doi: 10.1111/j.1365-2591.2006.01085.

A finite element analysis of ferrule design on

restoration resistance and distribution of stress
within a root

I. Ichim1, D. V. Kuzmanovic2 & R. M. Love2

1
Department of Oral Sciences; and 2Department of Oral Diagnostic and Surgical Sciences, University of Otago School of Dentistry,
Dunedin, New Zealand

Abstract increasing ferrule height the displacement and rotation

of the crown reduced in conventional and crown-
Ichim I, Kuzmanovic DV, Love RM. A finite element
lengthening models with maximum reduction occur-
analysis of ferrule design on restoration resistance and distri-
ring when the ferrule height reached 1.5 mm. In ferrule
bution of stress within a root. International Endodontic Journal.
models, higher levels of tensile stress developed in
39, 443–452, 2006.
internal (by a factor of 8) and mid-root palatal (by a
Aim To analyse the effect of ferrule height upon the factor of 90) dentine at the cervical margin of the
mechanical resistance and stress distribution within a preparation. With an increase in ferrule height, the area
root to explain variations in the pattern of root fracture. of tensile stress within the palatal mid-dentine expanded
Methodology An extracted, intact, caries free, max- towards the cervical margin. Similar patterns and stress
illary right central incisor was scanned by laser and values were recorded for the crown-lengthening models.
then reconstructed on a computer to produce a model Conclusion The study confirms that a ferrule increa-
of the tooth and associated periodontal ligament. A ses the mechanical resistance of a post/core/crown
simulated post/core/crown restoration was constructed restoration. However a ferrule creates a larger area of
on conventional tooth preparations with various palatal dentine under tensile stress that may be a
ferrules. The crown was loaded with a simulated favourable condition for a crack to develop. Crown-
500 N force and the simulated displacement of com- lengthening did not alter the levels or pattern of stress
ponents and the tensile and compressive stress within within compared with conventional ferrule preparations.
the tooth structure were recorded.
Keywords: ferrule, post, root fracture, tensile stress.
Results Without a ferrule preparation, the simulated
crown tilted to the labial and rotated distally. With Received 1 August 2005; accepted 9 November 2005

prosthesis (Sorensen & Martinoff 1984). If the amount

Introduction
of lost coronal tooth substance impairs retention of a
A number of factors influence the functional life span of a fixed prosthesis, e.g. a crown, a cast indirect post-and-
restoration on a root filled tooth, these include the core may be required to replace missing tooth structure
amount of remaining enamel and dentine, the properties and provide retention for the restoration. The main
of restorative materials and the design of the tooth function of a post is to provide intracanal retention of the
preparation. After root canal treatment the functional core/crown restoration and to distribute functional
and morphological integrity of a tooth is normally loadings to a larger area of the remaining coronal tooth
restored by an intracoronal restoration or fixed structure and root, however it has been shown that a post
does not strengthen the root (Guzy & Nicholls 1979).
It is generally accepted that when restoring a root
Correspondence: Robert M. Love, Professor, Department of filled tooth with a post/core and crown a ferrule design
Oral Diagnostic and Surgical Sciences, School of Dentistry,
University of Otago, PO Box 647, Dunedin, New Zealand (Tel.:
should be incorporated into the tooth preparation and
0064 3 479 7121; fax: 0064 3 479 7046; e-mail: robert. crown. The ferrule design incorporates a crown with a
[email protected]). 360
collar that surrounds the perimeter of the

ª 2006 International Endodontic Journal International Endodontic Journal, 39, 443–452, 2006 443
 FEA analysis of ferrule designs Ichim et al.

prepared parallel dentine walls and extends cervically maximum dimensional deviation of the digital model to
to the shoulder of tooth preparation. This design the real tooth of <100 lm.
enhances the mechanical resistance of the treated Using clinical guidelines (Rosenstiel et al. 2001) the
tooth by distributing forces on the remaining tooth digital model was then altered to accommodate a custom
structure (Tjan & Whang 1985, Barkhordar et al. made post and core and crown restoration. The diameter
1989, Sorensen & Engelman 1990) and thereby of the post at the cervical and apical sections was 1/3 of
minimising disruption of the integrity of the bond of the diameter of the root and the post extended apically to
the post/core or crown to the tooth and the potential 2/3 of the root length. The crown had a uniform
for root fracture. The ferrule may also affect the pattern thickness of 1 mm, except at the incisal edge where it
of root fracture in post/core restored teeth, from was 1.5 mm and the margin of the shoulder fol-
debonding of the post/core to vertical root fracture in lowed the pattern of the cementoenamel junction
nonferrule designs to oblique or horizontal root (Fig. 1). A periodontal ligament (PDL) was modelled as
fractures in designs incorporating a ferrule (Barkhordar a uniform 250-lm thick shell surrounding the root and
et al. 1989, Milot & Stein 1992, Ng et al. 2004). finished 1.5 mm apical to the cementoenamel junction.
Numerous studies have linked the ferrule design with The presence of the PDL was required to accurately
various factors such as remaining dentine thickness and reproduce constraints of the system and also to provide
distribution (Tjan & Whang 1985), different designs of the tensile component as discussed below.
the cervical collar (Barkhordar et al. 1989, Sorensen &
Engelman 1990), ferrule height and configuration
Ferrule design
(Sorensen & Engelman 1990, Libman & Nicholls 1995,
Isidor et al. 1999) and post design and material (Saupe From a starting point at the crown margin, which had
et al. 1996, Gegauff 2000), with ferrule height being no ferrule, four conventional ferrule preparations
identified as a critical factor to mechanical resistance. involving the simulated coronal dentine were produced
The aims of this study were to analyse the effect of ferrule at 0.5-mm intervals up to a height of 2 mm and at a
height upon the mechanical resistance of a reconstruc- constant 1-mm thickness of dentine (Fig. 2). To simu-
ted tooth by Finite Element Analyses (FEA) and to late a crown lengthening procedure, four models with
correlate the different ferrule designs with stress distri- similar dimensions were produced apical to the
bution within a root in order to explain variations in the cementoenamel junction extending down the radicular
pattern of root fracture. dentine. In these cases the simulated crown margin
extended down the root face and the level of the
periodontal ligament was adjusted to the relevant
Materials and methods
length of the extended ferrule preparation.
In order to obtain clinically relevant FEA results, The root was divided into two concentric areas
particular attention was given to an accurate computer around the post preparation to mark the inner-
reconstruction of a tooth. As the main focus was to (adjacent to the post), mid- and outer radicular dentine
assess the influence of the ferrule height, the mathe- and five stress reading sites were located at 3-mm
matical models were represented in a consistent man- intervals along each of these boundaries (Fig. 3). In
ner, maintaining the ferrule height as the unique order to accurately read the stress associated with a
geometry variable. ferrule preparation, the cervical mid-dentine reading
site was aligned to correspond to the angle between the
ferrule and the cervical seat of the preparation.
Three-dimensional model
An extracted, intact, caries free, maxillary right central
Boundary conditions and materials
incisor was scanned by a laser 3D scanner (Picza Pix-4,
Roland Co., Osaka, Japan). Starting from the point Tooth structures are normally subjected to loading
cloud from the scan, the tooth was reconstructed on a within their elastic range and the goal of this study was
computer (PIV 2.4 GHz; 2GB RAM, Windows XP, to analyse the consequences of design changes to the
Microsoft Corporation, Redmond, WA, USA) using ferrule preparation, hence static linear analysis was
current reverse engineering techniques in a generic performed for each case. In a static linear study, all
CAD package (Rhinoceros 3D; v.3. Robert McNeel & loads are applied slowly and gradually until they reach
Assoc., Seattle, WA, USA). This technique resulted in a their maximal values, there upon the loads remain

444 International Endodontic Journal, 39, 443–452, 2006 ª 2006 International Endodontic Journal
 Ichim et al. FEA analysis of ferrule designs

.9
7 16
.
11

0
1.

.2
10 2.
0
1. 5
0 1.
1.
0

Figure 1 Sample image of the 1 mm

5
ferrule model illustrating the dimensions 1.
of the model (millimetres).

No ferrule 0.5 mm ferrule 1.0 mm ferrule 1.5 mm ferrule 2.0 mm ferrule

Figure 2 The five models demonstrating different coronal dentine ferrule height.

constant (time-invariant) and the FE calculation is all the other parts of the assembly were of a gap contact
made until the solution is convergent (no variation). type. A gap contact condition allows the faces to move
The FE analysis was performed using a general purpose away from each other, while preserving the physical
FEA package (Cosmos DesignStar 4.0; Structural requirement of not penetrating each other. Also, to
Research and Analysis Corp., Santa Monica, CA, fully analyse the effect of the ferrule design, the friction
USA) using the computer as described above. The coefficient between the different components of the
models were meshed using linear tetrahedral elements assembly was set to zero, thus any displacement that
and in order to reduce the variability all the models occurred between the components was not restricted by
were meshed using 600 lm elements. This resulted in friction. The material properties of the various
an average of 27.000 elements per model (Fig. 4). simulated components were set in the solver (Table 1).
In order to simulate the effect of the inner socket To minimise variables between restorative materials,
wall, the models were fixed on the outer surface of the the properties of a nonprecious dental alloy were used
PDL with no rotation or translation allowed. The for the post and core and also the crown. The PDL
junction between the PDL and the external root surface elastic modulus was based on the standard properties of
was considered as bonded, while the contacts between a rubber (Saupe et al. 1996) and determined by a

ª 2006 International Endodontic Journal International Endodontic Journal, 39, 443–452, 2006 445
 FEA analysis of ferrule designs Ichim et al.

Figure 3 Representative diagram show-

ing the boundaries of inner-, mid- and
outer dentine and the location of the
stress reading sites in these areas (num-
bers 1–5).

Figure 4 Meshed model demonstrating

palatal surface loading area and
restraints at external root surface.

preliminary test by steadily increasing the modulus of onto a 5 mm2 area of the palatal surface incisal to the
elasticity of the material, sourced from the software cingulum (Milot & Stein 1992). This force corresponds
material library, until the vertical displacement of the to the upper range of occlusal load during function.
simulated tooth under a 300 N axial load was 200 lm
(Mühlemann 1960). This approach ensured that the
Post-processing
elastic deformation of the material was within the
physiologic range of the human PDL. After the numerical input was created and the problem
The crown was loaded with a simulated 500 N force solved, the displacements and principal stresses were
directed at 120
to the longitudinal axis of the tooth assessed. The principal stresses (r1 and r3) were used in

446 International Endodontic Journal, 39, 443–452, 2006 ª 2006 International Endodontic Journal
 Ichim et al. FEA analysis of ferrule designs

Table 1 Properties of materials used in FEA models. better understanding of the structure response.
Material Elastic modulus (MPa) Poisson ratio Although the study was not focused on predicting the
system’s failure, the principal stress determination can
Periodontal ligament 11.76 0.49
Steel 210 000 0.28 provide useful data considering that the maximum
Dentine 14 700* 0.31 stress criterion states that failure occurs when the
*(Sano et al. 1994).
maximum (normal) principal stress reaches either the
uniaxial tension strength or the uniaxial compression
the assessment because on one hand the root is made strength. Furthermore, the stiffness of the structures
out of dentine, which is a brittle material and on the was compared by means of displacements calculated on
other hand, a comparison between the tensile and the crown’s incisal margin.
compressive stresses generated by each design allows a
Results

Displacement of the simulated crown

An analysis of the displacement of the simulated crown
subsequent to the palatal force demonstrated that the
URES
crown tilted to the labial and rotated with the distal
0.512
0.469 incisal angle being displaced more towards the labial
0.426 compared with the mesial incisal edge (Fig. 5). This
0.384
0.341
axial rotation may be due to the asymmetric shape of
0.298 the cervical section or the morphology of the palatal
0.256 surface.
0.213
0.171
The displacement values recorded at the midline of
0.128 the palatal surface of the crown near the incisal edge
0.085 and the maximal displacement at the distal incisal edge
0.043
0.000
demonstrate that crown displacement reduced with
increasing ferrule height up to 1.5 mm, after which
there was no difference (Fig. 6). In the model a ferrule
Max: 0.512 of, or >1.5 mm resulted in approximately a 35%
reduction in displacement compared with the nonfer-
Figure 5 Image showing the point of application of the rule preparation. The anti-rotational effect of the ferrule
simulated force on the palatal surface with resulting uneven can be seen by the reduction in the difference between
displacement of the crown seen as a rotation of the distal the maximum and mid-point displacement when a
surface buccally. ferrule is present (Fig. 6).

Figure 6 The effect of ferrule height

on simulated crown displacement at
the distal (maximum displacement)
and mid-palatal incisal margin of
coronal dentine ferrule and radicular
dentine (crown extension) ferrule
designs.

ª 2006 International Endodontic Journal International Endodontic Journal, 39, 443–452, 2006 447
 FEA analysis of ferrule designs Ichim et al.

In the case of crown lengthening the crown dis-

placement showed a similar trend, although the values
were higher than the coronal dentine ferrules because
of the increase in the length of the extracoronal arm
(Fig. 6).

Stress analysis
Because of the direction of the force applied to the
palatal surface of the simulated crown there was a
tendency for the post to flex in a labial direction. This
will result in a compressive stress developing in the
labial dentine while the palatal dentine will be in
tension as it is restricted by the periodontal mem-
brane.

Tensile stress
Figure 7(a,b) demonstrates that when a ferrule was
present, a higher level of tensile stress (r1) developed in
internal and mid-root palatal dentine at the cervical
margin of the preparation compared with a preparation
with no ferrule. In internal dentine, the tensile stress for
all models increased along the root to approximately
the mid-root area, after which it decreased, with the
stress being higher in models with ferrule preparations
compared with no preparation (Fig. 7a). Compared
with the values at the cervical area, the stress in mid-
root palatal dentine was higher along the root when a
ferrule was not present and less with all ferrule designs,
with the values of the ferrule models similar to the
nonferrule model (Fig. 7b). The tensile stress on
the external palatal dentine was not influenced by the
presence of a ferrule preparation (Fig. 7c).
A colour plot analysis showed that there were two
peaks of maximal tensile stress on the palatal wall of
the root at the junction between the ferrule and the
Figure 7 Plots of the tensile stress in palatal (a) internal
cervical margin of the preparation and the other dentine, (b) mid-dentine, (c) external dentine at the various
located within the periphery of the root dentine in the recording points in the model.
cervical two-fifths (approximately 3–5 mm from palatal
cervical margin of the restoration) (Fig. 8a). With an the stress values within the dentine differed between
increase in ferrule height, the area of tensile stress the standard and crown-lengthened models (Fig. 10).
within the palatal mid-dentine expanded towards the
cervical margin creating an almost constant area
Compressive stress
under tension bridging the external two-fifths of the
root from the palatal aspect of the ferrule (Fig. 8b). The presence of a ferrule resulted in a marked decrease
However, the maximum nodal stress values showed in the r3 stress (compressive component) in the labial
very little difference between the samples. internal root dentine at the cervical measuring point,
Neither did the crown lengthening alter the pattern while the compressive stress at all other points for all
of tensile stress in any area of the root (Fig. 9) models were similar (Fig. 11a). In contrast, on the
compared with the standard models (Fig. 7b) nor were labial external root dentine, there was very little

448 International Endodontic Journal, 39, 443–452, 2006 ª 2006 International Endodontic Journal
 Ichim et al. FEA analysis of ferrule designs

(a) Princ_1
250.000
229.175
208.350
187.525
166.700
145.875
125.050
104.226
83.401
62.576
41.751
20.926
0.101

(b) Princ_1
250.000
229.167
208.333
187.500
166.667
145.833
125.000
104.167
83.333
62.500
41.667
20.833
0.000

Figure 8 Colour plots of the tensile stress in (a) 0.5 mm

ferrule and (b) 2.0 mm ferrule models demonstrating the Figure 10 Tensile stress plots comparing stresses at (a) inter-
extension of the area under tension with increased ferule nal dentine and (b) mid-dentine in standard 1 mm ferrule and
height. crown-lengthening models.

into the direction of forces playing on the restoration

and tooth structure. In the absence of a ferrule
preparation, an apical-labial directed force applied to
the palatal aspect of a crown will produce a tilting
force on the crown and core to the labial and an
axial force that displaces the post/core occlusally,
with the point of rotation below the contact point at
a point along the labial shoulder of the preparation
and the external root surface. As a result of the force
and the limited compliance of the labial periodontal
membrane, compressive stress develops in the labial
Figure 9 Plot of the tensile stresses in the palatal mid-dentine dentine and in the palatal wall of the root canal at
for the crown lengthening models. the apical extent of the post. The elastic restraint of
the palatal periodontal ligament and direction of the
difference in compressive stress, although there was a force results in tensile stress generated within the
trend for reduced stress with increasing ferrule height palatal root dentine and within the post in its
(Fig. 11b). A similar pattern and stress values were coronal aspect near the junction with the core
recorded for the crown-lengthening models. (Fig. 12). In a clinical situation, these stresses will
cause shearing stress on the luting cement and
subsequent break down of the material, resulting in
Discussion
movement of the restoration and microleakage of oral
An analysis of the displacement of components of a fluids which will accelerate the mechanical failure of
post/core/crown restoration can give some insight the restoration. The tensile stress values (Fig. 7) were

ª 2006 International Endodontic Journal International Endodontic Journal, 39, 443–452, 2006 449
 FEA analysis of ferrule designs Ichim et al.

is low risk of a crack initiating in the dentine,

however once the post/core becomes loose the
resulting lever action will favour fracture of the post
in the coronal portion or the development of a
vertical root fracture.
The results of the simulated crown displacement
(Fig. 6) confirm that a ferrule preparation increases
the mechanical resistance of the restoration (Tjan &
Whang 1985, Barkhordar et al. 1989, Sorensen &
Engelman 1990) by opposing displacement. This
becomes most efficient when the ferrule height is
greater than the rotation radius of the crown
(Fig. 13). Although a short ferrule may not efficiently
resist displacement, it does however reduce the axial
arm of the rotation force and this results in reducing
the bending force on the post with subsequent
reduction of stress within the post and the axial
displacement force. This reduces the potential for
debonding of the lute and post-fracture.
There were two main changes in the stress distribu-
Figure 11 Plots of the compressive stress on the (a) internal tion in root dentine associated with an increase in
and (b) external buccal root dentine. ferrule height, a decrease in compressive stress in labial
cervical dentine (Fig. 11) at levels below the compres-
much less than the tensile strength of dentine sive strength of dentine (297 MPa, Craig & Peyton
(185 MPa, Sano et al. 1994) and lower than com- 1958) and an increase in tensile stress in palatal
pressive stress values (Fig. 11) suggesting that there cervical dentine (Fig. 7a) to values approaching the

Intens

300.000
275.000
250.000
225.000
200.000
175.000
150.000
125.000
100.000
75.000
50.000
25.000
0.000

Figure 12 Diagram of a nonferrule model demonstrating the rotational movement that the post/core/crown will be subjected to
under load (dashed arrow), the area of labial dentine under compressive stress (solid inward arrows), the area of palatal dentine
under tensile stress (solid outward arrows) and tensile stress within the coronal portion of the post.

450 International Endodontic Journal, 39, 443–452, 2006 ª 2006 International Endodontic Journal
 Ichim et al. FEA analysis of ferrule designs

Figure 13 Diagram illustrating contact-induced displacement for (a) 0.5 mm and (b) 2 mm ferrule models. The arrows indicate
the rotation radius of the crown and demonstrate that a 0.5 mm ferrule is less than the rotational radius of the crown while a
2 mm ferrule exceeds it and offers more mechanical resistance.

tensile strength of dentine (185 MPa, Sano et al. 1994) may have resulted in higher stresses in the root.
(Fig. 7b) with an associated increased area of palatal Similarly the elastic limit of the PDL was set at
dentine under stress (Figs 8 and 14). This is probably 11.8 MPa and modelled as an elastic material while
due to the palatally applied force transmitting a flexing recent work has shown that the PDL is visco-elastic
force to the palatal wall of the ferrule leading to the with a modulus of around 1 MPa (Bourauel et al.
concentrated area of stress at the base of the palatal 1999). In the present study, the PDL values were set so
ferrule wall (Fig. 14). In the present study, the stress as to simulate the amount of displacement of a root
value in the cervical mid-dentine approached the that could occur clinically and the structural stresses
tensile stress for dentine (Fig. 7b) suggesting favourable were determined through the solid bodies (dentine, post
conditions for a crack to nucleate, which would run and crown) of the model. As such a moderate change
apically and obliquely along the palatal root dentine (increase of decrease) in the elastic modulus of the PDL
from the internal to the external root surface. These would not alter the stress results inside the dentine,
results may explain the mechanisms behind the differ- post and crown.
ent failure patterns observed by other researchers The results of the present study show that a
(Saupe et al. 1996). simulated crown-lengthening procedure and ferrule
Finite element models provide an insight into the preparation did not change the levels or pattern of
processes involved when loading teeth but do not fully stress within the models compared with conventional
represent the clinical situation and limitations of the ferrule preparations and does not support the notion
model should be recognised. For example in this model that crown-lengthening may result in a tooth with less
no alveolar bone was modelled and the model was fixed ability to withstand load (Gegauff 2000), however
on the outside surface of the periodontal ligament so as methodology differences between studies may explain
to represent the effect of the inner socket wall and this these differences.

(a) (b)
Intens Intens
250.000 250.000
229.175 229.167
200.350 200.333
187.525 157.500
166.700 166.667
145.875 145.833
125.050 125.000
104.226 104.167
83.401 83.333
62.576 62.500
41.751 41.667
20.926 20.833
0.101 0.000

Figure 14 Diagram of the principle stresses in (a) 0.5 mm and (b) 2 mm ferrule models. Changes in the principle stress pattern
with increasing ferrule height are evident as a decrease in compressive stress in labial dentine and within the root canal and
increased area of tensile stress in palatal dentine especially around the palatal ferrule wall (the post has been removed for clarity).

ª 2006 International Endodontic Journal International Endodontic Journal, 39, 443–452, 2006 451
 FEA analysis of ferrule designs Ichim et al.

Conclusion Guzy GE, Nicholls JI (1979) In vitro comparison of intact

endodontically treated teeth with and without endo-post
Using FEA, the study shows that a ferrule increases reinforcement. Journal of Prosthetic Dentistry 42, 39–44.
the mechanical resistance of a post/core/crown res- Isidor F, Brondum K, Ravnholt G (1999) The influence of post
toration by reducing the potential for displacement length and crown ferrule length on the resistance to cyclic
(labial and axial rotation) and compressive stresses loading of bovine teeth with prefabricated titanium posts.
within labial dentine and the canal wall. The results International Journal of Prosthodontics 12, 78–82.
suggest that the ferrule height should be determined Libman WJ, Nicholls JI (1995) Load fatigue of teeth restored
individually for each case based on the bucco-lingual with cast posts and cores and complete crowns. International
cervical diameter of the root. As a drawback the Journal of Prosthodontics 8, 155–61.
Milot P, Stein RS (1992) Root fracture in endodontically
presence of the ferrule creates a larger area of palatal
treated teeth related to post selection and crown design.
dentine under tensile stress, which may be a favour-
Journal of Prosthetic Dentistry 68, 428–35.
able condition for a crack to develop on the palatal Mühlemann HR (1960) 10 years of tooth mobility measure-
aspect of the root eventually leading to an oblique ments. Journal of Periodontology 31, 110–22.
root fracture. By comparison, a restoration without Ng CC, al-Bayat MI, Dumbrigue HB, Griggs JA, Wakefield CW
ferrule is prone to fail primarily by debonding and (2004) Effect of no ferrule on failure of teeth restored with
subsequently by root fracture through the lever bonded posts and cores. General Dentistry 52, 143–6.
action of the loose post. Crown-lengthening did not Rosenstiel SF, Land MF, Fujimoto J (2001) Contemporary Fixed
alter the levels or pattern of stress when compared Prosthodontics, 3rd edn. St Louis, MO, USA: Mosby.
with standard ferrule preparations. Sano H, Ciucchi B, Matthews WG, Pashley DH (1994) Tensile
properties of mineralized and demineralized human and
bovine dentin. Journal of Dental Research 73, 1205–11.
References Saupe WH, Gluskin AH, Radke RJ (1996) A comparative study
of fracture resistance between morphologic dowel and cores
Barkhordar RA, Radke R, Abbasi J (1989) Effect of metal
and resin-reinforced dowel system in the intraradicular
collars on resistance ofendodontically treated teeth to root
restoration of structurally compromised roots. Quintessence
fracture. Journal of Prosthetic Dentistry 61, 676–8.
International 27, 483–91.
Bourauel C, Freudenreich D, Vollmer D, Kobe D, Drescher D,
Sorensen J, Engelman M (1990) Ferrule design and fracture
Jager A (1999) Simulation of orthodontic tooth movements.
resistance of endodontically treated teeth. Journal of Pros-
A comparison of numerical models. Journal of Orofacial
thetic Dentistry 63, 529–36.
Orthopaedics 60, 136–51.
Sorensen JA, Martinoff JT (1984) Intracoronal reinforcement
Craig RG, Peyton FA (1958) Elastic and mechanical properties
and coronal coverage: a study of endodontically treated
of human dentin. Journal of Dental Research 37, 710–8.
teeth. Journal of Prosthetic Dentistry 51, 780–84.
Gegauff AG (2000) Effect of crown lengthening and
Tjan AH, Whang SB (1985) Resistance to root fracture of
ferrule placement on static load failure of cemented cast
dowel channels with various thicknesses of buccal dentin
post-cores and crowns. Journal of Prosthetic Dentistry 84,
walls. Journal of Prosthetic Dentistry 53, 496–500.
169–79.

452 International Endodontic Journal, 39, 443–452, 2006 ª 2006 International Endodontic Journal