A

PoIMifmtR“mCt©

K

1

otmnr

1d

Available online at www.sciencedirect.com

Medical Engineering & Physics 30 (2008) 218–225

A complete finite element model of a mandibular implant-retainedoverdenture with two implants: Comparison between rigid and

resilient attachment configurations

M. Daas a, G. Dubois b,c, A.S. Bonnet d,∗, P. Lipinski d, C. Rignon-Bret e

a Faculty of Dental Surgery, Rene Descartes University, Paris V, 92000 Montrouge, Franceb Laboratory of Physics and Mechanics of Materials, Ecole Nationale d’Ingenieurs de Metz, Ile du Saulcy, 57045 Metz Cedex, France

c Orthopedie Biomecanique Locomotion, 92320 Chatillon, France.d Laboratoire de Fiabilite Mecanique, Ecole Nationale d’Ingenieurs de Metz, Ile du Saulcy, 57045 Metz Cedex, France

e Faculty of Dental Surgery, Rene Descartes University, Paris V, 92000 Montrouge, France

Received 14 March 2006; received in revised form 9 February 2007; accepted 12 February 2007

bstract

urpose: The aim of this study was to evaluate the influence of the retention mechanism on the behavior of a mandibular implant-retainedverdenture (IRO) during the simulation of mastication. Therefore, a complete three-dimensional finite element model of a mandible with itsRO was developed.

aterials and methods: The geometry of the edentulous mandible and overdenture was generated from computed tomography. Two MKIII®

mplants (Nobel Biocare) with ball abutments and Dalbo Plus® (Cendres et Metaux) attachments were placed in the canine areas. Threeoodstuff positions were analyzed for two retention mechanisms, “resilient” or “rigid”. Special attention was given to the modeling of theandibular environment and of the existing contact between the different components. A probable muscular action was determined following

he minimal work principle.esults: The food-crushing force was provided by masseters with a two-third/one-third ratio between working and non-working sides. Theresilient” configuration provided a wider contact area between the mucosa of the denture bearing area and the prosthesis. An increase of theastication force transiting through the mucosa was also noted and lower stresses were observed in the bone surrounding implants.

onclusion: Resilient attachments allowed for an increase of the mastication load transiting through denture bearing surface. Furthermore,

his study proposed an accurate model of the mandibular IRO, including its environment and faithful behavior reproduction.2007 IPEM. Published by Elsevier Ltd. All rights reserved.

eywords: Biomechanics; Dental implants; Mandible; Overdenture attachment; Finite element model

bdHao

. Introduction

The treatment of a fully edentulous mandible by meansf implant-retained overdenture (IRO) has become a rou-ine therapy [1,2]. Many different attachments available todayay be used to support IRO. Currently, mandibular IRO con-

ected to two implants with ball attachments has become aeliable and well-documented treatment [3–7].

∗ Corresponding author. Tel.: +33 387344264E-mail address: bonnet@enim.fr (A.S. Bonnet).

tttrtr

350-4533/$ – see front matter © 2007 IPEM. Published by Elsevier Ltd. All rightoi:10.1016/j.medengphy.2007.02.005

Moreover, some authors reported overdentures supportedy a two-implant ball system result in a more favorable stressistribution in bone than a two-implant bar system [8–13].owever, various options exist about whether a “rigid” or“resilient” retention mechanism should be used. The use

f either retention mechanism is a subject of controversy inhe literature [8,14]. It was often claimed that resilient reten-ion mechanisms for overdenture stabilization should be used

o distribute tissue and implant support [15]. Other authorseported that there is no significant difference between thewo retention options and suggested a slight superiority for aigid system [14,16].

s reserved.

eering & Physics 30 (2008) 218–225 219

bawuams“ltt

mtvtaItbutalmswasotom

2

2

M1c(DamabptgdIf

F“

2

tehtootea3smc

mgatUraomC

2

a

M. Daas et al. / Medical Engin

Beyond clinical trials, other investigative methods haveeen used to address this controversy. Mericske-Stern [17], inn in vivo study on two implants, observed that lower forcesere recorded on both implants in all three directions andnder various test conditions for resilient configuration. Someuthors used a photoelastic model of a human edentulousandible to compare stress distribution with different sub-

tructure designs and in two configurations: “resilient” andrigid” [8,13]. Other studies developed models of a mandibu-ar IRO using strain gauges to compare several types of reten-ion mechanisms [18,19]. They concluded that resilient sys-ems induced the lowest forces and moments on the implants.

As components of the dental implant-bone system, attach-ents devices and prosthesis are geometrically complex,

hree-dimensional finite element analysis (FEA) has beeniewed as the most suitable tool to provide an answer tohis controversy. Menicucci et al. [20] used such a modelnd concluded that stress distribution with ball anchors forRO was more favorable than the bar/clips attachment. Ahree-dimensional finite element analysis was also achievedy Chun et al. [14] to study different overdenture attachmentsnder vertical and inclined loads. They used a simplifiedhree-dimensional geometry of the mandible, overdenturend mucosa. The main goal of the present study was to ana-yze the influence of the attachment resilience on overdenture

otions, load repartition between implants and mucosa, andtress states in both bone and implants. Six configurationsere analyzed corresponding to three foodstuff positions

nd two retention mechanisms. To this end, it was neces-ary to elaborate a precise and reliable finite element modelf a mandible equipped with an implant-retained overden-ure. Special attention was given to an accurate modelingf muscular actions, temporo-mandibular joint and contactanagement between the different components.

. Material and methods

.1. Implants and prosthetic components

The prosthetic solution retained was supported by twoKIII® implants (Nobel Biocare, diameter: 3.75 mm, length:

1.5 mm) located in the canine area. The overdenture wasonnected to the implants by means of two ball abutmentsNobel Biocare, diameter: 2.25 mm, length: 3 mm) and twoalbo Plus® matrices (Cendres et Metaux). Dalbo Plus

ttachments used here can lead to two different retentionechanisms depending on how they are set up. These mech-

nisms differ by the presence of a 0.4 mm axial gap that cane obtained by setting a spacer disc between male and femalearts during overdenture polymerization. In the present study,hese two configurations (named “rigid” for the case with no

ap and “resilient” for the other one) were compared. Theifference between these two models is presented in Fig. 1.t can be noted that in both cases, rotation between male andemale parts was not restrained.

o(

c

ig. 1. View of the attachment’s relative position in both “rigid” (a) andresilient” configurations (b).

.2. Meshing

The first step to model the mandible, the overdenture andheir environment was to obtain the geometry of a totallydentulous patient’s mandible. This patient was chosen foris “average” anatomy according to considerations in litera-ure [21]. A computed tomography examination was carriedut on the patient equipped with a radio-opaque duplicatef his overdenture, to get the accurate geometry of the pros-hesis and mandible. The files of the computed tomographyxamination were then imported into two software pack-ges namely Mimics 7.3 (Materialise, Leuven, Belgium) andMatic 2 (Materialise, Leuven, Belgium). They allowed con-truction of the three-dimensional geometry of the edentulousandible and overdenture and to export it into a file type that

an be exploited by classical CAD systems.Furthermore, knowing the actual relative position of the

andible and overdenture thanks to its duplicate, the exacteometry of the mucosa was deduced but considered onlyround the overdenture. The geometry was later meshed usinghe HyperMesh 7 software package (Altair, Troy, Michigan,SA). The interforaminal region was distinguished from the

est of the mandible. A refined mesh was needed in thisrea to reproduce faithfully the complex stress state generallybserved in the peri-implant bone. Implants and ball attach-ents were modeled using SolidWorks 2003 (SolidWorksorporation, Concord, Massachusetts, USA).

.3. Material properties

Implants were made of Ti6Al4V titanium alloy and ballbutments of grade 4 titanium. The caps were constituted

f two components: lamellae’s made of a gold-based alloyElitor) screwed into a grade 4 titanium housing.

The interforaminal region of the edentulous mandible wasomposed of a 2-mm constant cortical bone layer around a

220 M. Daas et al. / Medical Engineering

ctilVoanptpi

2

aiwtosjus

TM

B

CCMOTTE

dsctFcawf

2

mmll

mgptstmcbgm[ari

rfiguration was assumed for position (C) and a symmetricalone was supposed for position (I). The muscular actions weregenerated in order to obtain reaction forces on foodstuff of100 N in case (M), 55 N in case (C), and 40 N in case (I). These

Fig. 2. Global view of meshing.

ancellous bone core (Fig. 2). Elsewhere, bone was assumedo be homogenous. The material properties of bone out of thenterforaminal region were deduced by averaging the cancel-ous and cortical bone properties according to the classicaloigt [22] and Reuss [23] models with assumed proportionsf 60% and 40%, respectively. All these values were chosenccording to measures on the computed tomography exami-ation. The same method was applied to obtain overdentureroperties from those of resin and ceramic. Material proper-ies for cortical and cancellous bones, mucosa and titaniumarts were taken from literature [14,22,24,25] and are recalledn Table 1. All materials were assumed isotropic.

.4. Contact management

Contact was introduced between overdenture and mucosa,nd between both parts of the attachments to simulate thenteractions existing between these bodies. Contact at anchorsas essential to reproduce the behavior of the two reten-

ion mechanisms studied. Implants were considered totallysseointegrated, so a mechanically perfect interface was pre-umed between implants and bone. The temporo-mandibular

oint was modeled by a contact between rigid surfaces sim-lating mandibular fossae, and condyle elements. A rigidurface was also used to represent foodstuff in order to

able 1aterial properties

Young’s modulus (MPa) Poisson’s ratio

one out of theinterforminal region

4500 0.3

ortical bone 13700 0.3ancellous bone 1370 0.3ucosa 1 0.37verdenture 4500 0.35i6Al4V 135000 0.3itanium Grade 4 114000 0.3litor 97000 0.42 F

f

& Physics 30 (2008) 218–225

escribe precisely its interaction with overdenture. Thispherical surface had a 20 mm-diameter to avoid a localizedontact. The main interest using rigid surface models waso limit calculation time and facilitate contact management.oodstuff position was changed to simulate different bitingonfigurations: foodstuff on first molar (M), on canine (C)nd on incisors (I). For (M) and (C) configurations, foodstuffas located on the right side. A sagittal medial position of

oodstuff was chosen for the (I) configuration.

.5. Anatomic considerations

Special attention was given to accurate modeling ofuscular actions, temporo-mandibular joint and contactanagement between the different parts. Indeed, there is very

ittle data concerning muscular action during mastication initerature [26,27].

The four elevator muscular groups were modeled byeans of truss elements for masseters and medial ptery-

oids and of membrane elements for temporalis and lateralterygoids (Fig. 3). Muscular action was simulated by con-raction of these elements. Elements simulating tendons wereet up at the muscular insertion regions. Twenty-four calcula-ions were performed in order to determine possible working

uscle configurations. Indeed, each of the four elevator mus-ular groups was activated separately and six load repartitionsetween working and non-working sides were tested for eachroup. In these cases, foodstuff was located above the firstolar, representing the most frequent masticating situation

17,20]. The most probable muscular action was determinedccording to the minimal work principle. Once this configu-ation was defined, it was applied to remaining calculationsn case of position (M).

“Rigid” and “resilient” configurations were studied withespect to (M), (C) and (I) positions. The same muscular con-

ig. 3. Global view of the model with muscles, mandibular fossae andoodstuff.

eering & Physics 30 (2008) 218–225 221

valslw

3

3

mawgtmc

sfIdbcit

3

ritta

fmtotwawonstbmap

M. Daas et al. / Medical Engin

alues were chosen according to data from literature [28,29]nd were assimilated to the mastication force. These non-inear analyses were accomplished on the complete modelhown in Fig. 3, composed of 108781 elements. All calcu-ations were accomplished with the MSC/MARC 2005 soft-are package (MSC Software, Santa Ana, California, USA).

. Results

.1. Determination of a probable muscular action

The study relative to muscular action revealed that theasseters and temporalis produced minimal work when

pproximately two thirds of the force was applied on theorking side (see Fig. 4). The minimal work for medial ptery-oids was reached when about half of the force came fromhe working side. It can be noted that a 100 N contact nor-

al force could not be reached when lateral pterygoids wereontracting.

A comparison between the different muscular groupshowed that the mastication force of 100 N was generatedrom the smallest amount of work by masseters (Fig. 5).ndeed, the work done by masseters was half of that pro-uced by medial pterygoids. The one-fifth ratio was obtainedetween the work done by masseters and temporalis. As aonsequence, masseters were chosen to achieve the follow-ng analyses, with two thirds of the 100 N load resulting fromhe working side.

.2. Overdenture global behaviour

Comparisons between “rigid” and “resilient” configu-ations for various positions of the foodstuff showed an

mportant influence of the retention mechanism used. Firstly,he global behavior of the overdenture strongly depended onhe attachment configuration. The deformed shape of mucosand overdenture is illustrated in Figs. 6 and 7, respectively,

Fig. 4. Work needed by masseters versus force repartition.

rbfsl

ai“wotsre

3i

i(

Fig. 5. Work needed to obtain a 100 N force on the first molar.

or “rigid” and “resilient” configurations with a displace-ent magnification factor of 10. In all cases, a tilting of

he overdenture was observed. For position (M), a swingf the overdenture around an axis formed by the intersec-ion of the sagittal and horizontal planes could be noted. Theorking side of the overdenture was shifted down under the

ction of the foodstuff. This motion involved a lift of the non-orking side of the overdenture. For position (C), both sidesf the overdenture were rising with greater amplitude of theon-working side. Concerning position (I), an approximatelyymmetrical rise of both sides could be seen. The overdentureurned around an axis going through the centers of the twoall attachments. In the “resilient” configuration, the rockingotion was less important for each foodstuff position andvertical translation was superimposed leading to a com-

ression of the mucosa of the denture bearing area. This lastemark is confirmed by Figs. 8 and 9 where the contact areaetween overdenture and mucosa is illustrated, respectively,or “rigid” and “resilient” configurations. Dark areas corre-pond to contact zones between overdenture and mucosa. Aarger contact area was noted in the “resilient” configuration.

Whatever the foodstuff position, the vertical gap consider-bly changed the way the vertical load was transiting throughmplants and mucosa of the denture bearing area (Fig. 10). Inrigid” (M) and (I) configurations, most of the vertical loadas supported by the mucosa. For position (C), the main partf the load transited through the working-side implant. Inhe “resilient” configuration, the proportion of vertical loadupported by the mucosa increased for all foodstuff configu-ations and most of the vertical load was borne by the mucosa,ven in position (C).

.3. Stresses in peri-implant bone, ball abutments andmplants

The distribution of the stress tensor vertical componentn the peri-implant bone, reported in Fig. 11, for positionM), indicated that a bending of the implants occurred, since

222 M. Daas et al. / Medical Engineering & Physics 30 (2008) 218–225

Fig. 6. View of overdenture motion in “rigid” configuration with foodstuff on: (a) molar, (b) canine, (c) incisor.

guratio

nosiosmtihi

inttc

Fig. 7. View of overdenture motion in “resilient” confi

egative and positive values were reported on diametricallypposite zones. This phenomenon was visible for each food-tuff position (cases (C) and (I) not reported in Fig. 11) andn both rigid and resilient configurations. The bending planerientation varied as a function of foodstuff position, but noignificant change was observed comparing both retentionechanisms. In case (C), an important bending moment of

he right implant occurred whereas the bone surrounding leftmplant was almost unloaded. It can also be observed that theighest stress values were found in the cortical bone near themplant necks.

ainf

Fig. 8. Contact area of overdenture on the mucosa in “rigid” config

Fig. 9. Contact area of overdenture on the mucosa in “resilient” confi

n with foodstuff on: (a) molar, (b) canine, (c) incisor.

Maximal values of the stress tensor vertical componentn the peri-implant bone are reported in Table 2. It can beoted that when foodstuff was situated above the first molar,he corresponding maximal value was about 30% lower inhe “resilient” configuration than in the “rigid” one. In otherases, differences were less significant.

Maximal values of von Mises equivalent stress on implants

nd ball abutments are reported in Table 3. For the rightmplant and ball abutment, von Mises stresses were sig-ificantly reduced in “resilient” configuration, whatever theoodstuff position. Conversely, for the left side components,

uration with foodstuff on: (a) molar, (b) canine, (c) incisor.

guration with foodstuff on: (a) molar, (b) canine, (c) incisor.

M. Daas et al. / Medical Engineering & Physics 30 (2008) 218–225 223

Fig. 10. Distribution of the mastication force between implants and mucosa according to foodstuff position. Positive values correspond to forces directedupwards.

Fig. 11. Vertical component of stress in the interforaminal region when foodstuff is above the first molar (MPa): (a) “rigid” configuration, (b) “resilient”configuration.

Table 2Vertical component of stress in the peri-implant bone (MPa)

Stress valuesin MPa

First molar Canine Incisors

“Rigid” “Resilient” “Rigid” “Resilient” “Rigid” “Resilient”

Right sideMinimal value −29.82 −18.37 −33.73 −32.00 −12.83 −11.08Maximal value 32.05 23.14 29.12 34.87 23.73 26.45

Left sideMinimal value −26.61 −16.28 −10.49 −4.59 −26.45 −28.93Maximal value 31.57 20.02 11.24 5.59 33.15 32.86

Table 3Maximal equivalent von Mises stress on implants and ball abutments (MPa)

Stress values in MPa First molar Canine Incisors

“Rigid” “Resilient” “Rigid” “Resilient” “Rigid” “Resilient”

RL

scp

4

wp

ats[

t

ight implant and ball abutment 98.3 65.6eft implant and ball abutment 45.7 120.1

tresses were increased in the “resilient” configuration. Itan be noted that the highest stress values were obtained forosition (C).

. Discussion

Results obtained in muscular action simulations complyith existing literature. The fact that lateral pterygoids do notrovide a considerable part of the food-crushing force is in

tats

120.0 104.6 50.4 45.7103.7 147.6 52.5 62.1

ccordance with the findings of Van Eidjen et al. [30]. Thewo third/one third ratio between working and non-workingides for the masseter contraction is also reported in literature30,31].

Concerning the overall behavior, our analyses confirmedhe global motion of the mandibular IRO during mastica-

ion reported in literature [17]. The “resilient” configurationllows for some movements of the denture. To a certain extenthis was due to the depressibility of the denture bearing tis-ues and ball-and-socket joints constituting each attachment.

2 eering

Htbo

mltitiptw

ioimtt“

ve“aicfftcIaudSibqtmbd

nboemfitaes

shtta

att“lltqmci[octoohCaf

tintatcw[

tf[ie

5

bhmt

24 M. Daas et al. / Medical Engin

owever, this remark must be moderated because overden-ure motions are actually limited by occlusion contacts andy peripheral tissues which were not taken into account inur simulations.

The denture bearing load on the alveolar ridge cannot beeasured in vivo [3]. As reported by other authors [18], the

oad on the denture bearing area in the “resilient” configura-ion was higher than in the “rigid” one. This can be explainedn our calculations by the fact that the vertical gap delayedhe axial contact between female and male parts. Therefore,n the “resilient” case, the implants supported only a weakart of the contact force. The higher involvement of the den-ure bearing area mucosa due to the axial gap is stated by theider contact area observed in that case.All the results concerning the load repartition between

mplants and mucosa of the denture bearing surface showedne of the main interest in using the mandibular IRO. Indeed,n a classical mastication configuration, i.e. foodstuff on first

olar, implants only supported a small part of the mastica-ion force, reducing the risk of overload. Besides, the loadransiting through the implants was reduced by the use ofresilient” attachments.

In opposition, the overdenture motion induced by theertical gap engendered a bending moment. Consequently,ven if the load transiting through implants decreased in theresilient” case, high stresses appeared in implants and ballbutments. These stresses were spread in bone, explainingn part the high values observed in the peri-implant boneompared to other studies [10,14,20]. In our study, they wereound mainly in the cortical bone and were caused by the per-ect contact considered between bone and implants. In fact,his assumption induces exaggerated tensile stresses in theortical bone and very low stresses in the cancellous bone.n spite of the bending moment, the interest of “resilient”ttachments was confirmed by the 30% lower stress val-es obtained in the peri-implant bone for position (M) (i.e.uring mastication). This observation was also described byadowsky and Caputo [8], who reported higher stresses on

mplants in the case of ball anchors without spacer. It shoulde highlighted that values of stresses were not compareduantitatively between the different foodstuff locations dueo the important differences of loading conditions. Maxi-

al values of von Mises equivalent stress in implants andall abutments testified that these parts were not endangereduring the mastication process.

The analysis of the influence of the retention mecha-ism was made possible thanks to the precise modeling ofoundary conditions and contact management. Indeed, it isf great importance to accurately model all the interactionsxisting in the mandibular environment. Most finite elementandibular models encountered in the literature use simpli-ed boundary conditions [20,32]. For example, modeling the

emporo-mandibular joint as a pivot connexion [20] or fixingn overdenture node in vertical direction to simulate the pres-nce of a foodstuff [20] is too restraining compared to trueituation. Some degrees of freedom are then wrongly con-

stsc

& Physics 30 (2008) 218–225

trained, altering global and local overdenture behaviors. Theigh stresses obtained in the peri-implant bone compared tohose found by Menicucci et al. [20] can also be explained byhe fact that they drastically simplified the mandible geometrynd the interactions with its environment.

It is generally assumed that loading the alveolar ridgebove a certain level may lead to its atrophy. In our study,he load of denture-bearing area is mainly determined byhe number of degrees of freedom of the attachment. In therigid” configuration, the vertical translation is suppressedeading to reduced forces acting on posterior ridge which mayimit eventual bone loss in the posterior ridge. Nevertheless,o avoid bone ridge resorption, it is essential to get an ade-uate denture bearing surface, a proper fit and occlusion of theandibular IRO. On the other hand, we have seen that rigid

onfiguration can generate higher stress values on the peri-mplant bone during mastication as confirmed in literature14,18–20]. It is accepted that peri-implant bone resorptionccurs if implant is overloaded. However, the level of stressorrelated to bone resorption is not clearly defined in litera-ure. Therefore, the question remains open to define the levelf load that can be exerted before bone loss occurs [3]. Inur case, stresses in the peri-implant bone were considerablyigher than those reported on the denture bearing surface.onsequently, resorption might rather occur in cortical boneround the implant necks where stress concentrations wereound.

Another important point concerns peri-implant bone reac-ion. The cortical bone layer sometimes observed aroundmplants and resulting from osseointegration [20,24,28] wasot considered here and would probably have influencedhe results. At this stage of the study no criterion waspplied to predict how bone is affected by implant transmit-ed loads and by damage caused by surgery [12]. Indeed, itan be assumed that bone remodeling could happen and thenould change the properties of the bone-implant interface

28,33,34]. Future study will relate to this topic.Clinically, it could be suggested that attachment sys-

ems that provide the most equitable transfer of occlusalorces among abutments are preferred for bone preservation8,18]. However, it has to be mentioned that numerical stud-es present clear limitations and any conclusions should beventually verified in vivo.

. Conclusion

Within the limits of this study, various conclusions cane drawn. The boundary conditions retained in our modelave revealed that masseters are the muscles that provideost of the food-crushing force during mastication. About

wo thirds of the force came from the working side mas-

eter. “Resilient” attachments allowed the reduction of loadransmission throughout the implants. They also reduced thetresses in the bone surrounding the implant during the masti-ation process. Besides, attachment configuration influenced

eering

tmantrwimalb

A

Looi

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

M. Daas et al. / Medical Engin

he global behavior of the mandibular IRO. Important rockingotions occurred in the “rigid” cases whereas more moder-

te tilting motions accompanied by a vertical translation wereoted in the “resilient” cases. This could be achieved thanks tohe quality of boundary conditions and contact managementetained. During the mastication process, load transferredas mainly supported by the mucosa of the denture bear-

ng surface in both configurations. This phenomenon wasore pronounced in the “resilient” configuration. Finally, it

ppeared that the “resilient” attachment allowed for a betteroad distribution between the dental implants and the dentureearing surface.

cknowledgments

The authors wish to thank Orthopedie Biomecaniqueocomotion (OBL, Chatillon, France) for the achievementf the three-dimensional reconstruction of the mandible andverdenture geometry from the computed tomography exam-nations.

eferences

[1] Mericske-Stern R, Steinlin ST, Marti P, Geering AH. Peri-implantmucosal aspects of ITI implants supporting overdentures. A five-yearlongitudinal study. Clin Oral Implants Res 1994;5:9–18.

[2] Jemt T, Chai J, Harnett J, Health MR, Hutton JE, Johns RN. A5-year prospective multicenter follow-up report on overdentures sup-ported by osseointegrated implants. Int J Oral Maxillofac Implants1996;11:291–8.

[3] Heckmann SM, Winter W, Meyer M, Weber HP, Wichman MG.Overdenture attachment selection and the loading of implant anddenture-bearing area. Part 1. In vivo verification of stereolithographicmodel. Clin Oral Impl Res 2001;12:617–23.

[4] Oetterli M, Kiener P, Mericske-Stern R. A longitudinal study onmandibular implants supporting an overdenture: the influence of reten-tion mechanism and anatomic-prosthetic variables on peri-implantparameters. Int J Prosthodont 2001;14:536–42.

[5] Tokuhisa M, Matsushita Y, Koyano K. ln vitro study of a mandibularimplant overdenture retained with ball, magnet, or bar attachments:comparison of load transfer and denture stability. Int J Prosthodont2003;16(2):128–34.

[6] Gotfredsen K, Holm B. Implant-supported mandibular overdenturesretained with ball or bar attachments: a randomized prospective 5-yearstudy. Int J Prosthodont 2000;13(2):125–30.

[7] Cune M, Van Kampen F, Van der Bilt A, Bosman F. Patient satisfactionand preference with magnet, bar-clip, and ball-socket retained mandibu-lar implant overdentures: a cross-over clinical trial. Int J Prosthodont2005;18(2):99–105.

[8] Sadowsky SJ, Caputo AA. Effect of anchorage systems and exten-sion base contact on load transfer with mandibular implant-retainedoverdentures. J Prosthet Dent 2000;84:327–34.

[9] Meijer HJ, Kuiper JH, Starmans FJM, Bosman F. Stress distributionaround dental implants: influence of superstructure, length of implants,and height of mandible. J Prosthet Dent 1992;68:96–101.

10] Meijer HJ, Starmans FJ, Steen WH, Bosman F. Location of implants

in the interforaminal region of the mandible and the consequences forthe design of the superstructure. J Oral Rehabil 1994;21(1):47–56.

11] Menicucci G, Lorenzetti M, Pera P, Preti G. Mandibular implant-retained overdenture: a clinical trial of two anchorage systems. Int JOral Maxillofac Implants 1998;13(6):851–6.

[

& Physics 30 (2008) 218–225 225

12] Brunski JB. Biomaterials and biomechanics in dental implant design.Int J Oral Maxillofac Implants 1988;3:85–97.

13] Kenny R, Richards MW. Photoelastic stress patterns produced byimplant-retained overdentures. J Prosthet Dent 1998;80:559–64.

14] Chun HJ, Park DN, Han CH, Heo SJ, Heo MS, Koak JY. Stress distribu-tions in maxillary bone surrounding overdenture implants with differentoverdenture attachments. J Oral Rehabil 2005;32(3):193–205.

15] Walton JN, MacEntee ML. Problems with prostheses on implants: aretrospective study. J Prosthet Dent 1994;71:283–8.

16] Dudic A, Mericske-Stern R. Retention mechanisms and prosthetic com-plications of implant-supported mandibular overdentures: long-termresults. Clin Oral Implants Res 2002;4(4):212–9.

17] Mericske-Stern R. Three-dimensional force measurements withmandibular overdentures connected to implants by ball-shaped reten-tive anchors. A clinical study. Int J Oral Maxillofac Implants 1998;13:36–43.

18] Heckmann SM, Winter W, Meyer M, Weber HP, Wichman MG.Overdenture attachment selection and the loading of implant anddenture-bearing area. Part 2. A methodical study using five types ofattachments. Clin Oral Impl Res 2001;12:640–7.

19] Porter JA, Petropoulos VC, Brunski JB. Comparaison of load distribu-tion for implant overdenture attachments. Int J Oral Maxillofac Implants2002;17:651–62.

20] Menicucci G, Lorenzetti M, Pera P, Preti G. Mandibular implant-retained overdenture: finite element analysis of two anchorage systems.Int J Oral Maxillofac Implants 1998;13:369–76.

21] Zarb GA, Bolender CL, Carlsson GE. In: Boucher’s prothodontic treat-ment for edentulous patient. 11th ed. St Louis: CV Mosby; 1997. p.8–29.

22] Voigt LJ. Uber die Bezechnung zwischen den beidem Elas-tizitatskonstanten isotroper Korper. Wied Ann 1889;33:573–87.

23] Reuss A. Berechnung der Fliessgrenze von Mischkristalen auf grundder Plastizitatsbedigung fur Einkristalle. Z Angew Math Mech1929;9:49–58.

24] Brunski JB, Puleo DA, Nanci A. Biomaterials and biomechanics of oraland maxillofacial implants: current status and future developments. IntJ Oral Maxillofac Implants 2000;15:15–46.

25] Takayama Y, Yamada T, Araki O, Seki T, Kawasaki T. The dynamicbehaviour of a lower complete denture during unilateral loads: analysisusing the finite element method. J Oral Rehabil 2001;28:1064–74.

26] Throckmorton GS, Throckmorton L. Quantitative calculations oftemporomandibular joint reaction forces-I. The importance of the mag-nitude of the jaw muscle forces. J Biomechanics 1985;18(6):445–52.

27] Koolstra JH. Number crunching with the human masticatory system. JDent Res 2003;82(9):672–6.

28] Ogata K, Satoh M. Centre and magnitude of the vertical forces incomplete denture weares. J Oral Rehabil 1995;22:113–9.

29] Cruz M, Wassall T, Toledo EM, Da Silva Barra LP, De Castro LemongeAC. Three-dimensional finite element stress analysis of a cuneiform-geometry implant. Int J Oral Maxillofac Implants 2003;18(5):675–84.

30] Van Eidjen TMGJ, Brugman P, Weijs WA, Oosting J. Coactivation ofjaw muscles: recruitment order and level as a function of bite forcedirection and magnitude. J Biomech 1990;23:475–85.

31] Tortopidid D, Lyons MF, Baxendale RH. Bite force, endurance andmasseter muscle fatigue in healthy edentulous subjects and those withTDM. J Oral Rehabil 1999;26:321–8.

32] Vanzyl PP, Grundling NL, Jooste CH, Terblanche E. Three-dimensionalfinite element model of a human mandible incorporating six osseointe-grated implants for stress analysis of mandibular cantilever prostheses.Int J of Oral Maxillofac Implants 1995;10:51–7.

33] Carter DR, Fyhrie DP, Whalen RT. Trabecular bone density and loading

history: regulation of connective tissue biology by mechanical energy.J Biomech 1987;20(8):785–94.

34] Lacroix D, Prendergast PJ, Li G, Marsh D. Biomechanical model tosimulate tissue differentiation and bone regeneration: application tofracture healing. Med Bio Eng Computing 2002;40:14–21.

本文献由“学霸图书馆-文献云下载”收集自网络，仅供学习交流使用。

学霸图书馆（www.xuebalib.com）是一个“整合众多图书馆数据库资源，

提供一站式文献检索和下载服务”的24 小时在线不限IP

图书馆。

图书馆致力于便利、促进学习与科研，提供最强文献下载服务。

图书馆导航：

图书馆首页 文献云下载 图书馆入口 外文数据库大全 疑难文献辅助工具