PRD 22 5 Magne1
PRD 22 5 Magne1
COPYRIGHT © 2002 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY. NO PART OF THIS ARTICLE MAY BE
REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
3
REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
COPYRIGHT © 2002 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY. NO PART OF THIS ARTICLE MAY BE
Pascal Magne, Dr Med Dent*/Urs C. Belser, Prof Dr Med Dent** Reproducing the original perfor-
mance of the intact tooth (“bio-
This study considered the loading configuration of intact teeth by using finite ele-
mimetics”) should be a driving force
ment analyses to rationalize the clinical and biologic advantage inherent to posteri-
in restorative dentistry.1 It requires
or tooth shape. The biomechanical behavior of opposing molars was investigated
fundamental understanding of the
in consideration of different loadcases simulating working, nonworking, and vertical
natural tooth with regard to its inti-
closing micromotions starting in a position close to maximum intercuspation. The
resulting stress distribution was assessed in a numerical model, reproducing 2-D mate structure, biology, morphol-
buccolingual cross sections of maxillary and mandibular molars. In each case (work- ogy, and external shape (Fig 1). For
ing/nonworking/closure), the stroke was applied to the mandibular tooth in a step- posterior teeth, chief advances have
ping procedure (nonlinear contact analysis) until a total external force of 200 N was resulted from the study and under-
attained on the contact nodes. The principal stress distribution and modified Von standing of cuspal flexure and plas-
Mises stresses were extracted from the postprocessing files. Vertical loading of the tic yielding, which represent key
teeth did not generate harmful concentrations of stress. More challenging situa- parameters in the performance of
tions were encountered during working and nonworking micromotions, both of the tooth-restorative complex.2,3
which generated inverted stress patterns. Supporting cusps were generally well Subclinical cuspal microdeformation,
protected during both working and nonworking cases (mostly subjected to com-
ie, below the threshold of chairside
pressive stresses). Nonsupporting cusps tended to exhibit more tensile stresses.
observation, was identified in the
High stress levels were found in the central groove of the maxillary molar during
early 1980s,2,4,5 and it is now well
nonworking micromotion and at the lingual surface of enamel of the mandibular
accepted that intact teeth demon-
tooth during single-contact working micromotion. The occlusal load configuration
as well as geometry and hard tissue arrangement had a marked influence on the strate cuspal flexure because of their
stress distribution within opposing molars. Additional computations demonstrated morphology and occlusion.
the essential role of enamel bridges and crests to protect the crown from harmful Restorative procedures can increase
tensile stresses. (Int J Periodontics Restorative Dent 2002;22:xxx–xxx.) cuspal movement under occlusal
load,2,6 which may result in altered
**Senior Lecturer, Department of Prosthodontics, School of Dental Medicine, strength, fatigue fracture, and
University of Geneva, Switzerland; and Adjunct Associate Professor,
Department of Oral Science, University of Minnesota, Minneapolis.
cracked tooth syndromes.7–9 Such
**Professor and Chair, Department of Prosthodontics, School of Dental knowledge allowed the develop-
Medicine, University of Geneva, Switzerland, ment of methods to improve fracture
resistance of teeth10,11 through var-
**Reprint requests: Dr Pascal Magne, University of Geneva, School of Dental
Medicine, Department of Prosthodontics, 19 rue Barthélemy-Menn, CH- ious forms of complete or partial
1205 Genève, Switzerland. e-mail: Pascal.Magne@medecine.unige.ch coverage12–14 and by introducing
the more conservative adhesive ture. By solving the deformation of Current literature provides little
techniques.3,15,16 all the small elements simultane- information about the effect of real-
Most of the in vitro investiga- ously, the deformation of the struc- istic biting configurations on the bio-
tions mentioned above were carried ture as a whole can be assessed. mechanical behavior of posterior
out using nondestructive strain Using the traditional biophysical teeth. Special attention was there-
gauge methods along with load-to- knowledge database in a rational fore given to the simulation of
failure tests. These traditional “load- validation process, FE analysis has opposing contacts during micromo-
point” experiments provide insights been significantly refined during tions including clenching, working,
into a number of biomechanical recent years.17 Nowadays, experi- and nonworking excursions.
issues, yet they do not reveal the mental-numerical approaches
stress distribution within the tooth- undoubtedly represent the most
restoration complex during biting comprehensive in vitro investigation Method and materials
and clenching. Knowledge of stress methods. The present work used
distribution is of paramount impor- two-dimensional FE models, the Mesh generation and material
tance in the biomimetic approach accuracy of which, considered in a properties (preprocessing)
(especially in the optimization of buccolingual cross section, has been
adhesive restorative techniques), but demonstrated and validated on sev- Buccolingual cross sections of nat-
requires complex modeling tools eral occasions by experimental strain ural maxillary and mandibular molars
such as the finite element (FE) measurements on both anterior and were digitized using a charge cou-
method. In FE analysis, a large struc- posterior teeth.1,4,5,18 pled device camera (Sony DXC-
ture is divided into a number of small The aim of this study was to 151A) attached to a stereomicro-
simple-shaped elements, for which describe the biomechanical re- scope (Olympus SZH10) and an
individual deformation (strain and sponse of intact opposing molars in image-analysis software program
stress) can be more easily calculated terms of stress and strain distribution (Optimas 5.22). The contours of the
than for the entire undivided struc- based on 2-D FE simulations. enamel, dentin, and pulp areas were
manually traced using a PC worksta- Versluis et al,19 a correct ratio of displacement of a rigid base (non-
tion and graphic software (Freelance moduli (enamel:dentin) is necessary deformable body) glued to the cut
Graphics, Lotus). An image-process- for a qualitative analysis. Moduli of plane of the root. Contact between
ing program (NIH Image, developed 50 GPa and 12 GPa were chosen for bodies in Mentat 2000 software was
at the Research Services Branch of enamel and dentin, respectively, controlled according to the deform-
the National Institute of Mental yielding a ratio of 4.2. Poisson’s ratios able-deformable method (“double-
Health) was used to record the coor- of 0.23 for dentin20 and 0.30 for sided,” ie, presence of two deform-
dinates of all of the structures and enamel21 were assumed. able bodies) in a static mechanical
defined contours. These geometric loadcase with a stepping procedure
data were then transferred to an (50 steps). Preliminary testing
interactive FE program for mesh Boundary conditions, loadcase, included the application of Coulomb
generation and preprocessing and data processing friction (enamel friction coefficient
(Mentat 2000, MSC Software). of 0.4), but this feature was ultimately
Although teeth are 3-D structures, a Contours of opposing teeth were abandoned because no difference
2-D FE model with plane-strain ele- positioned as close as possible to could be detected, probably be-
ments (linear, four-node, isopara- maximum intercuspation, but with- cause of the very limited sliding of
metric, and arbitrary quadrilateral) out contacts. Teeth were defined as the cusps. In a first approach, four
was chosen (Fig 2). A 3-D model, deformable contact bodies. Fixed different loadcases were simulated:
although more realistic, would have zero displacement in both the hori-
resulted in coarser meshes. The zontal and vertical directions was 1. Working micromotion on two
increased memory requirements for assigned to the cut plane of the root contacts. The movement of the
3-D models combined with the non- of the maxillary molar, approximately base simulated a working motion
linear nature of this analysis would 1.5 mm beyond the cementoenamel (“W” in Fig 2): three steps of 16
not have allowed fine representation junction. The motion was applied to µm were required to reach con-
of the geometry. As reported by the mandibular tooth through the tact in two locations (Fig 3a). The
a b c d
Fig 3 First principal stresses within the molar cross sections for each loadcase. Negative val-
ues of stress appear in gray and delineate the areas of compressive stresses. Color shadings
indicate the different levels of tensile stresses. In each loadcase (a to d), the sum of external
forces on the contact nodes is ≈ 200 N. Note the tensile stresses at the lingual surface in Fig 3b
(arrowheads). * = area of contact nodes
motion continued for nine addi- reach a total force of ≈ 200 N on Results
tional steps (144 µm) to reach a the contact nodes.
total force of ≈ 200 N on the con- 4. Vertical micromotion. The move- In a systematic approach to under-
tact nodes. ment of the base simulated a ver- stand the deformation mode of the
2. Working micromotion on one tical closure (“C” in Fig 2): seven tooth, it is appropriate to analyze
contact. The morphology of the steps of 8 µm were required to stresses in a direction for which the
lingual cusp of the mandibular reach contact in three locations x and y components of stresses will
molar was slightly modified to (Fig 3d). The motion continued display their maximum values. The
avoid contact during the working for five additional steps (40 µm) resulting analysis (Fig 3) outlines the
motion: three steps of 16 µm to reach a total force of ≈ 200 N principal stresses in the form of areas
were required to reach contact on the contact nodes. At all times of compression and tension.
between the buccal cusps only in this specific case, the man-
(Fig 3b). The motion continued dibular molar was allowed to
for 10 additional steps (160 µm) move laterally on its base (along Working motions
to reach a total force of ≈ 200 N the x axis), which ultimately pro-
on the contact nodes. vided an optimal centering in Both loadcases simulating a working
3. Nonworking micromotion. The maximum intercuspation. motion (Figs 3a and 3b) generated
movement of the base simulated on the supporting cusps compres-
a nonworking motion (“NW” in The stress distribution within sive stress in the enamel and tension
Fig 2): four steps of 16 µm were both molars was solved using the in the major area of dentin. However,
required to reach contact (Fig MARC 2000 Analysis solver (MSC a definite portion of enamel at the
3c). The motion continued for 10 Software). The postprocessing file nonsupporting cusps (buccal surface
additional steps (160 µm) to was accessed through Mentat. of the maxillary tooth and lingual sur-
Fig 4 Modified Von Mises stresses (MPa) for each loadcase. Path plot proceeds along the
enamel surface (dotted arrows) from the lingual/palatal cementoenamel junction (A) to the
buccal side (B). Note the stress peak at the sulcus of the maxillary tooth during the nonwork-
ing motion (red curve) and the elevated stresses at the buccal cervical aspect in the same
curve (*, left plot). Note also the stress peak at the lingual surface of the mandibular molar
during the working motion with a single contact (*, right plot, gray curve).
The biomimetic approach physical functions of biologic “com- General stress patterns
posites” and the designing of new
The results presented here might be and improved substitutes. In restora- Inversely to anterior teeth, cusps do
questioned because they have been tive dentistry, it starts with the under- not deform under load as simple
produced in an FE environment. The standing of hard tissue arrangement cantilever beams.30 The deforma-
methods used in this study, how- and related stress distribution within tion mode is complicated by the
ever, are based on several preexist- the intact tooth.1,28 Chief advances numerous possibilities in the appli-
ing validation studies that have have resulted from such an ap- cation of loads (working, nonwork-
proven the relevance of these con- proach in the field of anterior ing, closure). General assumptions
cepts.1,4,5,18 Even though some dif- bonded ceramic restorations.29 New claiming the harmful effect of lateral
ferences can remain between reality restorative approaches do not aim loading have been confirmed in the
and the FE environment, there are to create the strongest restoration, present work. Vertical loading of the
still at least two reasons that justify but rather a restoration that is com- tooth (in the direction of its main
the use of numerical modeling: (1) it patible with the mechanical and axis) did not generate harmful con-
is able to reveal the otherwise inac- biologic properties of underlying centrations of stress. More chal-
cessible stress distribution within the dental tissues—the biomimetic prin- lenging situations were encountered
tooth-restoration complex, and (2) it ciple. Enamel, dentin, and modern during working and nonworking
has proven to be an essential tool in dental materials (ceramics, com- micromotions, both of which gener-
the thinking process for the under- posite resins) are characterized by ated inverted stress patterns. From
standing of tooth biomechanics and their brittle behavior. It is of para- the current analysis, it appears that
the biomimetic approach. Biomime- mount importance to locate harmful supporting cusps are generally well
tics is a newly emerging interdisci- stresses during function, which con- protected during both working and
plinary material science27 involving stituted one aim of the present nonworking loadcases (mostly sub-
investigation of both structures and study. jected to compressive stresses).
Nonsupporting cusps tend to exhibit motion in Fig 4). This situation must two extreme geometric designs on
more tensile stresses. The most sur- be evaluated in view of additional the mVM stresses. The results are
prising effect was found in the form data to understand the effect of presented in Fig 7: Enamel bridging
of a stress peak at the lingual surface geometry and shape on the stress not only reduces the stresses locally
of enamel (mandibular molar) in the distribution. An experimental- (see the sulcus of the maxillary
single-contact working case. It can numerical study carried out on ante- molar), but can also protect distant
be explained by the massive oblique rior teeth31 demonstrated low stress enamel (see the lingual enamel of
compression of dentin in the lingual levels in enamel surfaces of maxi- the mandibular tooth). The denti-
cusp. The surface of enamel must mum convex curvature, namely the noenamel junction (DEJ) constitutes
stretch to conform to the buckling of cingulum and the cervical part of the another crucial element that must
the cusp (Fig 5). The presence of a facial surface of incisors. It was con- be mentioned among the natural
second contact between the palatal cluded that convex surfaces with protective mechanisms of the tooth.
and lingual cusps significantly thick enamel raise less concentrated The crack-arresting effect of dentin
reduced that phenomenon by trans- stresses than concave areas, which and of the thick collagen fibers at the
ferring tension in the central groove. tend to concentrate stresses. These DEJ32 compensate for the inherently
principles can be applied to poste- brittle nature of enamel. The com-
rior teeth. Figure 5 illustrates the vari- plex fusion found at the DEJ can be
Natural protective mechanisms able morphology of the central sul- regarded as a fribril-reinforced
of the intact tooth cus within the same tooth, which can bond, 33 and scanning electron
exhibit a marked crest or enamel microscopic fractographs of DEJ
High levels of mVM stresses were bridge (Fig 6a) or a deep fissure with specimen have demonstrated crack
found at certain locations in enamel, an extreme concavity (Fig 6b). deflection to another fracture plane
especially in the central groove (max- Additional computations were car- when forced through the DEJ
illary tooth during nonworking ried out to explore the effect of these zone.32
a b
Fig 6 The same tooth can display extreme morphologic types within the occlusal table, either
with an enamel bridge or crest (a) or a deep fissure (b) according to the cross-sectional area.
These elements were considered for additional computations presented in Fig 7.
Fig 7 Modified Von Mises stresses (MPa) for the two additional loadcases with altered
central grooves (crest and fissure): (I) maxillary molar during nonworking motion, and (II)
mandibular tooth during working motion with a single contact. Same path plot as in Fig 4 (A
to B). Note marked reduction of stress peak at sulcus of maxillary tooth in the presence of
the crest (red curve). Note also the reduction of stresses at lingual surface of mandibular
molar because of the distant presence of the enamel bridge.
• High stress levels were found in 8. Cameron CE. The cracked tooth syn- 21. Craig RG. Restorative Dental Materials,
drome: Additional findings. J Am Dent ed 6. St Louis: Mosby, 1980:76.
the central groove of the maxil-
Asssoc 1976;93:971–975.
lary molar during nonworking 22. De Groot R, Peters MCRB, De Haan YM,
9. Cavel WT, Kelsey WP, Blankenau RJ. An Dop GJ, Plasschaert AJM. Failure stress
micromotion and at the lingual in vivo study of cuspal fracture. J Prosthet criteria for composite resin. J Dent Res
surface of enamel of the man- Dent 1985;53:38–42. 1987;66:1748–1752.
dibular tooth during single-con- 10. Hood JAA. Methods to improve fracture 23. Pameijer JH, Glickman I, Roeber FW.
tact working micromotion. resistance of teeth. In: Vanherle G, Smith Intraoral occlusal telemetry. 3. Tooth con-
• Enamel bridges and crests prov- DC (eds). International Symposium on tacts in chewing, swallowing and bruxism.
Posterior Composite Resin Restorative J Periodontol 1969;40:253–258.
ed to be essential mechanisms to Materials. St Paul, Minn, 1985:443–450.
24. Gibbs CH. Electromyographic activity
protect crown biomechanics. [AU: Who is the publisher?]
during the motionless period in chew-
11. Douglas WH. Methods to improve frac- ing. J Prosthet Dent 1975;34:35–40.
ture resistance of teeth. In: Vanherle G,
25. Miura H, Watanabe S, Isogai E, Miura K.
Acknowledgments Smith DC, eds. International Symposium
Comparison of maximum bite force and
on posterior composite resin restorative
dentate status between healthy and frail
materials. St Paul, Minn, 1985:433–441.
This study was supported in part by the Minne- elderly persons. J Oral Rehabil 2001;28:
sota Dental Research Center for Biomaterials 12. Malcolm PJ, Hood JAA. The effect of cast 592–595.
and Biomechanics (image processing and restorations in reducing cusp flexibility in
26. Nishigawa K, Bando E, Nakano M. Quan-
numerical modeling) and the Swiss Foun- restored teeth. J Dent Res 1971;56:D207.
titative study of bite force during sleep
dation for Medical-Biological Grants. The 13. Reeh ES, Douglas WH, Messer HH. associated bruxism. J Oral Rehabil 2001;
authors wish to express their gratitude to Dr Stiffness of endodontically-treated teeth 28:485–491.
Antheunis Versluis for his help in refining the related to restoration technique. J Dent
27. Sarikaya M. An introduction to bio-
finite element modeling and numerical output. Res 1989;68:1540–1544.
mimetics: A structural viewpoint. Microsc
14. Linn J, Messer HH. Effect of restorative Res Technol 1994;27:360–375.
procedures on the strength of endodon-
28. Magne P, Douglas WH. Rationalization of
References tically treated molars. J Endod 1994;20:
esthetic restorative dentistry based on
479–485.
biomimetics. J Esthet Dent 1999;11:5–15.
1. Magne P, Douglas WH. Rationalization of 15. McCullock AJ, Smith BG. In vitro studies
29. Magne P, Perroud R, Hodges JS, Belser U.
esthetic restorative dentistry based on bio- of cusp reinforcement with adhesive
Clinical performance of novel-design
mimetics. J Esthet Dent 1999;11:5–15. restorative material. Br Dent J 1986;161:
porcelain veneers for the recovery of coro-
450–452.
2. Douglas WH. Considerations for model- nal volume and length. Int J Periodontics
ing. Dent Mater 1996;12:203–207. 16. Macpherson LC, Smith BG. Reinforce- Restorative Dent 2000;20:441–457.
ment of weakened cusps by adhesive
3. Morin D, Delong R, Douglas WH. Cusp 30. Jantarat J, Panitvisai P, Palamara JE,
restorative materials: An in-vitro study. Br
reinforcement by the acid-etch technique. Messer HH. Comparison of methods for
Dent J 1995;178:341–344.
J Dent Res 1984;63:1075–1078. measuring cuspal deformation in teeth. J
17. Korioth TWP, Versluis A. Modeling the Dent 2001;29:75–82.
4. Morin DL, Douglas WH, Cross M, Delong
mechanical behavior of the jaws and their
R. Biophysical stress analysis of restored 31. Magne P, Versluis A, Douglas WH. Ration-
related structures by finite element (FE)
teeth: Experimental strain measurements. alization of incisor shape: Experimental-
analysis. Crit Rev Oral Biol Med
Dent Mater 1988;4:41–48. numerical analysis. J Prosthet Dent 1999;
1997;8:90–104.
81:345–355.
5. Morin DL, Cross M, Voller VR, Douglas
18. Magne P, Douglas WH. Porcelain veneers:
WH, Delong R. Biophysical stress analysis 32. Lin CP, Douglas WH. Structure-property
Dentin bonding optimization and bio-
of restored teeth: Modeling and analysis. relations and crack resistance at the
mimetic recovery of the crown. Int J
Dent Mater 1988;4:77–84. bovine dentin-enamel junction. J Dent
Prosthodont 1999;12:111–121.
Res 1994;73:1072–1078.
6. Assif D, Marshak BL, Pilo R. Cuspal flexure
19. Versluis A, Douglas WH, Cross M,
associated with amalgam restorations. J 33. Lin CP, Douglas WH, Erlandsen SL. Scan-
Sakaguchi RL. Does an incremental filling
Prosthet Dent 1990;63:258–262. ning electron microscopy of type I colla-
technique reduce polymerization shrink-
gen at the dentin-enamel junction of hu-
7. Cameron CE. The cracked tooth syn- age ? J Dent Res 1996;75:871–878.
man teeth. J Histochem Cytochem 1993;
drome. J Am Dent Assoc 1964;68:
20. Watts DC, Mowafy OM, Grant AA. 41:381–388.
405–411.
Temperature-dependence of compres-
sive properties of human dentin. J Dent
Res 1987;66:29–32.