetting agent.
One of the limitations of a single compressive load to
failure is that it does not replicate all of the clinical loads
to which the restoration is exposed.19 Fatiguing of ma-
terials can occur when cyclical loads are applied, such as
Fig. 7. SEM of fractured Procera crown, showing veneer under the physiologic conditions of mastication. This
porcelain (V), Procera core (C), Hackle fracture lines parallel experiment, therefore, addressed one aspect of loading,
to crack propagation (H), Wallner fracture lines perpendicu- which was chosen because it appeared to be appropriate
lar to crack propagation (W). for posterior teeth. Waltimo and Konenen18 found that
the maximum biting force in the molar region was 847
N for men and 597 N for women. Although the results
also applied. There was no significant difference in load of this study cannot be directly compared with the in
at fracture for any of the pairwise comparisons between vivo situation, the mean load at fracture was approxi-
the 3 test groups. mately 2200 to 2600 N and therefore far greater than
The mode of fracture, frequency, and mean load at clinically anticipated loads. This would appear to be con-
which the various fracture modes occurred were re- sistent with reports that a die with a high modulus of
corded for each specimen (Table III). Procera crowns elasticity can result in increased fracture loads of ceram-
exhibited delamination of the veneer, exposing the core ic.21,22
most frequently (Fig. 7), whereas the In-Ceram tended There was no significant difference in load at fracture
to fracture in a manner that resulted in the brass die between the 3 groups. It is interesting to note that the
being exposed. group that had the lowest load at fracture had the light-
est axial tooth reduction (1.0 mm). The median load at
fracture tended to be greater when the fracture propa-
DISCUSSION
gation path extended through the core than when the
This in vitro study aimed to determine the influence veneer sheared cohesively or delaminated from the core.
of the thickness of veneer porcelain on the load at frac- This was attributed to the increased energy required to
ture of all-ceramic crowns with high alumina cores. The cause fracture of the core.
results provide an indication of how materials may be- The veneer porcelain thickness did not have any sig-
have relative to each other but cannot be directly extrap- nificant effect on the load at fracture of the restorations,
olated to the clinical situation. nor was there any significant effect of the materials, Pro-
The methods used in this study were similar to pre- cera or In-Ceram, on the load at fracture. However, the
vious studies.16,17 For experimental convenience, the load at fracture for both Procera and In-Ceram crowns
occlusal surfaces of the preparations and ceramic resto- exhibited a large SD. The predictability of the success or
rations were flat to assist in the ease of fabrication and failure rate of all-ceramic restorations may therefore be
testing and did not replicate the cuspal inclines found uncertain, given their large SD values.
clinically. However preparation of a relatively flat occlu- The fracture pattern of the Procera and In-Ceram
sal surface to facilitate the scanning process for Procera groups differed in that the Procera cores tended to re-

FEBRUARY 2003 159

THE JOURNAL OF PROSTHETIC DENTISTRY WEBBER, McDONALD, AND KNOWLES

main intact when failure occurred, whereas the In-Ce- 8. Zeng K, Oden A, Rowcliffe D. Evaluation of mechanical properties of
dental ceramic core materials in combination with porcelains. Int J Prosth-
ram cores fractured in 18 of the 20 specimens. This may odont 1998;11:183-9.
have been due to greater fracture resistance of the Pro- 9. Neiva G, Yaman P, Dennison JB, Razzoog ME, Lang BR. Resistance to
cera core or to a weaker bond of the AllCeram veneer fracture of three all-ceramic systems. J Esthet Dent 1998;10:60-6.
10. Kelly JR, Giordano R, Pober R, Cima MJ. Fracture surface analysis of dental
porcelain to the core. It may also be that the surface ceramics: clinically failed restorations. Int J Prosthodont 1990;3:430-40.
preparation of the Procera core used in this study (air 11. Thompson JY, Rapp MM, Parker AJ. Microscopic and energy dispersive
abrasion and steam-cleaning) did not produce an opti- x-ray analysis of surface adaptation of dental cements to dental ceramic
surfaces. J Prosthet Dent 1998;79:378-83.
mal bond. 12. Awliya W, Oden A, Yaman P, Dennison JB, Razzoog ME. Shear bond
This study was in agreement with the findings of strength of a resin cement to densely sintered high-purity alumina with
Neiva et al9 who found no difference in load to fracture various surface conditions. Acta Odontol Scand 1998;56:9-13.
13. Thompson JY, Anusavice KJ, Naman A, Morris HF. Fracture surface char-
between Procera and In-Ceram crowns that were resin- acterization of clinically failed all-ceramic crowns. J Dent Res 1994;73:
cemented. This may be attributed to resin cementation 1824-32.
of die and ceramic, which act as a bonded system with 14. Kelly JR. Perspectives on strength. Dent Mater 1995;11:103-10.
15. Wakabayashi N, Anusavice KJ. Crack initiation modes in bilayered alu-
load transfer through each interface. mina/porcelain disks as a function of core/veneer thickness ratio and
The null hypothesis that there was no difference in supporting substrate stiffness. J Dent Res 2000;79:1398-404.
fracture strength between the groups was upheld. 16. Sobrinho LC, Cattell MJ, Glover RH, Knowles JC. Investigation of the dry
and wet fatigue properties of three all-ceramic crown systems. Int J Prosth-
odont 1998;11:255-62.
17. Sobrinho LC, Consani S, Knowles JC. Effect of convergence angle and
CONCLUSIONS luting agent on the fracture strength of In Ceram crowns. J Mater Sci:
Materials in Medicine 1999;10:493-6.
Values derived from this experimental method are 18. Waltimo A, Kononen M. A novel bite force recorder and maximal iso-
useful for comparative purposes only, and extrapolation metric bite force values for healthy young adults. Scand J Dent Res
1993;101:171-5.
to the clinical situation is not possible. Within the limi- 19. Probster L. Compressive strength of two modern all-ceramic crowns. Int J
tations of this study, the median load at fracture tended Prosthodont 1992;5:409-14.
to be greater when the fracture propagation path ex- 20. Sulaiman F, Chai J, Jameson LM, Wozniak WT. A comparison of the
marginal fit of In-Ceram, IPS Empress, and Procera crowns. Int J Prosth-
tended through the core than when the veneer sheared odont 1997;10:478-84.
cohesively or delaminated from the core. Also, there was 21. Scherrer SS, de Rijk WG. The fracture resistance of all-ceramic crowns on
no significant difference in load at fracture between the supporting structures with different elastic moduli. Int J Prosthodont 1993;
6:462-7.
3 experimental groups. 22. Anusavice KJ, Tsai YL. Stress distribution in ceramic crown forms as a
function of thickness, elastic modulus, and supporting substrate. In: Bum-
gardner JD, Puckett AD, editors. Proceedings of the Sixteenth Southern
REFERENCES Biomedical Engineering Conference. Biloxi (MS): Institute of Electrical and
Electronics Engineers; 1997. p. 246-67.
1. Donovan TE, Chee WW. Conservative indirect restorations for posterior
teeth. Cast versus bonded ceramic. Dent Clin North Am 1993;37:433-43. Reprint requests to:
2. Craig RG, Powers JM. Restorative Dental Materials. 11th ed. St Louis: DR AILBHE MCDONALD
Mosby; 2002. p. 551-92. DEPARTMENT OF CONSERVATION
3. Andersson M, Razzoog ME, Oden A, Hegenbarth EA, Lang BR. Procera: a EASTMAN DENTAL HOSPITAL AND INSTITUTE FOR ORAL HEALTH CARE SCIENCES
new way to achieve an all-ceramic crown. Quintessence Int 1998;29:285- 256 GRAYS INN RD.
96. LONDON WCIX 8LD,
4. Zeng K, Oden A, Rowcliffe D. Flexure tests on dental ceramics. Int J UNITED KINGDOM
Prosthodont 1996;9:434-9. FAX: 44-020-7915-1028
5. White SN, Caputo AA, Vidjak FM, Seghi RR. Moduli of rupture of layered EMAIL: a.mcdonald@eastman.ucl.ac.uk
dental ceramics. Dent Mater 1994;10:52-8.
6. Kelly JR, Tesk JA, Sorensen JA. Failure of all-ceramic fixed partial dentures Copyright © 2003 by The Editorial Council of The Journal of Prosthetic
in vitro and in vivo: analysis and modeling. J Dent Res 1995;74:1253-8. Dentistry.
7. Wagner WC, Chu TM. Biaxial flexural strength and indentation fracture 0022-3913/2003/$30.00 ⫹ 0
toughness of three new dental core ceramics. J Prosthet Dent 1996;76:
140-4. doi:10.1067/mpr.2003.85

160 VOLUME 89 NUMBER 2