Please cite this article in press as: Ramlee MH, et al.

l. Finite element analysis of three commonly used external xation devices for treating
Type III pilon fractures. Med Eng Phys (2014), http://dx.doi.org/10.1016/j.medengphy.2014.05.015
ARTICLE IN PRESS
G Model
JJBE-2506; No. of Pages 9
Medical Engineering & Physics xxx (2014) xxxxxx
Contents lists available at ScienceDirect
Medical Engineering & Physics
j o ur nal homepage: www. el sevi er . com/ l ocat e/ medengphy
Technical note
Finite element analysis of three commonly used external xation
devices for treating Type III pilon fractures
Muhammad Hanif Ramlee
a
, Mohammed Raq Abdul Kadir
a,
,
Malliga Raman Murali
b
, Tunku Kamarul
b
a
Medical Devices and Technology Group (MEDITEG), Faculty of Biosciences and Medical Engineering, Universiti Teknologi Malaysia, 81310 Johor Bahru,
Johor, Malaysia
b
Tissue Engineering Group (TEG), National Orthopaedic Centre of Excellence in Research and Learning (NOCERAL), Department of Orthopaedic Surgery,
Faculty ofMedicine, University of Malaya, 50603 Lembah Pantai, Kuala Lumpur, Malaysia
a r t i c l e i n f o
Article history:
Received 28 October 2013
Received in revised form19 May 2014
Accepted 24 May 2014
Keywords:
Finite element
External xator
Pilon fractures
Stability
Biomechanics
Micromovement
a b s t r a c t
Pilon fractures are commonly caused by high energy trauma and can result in long-term immobilization
of patients. The use of an external xator i.e. the (1) Delta, (2) Mitkovic or (3) Unilateral frame for treating
type III pilon fractures is generally recommended by many experts owing to the stability provided by
these constructs. This allows this type of fracture to heal quickly whilst permitting early mobilization.
However, the stability of one xator over the other has not been previously demonstrated. This study was
conducted to determine the biomechanical stability of these external xators in type III pilon fractures
using nite element modelling. Three-dimensional models of the tibia, bula, talus, calcaneus, navicu-
lar, cuboid, three cuneiforms and ve metatarsal bones were reconstructed from previously obtained
CT datasets. Bones were assigned with isotropic material properties, while the cartilage was assigned as
hyperelastic springs with MooneyRivlin properties. Axial loads of 350 N and 70 N were applied at the
tibia to simulate the stance and the swing phase of a gait cycle. To prevent rigid body motion, the calca-
neus and metatarsals were xed distally in all degrees of freedom. The results indicate that the model
with the Delta frame produced the lowest relative micromovement (0.03 mm) compared to the Mitkovic
(0.05 mm) and Unilateral (0.42mm) xators during the stance phase. The highest stress concentrations
were found at the pin of the Unilateral external xator (509.2MPa) compared to the Mitkovic (286.0 MPa)
and the Delta (266.7MPa) frames. In conclusion, the Delta external xator was foundto be the most stable
external xator for treating type III pilon fractures.
2014 IPEM. Published by Elsevier Ltd. All rights reserved.
1. Introduction
A pilon fracture is a general description of a comminuted frac-
ture at the distal tibia involving the ankle joint that occurs as the
result of high-energy vertical axial loading. This can occur as the
result of a fall from a substantial height, road trafc accidents,
industrial mishaps or sporting injuries, especially those involving
contact sports [16]. These fractures are uncommon and represent
up to 710% of tibia fractures and less than 1% of all lower extrem-
ity fractures [4,7]. The mechanism of injury varies from simple
rotational fractures to high energy axial compression injuries com-
plicated by shearing, rotation and bending forces [4,8,9]. In 1969,

Corresponding author. Tel.: +6 07 5535961; fax: +6 07 5536222.

E-mail addresses: mhanif008@gmail.com(M.H. Ramlee),
raq@biomedical.utm.my (M.R. Abdul Kadir), mrmurali08@gmail.com
(M.R. Murali), tkzrea@um.edu.my (T. Kamarul).
Ruedi and Allgower [1016] classied the pilon fractures into three
types: type I is an intra-articular fracture of the distal tibia with or
without minimal displacement; type II is a displaced intra-articular
fracturewithor without minimal comminution; andtypeIII has sig-
nicant comminution and impaction of the intra-articular surface
with displacement.
Treatment of pilon fractures is targeted to reduce the frac-
ture, align the ankle position, provide fast soft tissue healing,
be minimally invasive, allow the recreation of the joint surfaces
and provide early ankle function [1719]. Type I and type II frac-
tures can be almost effortlessly restored using internal xation,
and the results are promising without any major complications
[4,17,20,21]. However, the treatment of a type III fracture is still
controversial since it involves an intra-articular fracture with dis-
placement, signicant comminution and is associated with high
rate of complications [1,22]. Immediate treatment is reported to
produce complications such as infection, loss of reduction, non-
union, malunion and deformity [2326]. A systemic step-wise
http://dx.doi.org/10.1016/j.medengphy.2014.05.015
1350-4533/ 2014 IPEM. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Ramlee MH, et al. Finite element analysis of three commonly used external xation devices for treating
Type III pilon fractures. Med Eng Phys (2014), http://dx.doi.org/10.1016/j.medengphy.2014.05.015
ARTICLE IN PRESS
G Model
JJBE-2506; No. of Pages 9
2 M.H. Ramlee et al. / Medical Engineering & Physics xxx (2014) xxxxxx
approach, consisting of bular plating and temporary bridging
external xation, later substituted by a denitive external xation,
was reported to be favourable for type III fracture treatment [2,3,5].
The range of external xators used for the treatment of type
III pilon fractures includes spanning xator with rods and clamps,
dynamic or articulated devices, and ring frames [13,5,2730]. The
clinical outcome of using external xators has been reported to be
superior over those treated with internal plates and screws [28,29].
The three most popular spanning external xators in the litera-
ture include the Delta, Mitkovic and Unilateral devices [7,27,30].
Although the mid-term clinical outcomes have been shown to be
good in terms of successful healing process, the biomechanical sta-
bility of these devices has not been well investigated. Furthermore,
there is no clear evidence in the present literature with regard to
these types of implants and whether one of these will produce
the best stability in treating type III pilon fractures. Therefore, the
overall aimof this study was to understand the underlying biome-
chanics of the three most commonly used external xators for type
III pilon fractures. Finite element method was used to (1) assess the
stability of the aforementioned three external xators (2) investi-
gate the stress distribution in the xator and bone to highlight the
likelihood of the particular areas of the implant and bone that may
be subjected to excessive mechanical stress.
2. Materials and methods
2.1. Three-dimensional modelling
CT images of the right lower limb used in this study were
acquired with the approval of the medical ethical committee of
the Hospital Tengku Ampuan Afzan, Kuantan, Malaysia [65]. The
slice thickness of the CT images was 1.5mmin a 512512 matrix.
The DICOM data sets, which consist of 225 CT images, were then
imported into Mimics 15.1 software (Materialise, Leuven, Belgium)
to reconstruct the surface geometry of the tibia, bula, talus, calca-
neus, cuboid, navicular, cuneiforms andmetatarsals. Thetibial bone
was cut approximately 20cm above the medial tibial malleolus
[31,32]. A threshold of 700 Hounseld units was used to differen-
tiate between cortical and cancellous bone [33]. To simulate type
III pilon fractures, a total of eight fragments (Fig. 1) were modelled
basedontheRuedi andAllgower classication[1016,30]. Aperfect
t of the interfragments was simulated, i.e. there were no fracture
gaps betweenthefragments. However, thefragments wereallowed
to move relative to each other with an assigned friction coefcient
of 0.3 [34]. All contact between articulating surfaces was assigned a
friction coefcient of 0.3 [34]. All three-dimensional models of the
bones were converted to a surface triangular mesh and saved in a
stereolithographic le format.
2.2. Cartilage and ligaments
Thecartilagewas modelledmanuallywithanestimateduniform
thickness of 1mm for the tibia, bula, talus and calcaneus (Fig. 1)
[35,36]. The behaviour of the cartilage was simulated using the
MooneyRivlin hyper-elastic material properties with coefcients
of C
01
=0.41MPa and C
10
=4.1 MPa [3740]. A total of 34 ligaments
and three plantar fascias were modelled using linear spring ele-
ments with an assigned specic stiffness (Table 1). The use of linear
links to simulate the ligaments was found to be adequate and has
been reported in previous studies [4043]. Multiple parallel linear
springs were used to better mimic the distribution of the origin and
insertionof the ligaments [39,40]. The positionandinsertionpoints
of all the ligaments were estimated based on Netter (2003) [44].
2.3. Finite element modelling
The bones and cartilage in the STL les were imported into
Marc.Mentat (MSC.Software, Santa Ana, CA). The software was
used to convert the completed three-dimensional model to linear
rst order tetrahedral elements. Bones were assigned using lin-
ear isotropic material properties with elastic modulus of 7300MPa
for cortical bone [50,51] and 1100MPa for cancellous bone [52].
Fig. 1. Finite element model; (a) Delta frame, (b) Mitkovic xation, (c) Unilateral external xator. Fragment 8 is located at the posterior of distal tibia.
Please cite this article in press as: Ramlee MH, et al. Finite element analysis of three commonly used external xation devices for treating
Type III pilon fractures. Med Eng Phys (2014), http://dx.doi.org/10.1016/j.medengphy.2014.05.015
ARTICLE IN PRESS
G Model
JJBE-2506; No. of Pages 9
M.H. Ramlee et al. / Medical Engineering & Physics xxx (2014) xxxxxx 3
Table 1
Stiffness of ligaments.
Ligaments represented
in the models
Stiffness (N/mm) References
Interosseous membrane (4 ligaments) 400 [45]
Anterior tibiobular (distal) 78 [46]
Posterior tibiobular (distal) 101 [47]
Anterior talobular 90 [46]
Posterior talobular 70 [48]
Calcaneobular 70 [48]
Anterior tibiotalar 70 [46]
Posterior tibiotalar 80 [46]
Tibiocalcaneal 122 [46]
Tibionavicular 40 [46]
Interosseous talocalcaneal 70 [45,48]
Lateral talocalcaneal 70 [48]
Medial talocalcaneal 70 [48]
Posterior talocalcaneal 70 [48]
Dorsal talonavicular (2 ligaments) 70 [46]
Calcaneonavicular (dorsal & plantar) 70 [46,48]
Calcaeocuboid (dorsal & short plantar) 70 [46,48]
Cuboideonavicular (dorsal & plantar) 70 [45,46]
Cuneonavicular (dorsal & plantar) 70 a
Intercuneiform(dorsal & plantar) 70 a
Tarsometatarsal (dorsal & plantar) 70 a
Metatarsal (dorsal & plantar) 70 a
Medial plantar fascia 200 [49]
Central plantar fascia 230 [49]
Lateral plantar fascia 180 [49]
Long plantar 70 [45]
a: assumed fromneighboring ligaments.
Poissons ratio for both cortical and cancellous bones was simu-
lated with value of 0.3 [50,51] and 0.26 [52], respectively. Three
external xator frames, i.e. the Delta, Mitkovic and Unilateral sys-
tems, weredesignedusingthree-dimensional (3D) Computer Aided
Design (CAD) software (Solidworks 2012, Dassault Systemes Solid-
works Corp., USA) with 5mm pins and 11mmrods. To simulate the
Delta and Mitkovic xators, two pins were positioned at the tibial
diaphysis, one pin at the body of the calcaneus and another pin at
the rst metatarsal (Fig. 1) [27,30]. For the Unilateral frame, only
one pinwas positionedat the tibial diaphysis andanother pinat the
body of calcaneus (Fig. 1) [7]. All the xators were meshed using
3-Matic 7.1 (Materialise, Belgium) and were assigned with tita-
niummaterial properties with a Youngs modulus of 110,000MPa
and Poissons ratio of 0.3 [53,54]. Mesh convergence analyses were
performed and resulted in a variation in mesh size throughout the
model. The smallest mesh of size 1 was used for the pin-bone inter-
face, whereas a larger meshsize of 3was usedfor the bone. The total
number of elements and nodes for the Delta xator was 675,000
and157,000, respectively; for theMitkovic xator, this was 588,000
and 140,000, respectively, and for the Unilateral frame, this was
510,000 and 112,000, respectively. Contact condition between the
external xators and bone was set as an explicit contact with a tan-
gential friction coefcient of 0.4 [52,55,56]. Radial pre-stress was
not modelled at the interface between the bone and the xators.
2.4. Boundary conditions
In order to simulate human walking conditions, two physiolog-
ical loads were applied in this study: (1) the swing phase [5759]
and (2) the stance phase [60,61] of a gait cycle, where the force
value was determined fromthe adjacent muscles such as the gas-
trocnemius andsoleus. For the swing phase, 10%of the body weight
was recorded on these particular muscles [5759]. We assumed a
body weight of 70kg in our study; therefore, 70N was applied to
the tibia in the axial direction to simulate the swing phase. For the
stance phase, 50% (350N) of the body weight was applied onto the
foot, as has been reported by Cheung et al. and Simkin [60,61]. The
use of axial weight loadinghas become popular since this technique
is a way of testing bone quality and bone healing process [62]. In
order to prevent rigidbody movements during the analysis, the dis-
tal surfaces of the calcaneus and all metatarsal bones were xed in
all directions (Fig. 1). The relative micromovement of all simulated
models were measured between the proximal and distal fragments
at the lateral side.
3. Results
3.1. Stress distribution
The von Mises stress at the pin-bone interface at the tibia
and calcaneus is shown in Fig. 2. During the swing phase, the
observed peak values for the pin-bone interface at the tibia were
24.9MPa, 35.6MPa and 84.9MPa for the Delta, Mitkovic and Uni-
lateral xation devices, respectively. The difference in magnitude
was even higher during the stance phase for the pin-bone inter-
face at the tibia, where the Unilateral showed two times greater
stress (399.0MPa) than the Mitkovic (206.6MPa) and three times
greater stress than the Delta xator (130.3MPa). Generally, at the
tibia, the peak von Mises stress was found at the entrance cortex
of the pin-bone interface. During the swing phase, the magnitude
of the maximum stress was 0.8MPa, 3.6MPa and 4.5 MPa for the
Delta, Mitkovic and Unilateral xators, respectively. On the other
hand, the FE results in terms of von Mises stress were greater for
the simulated stance phase where the Unilateral generated at least
1.4 times greater (121.4MPa) than the Mitkovic (87.2MPa) and 27
times greater than the Delta (4.5MPa) frames. At the calcaneus,
high von Mises stress was observed at the pin-bone interface for
both the swing and the stance phase. During the swing phase,
the magnitude of stress for the Unilateral xator was at least four
times greater (98.5MPa) as compared to the Mitkovic (20.7MPa)
and Delta (7.9MPa) frames. Additionally, greater stresses were
observed during the stance phase, where the Unilateral xator pro-
duced 382.5MPa, 3.9 times larger than the Mitkovic (98.0MPa) and
9.2 times larger than Delta external xator (41.5MPa)
The stress distribution amongst the three external xators is
illustrated in Fig. 3. During the swing phase, higher von Mises
stresses werepredictedat thecalcaneus pinfor theUnilateral exter-
nal xator (113.6MPa) followed by the Mitkovic (80.0MPa) and
the Delta (45.6MPa) frame. For the Delta and Mitkovic systems,
the proximal pin-bone interface of the tibia bone produced a small
stress of 25.0MPa and 44.5MPa, respectively, as compared to the
distal pin-bone interface with a value of 27.0MPa and 54.5MPa,
respectively. At the rst metatarsal bone, high stress of 43.5MPa
was found for the Mitkovic xator, whilst the Delta frame only
showed 7.5MPa of stress at that particular bone. During the stance
phase, thestress at thecalcaneus pinfor theUnilateral external xa-
tor (509.2MPa) was at least 1.8times greater thanwiththeMitkovic
(286.0MPa) and Delta (266.7MPa) xators. At the proximal pin of
theMitkovic frame, thevonMises magnitude(125.9MPa) was close
to magnitude of the Delta frame (120.2MPa). On the other hand, a
small stress of 49.5MPa was generated at the rst metatarsal pin
of the Delta external xator, whilst the Mitkovic frame (286.0MPa)
showed 4.5 times greater stress.
3.2. Displacement and micromovement
Fig. 4 shows the displacement plot for the tibia bone. For the
swing phase, the greatest relative micromovement was observed
for the Unilateral xator (0.3mm) as compared to the Mitkovic
and Delta (0.02mm) frames. In contrast, the relative micromove-
ment was higher during the stance phase. The Unilateral xator
(0.42mm) generated 8.4 times greater micromovement than the
Please cite this article in press as: Ramlee MH, et al. Finite element analysis of three commonly used external xation devices for treating
Type III pilon fractures. Med Eng Phys (2014), http://dx.doi.org/10.1016/j.medengphy.2014.05.015
ARTICLE IN PRESS
G Model
JJBE-2506; No. of Pages 9
4 M.H. Ramlee et al. / Medical Engineering & Physics xxx (2014) xxxxxx
Fig. 2. The von Mises stress distribution for the tibia and calcaneus bones for (a) Delta frame, (b) Mitkovic xation and (c) Unilateral external xation.
Mitkovic (0.05mm) and 14 times greater micromovement than the
Delta (0.03mm) external xators.
The contour plot for the displacement of external xators is
shown in Fig. 5. When simulating the swing phase, it was demon-
stratedthat themaximumdisplacement producedbytheUnilateral
xator (8.7mm) was at least three times greater than the dis-
placement produced by the Mitkovic (3.0mm) and Delta (0.8mm)
systems. During the stance phase, the Unilateral xator generated
the highest magnitude of displacement (34.8mm). The Delta sys-
temwas themost stablexationsystemamongthethreeconstructs
examined with a maximumdisplacement of 3.8mm. The Mitkovic
systemshowed a maximumdisplacement value of 13.4mm.
4. Discussion
The use of an external xator for treating type III pilon frac-
tures is a well-acceptedsurgical option. This systemnot only allows
minimally invasive surgery of the soft tissue to be performed, but
also maintains the ankle alignment whilst allowing early ankle
mobilization [1719]. The results of using an external xator such
as the Delta, Mitkovic or Unilateral frame have been shown to
produce favourable clinical outcomes as compared to internal xa-
tors [1719,2326]. Stable xation and early anatomic reduction
of all fractures and dislocations can minimize long-term mor-
bidity and hasten soft tissue healing [63,64]. The biomechanical
Please cite this article in press as: Ramlee MH, et al. Finite element analysis of three commonly used external xation devices for treating
Type III pilon fractures. Med Eng Phys (2014), http://dx.doi.org/10.1016/j.medengphy.2014.05.015
ARTICLE IN PRESS
G Model
JJBE-2506; No. of Pages 9
M.H. Ramlee et al. / Medical Engineering & Physics xxx (2014) xxxxxx 5
Fig. 3. The von Mises stress plot for the external xator during swing and stance phase. (a) Delta, (b) Mitkovic and (c) Unilateral.
stability produced by these xators, provides improved fracture
healing to occur. A correlation between improved stability (albeit
still allowing micromotion to occur) with improved healing has
been demonstrated in a previous study [65]. However, it is clear
that fromour extensive literature reviewthat a comparison of the
most commonconstructs used for type III pilonfracture treatments
appears to be lacking. Hence, the results presented in the present
study are the rst that we are aware of. The results presented here
are of value not only for future research, but also serve as an objec-
tive measure for surgeons to justify the choice of one construct
over the other, although it may be necessary for further studies to
be conducted in order to support the choices made.
To build condence in our FE results, we corroborated our data
with experimental studies of Bergmann et al. [66] and Wang et al.
[32]. The former measured hip contact and ground reaction forces
for four patients during the most frequent activities of daily living.
Inorder tocompareour results withtheworkof Bergmannet al., we
simulated the ankle region of interest without a xator. Our results
showed minimum and maximum ground reaction forces of 0.1N
and 175.9N, respectively, compared to 0.1N and 108.7N obtained
Please cite this article in press as: Ramlee MH, et al. Finite element analysis of three commonly used external xation devices for treating
Type III pilon fractures. Med Eng Phys (2014), http://dx.doi.org/10.1016/j.medengphy.2014.05.015
ARTICLE IN PRESS
G Model
JJBE-2506; No. of Pages 9
6 M.H. Ramlee et al. / Medical Engineering & Physics xxx (2014) xxxxxx
Fig. 4. The displacement plot for the tibia bone; (a) Delta xation, (b) Mitkovic frame and (c) Unilateral external xator.
by Bergmann et al. [66]. Wang et al. [32], on the other hand, used
nine ankles from cadavers to measure the contact pressure of the
subtalar joint using pressure-sensitive lms. For the 600N load
applied at the tibia, they reported a maximum contact pressure
of 5.131.16MPa. This is close to the result that we obtained from
our FE simulation, which was 5.48MPa.
In thepresent study, thehighstress at thepin-boneinterfaceand
the surrounding tissues were in agreement with previous reports
[67,68]. The stress concentrated at this particular region is one of
the causes of an unstable external xator construct and can lead
to pain and implant loosening [12,6972]. However, this cannot be
avoidedwhenusinganexternal xator sinceanydesignedimplants
Please cite this article in press as: Ramlee MH, et al. Finite element analysis of three commonly used external xation devices for treating
Type III pilon fractures. Med Eng Phys (2014), http://dx.doi.org/10.1016/j.medengphy.2014.05.015
ARTICLE IN PRESS
G Model
JJBE-2506; No. of Pages 9
M.H. Ramlee et al. / Medical Engineering & Physics xxx (2014) xxxxxx 7
Fig. 5. The displacement plot for the external xator during the swing and stance phase. (a) Delta, (b) Mitkovic and (c) Unilateral.
using this concept will inherently have to face this issue. Never-
theless, our results showed a more favourable result for the Delta
external xator thanthe Mitkovic andUnilateral systems due tothe
lower stress magnitudes as well as better stability with lower rel-
ative micromovement values. Our ndings appear to be supported
by previously published clinical studies such as those by Cheema
et al. [27]. In their study, they observed that the design of Delta
frames provided the most stable construct, while preventing a high
incidence of getting deformities [27]. For long-termclinical results,
the use of the Unilateral external xator should be avoided since
the nite element predictions showed the highest relative micro-
movement as compared to the Delta and the Mitkovic constructs.
In addition, the Unilateral system, which does not use a pin at the
rst metatarsal bone, can cause forefoot equinus deformities [27].
Interfragment micromovement at the fracture site has been
reported as an important parameter that will assist in bone healing
process. Several studies using ovine has demonstrated that micro-
movement between 0.15mm and 0.4mm can assist in the healing
of a fracture gap not more than 3mm [73,74]. In our analysis, we
simulated a perfect t for the interfragments, so that there were no
fracture gaps present. The relative micromovement for the swing
phase for these bone fragments was therefore less than 0.02mm
(Mitkovic and Delta) and 0.3mm (Unilateral). During the the stance
phase, micromotions were observed to 0.03mm (Delta), 0.05mm
(Mitkovic) and 0.42mm (Unilateral). This suggests that in vivo,
bone regeneration can be expected.
Nevertheless, it is worth noting that unstable fracture xa-
tion leads to increasing stresses on implants [7577]. In several
studies it has been shown that stress onto implants xed at unsta-
ble sites have recoded levels in access maximum principal of
370600MPa [75] and von Mises stresses of 436750MPa [76,77]
for the plates. Similarly in our analysis, all three external xators
alsodemonstratedmaximumstress magnitudes at thexator pin. It
is fortunate however, that the magnitude of these stresses did not
exceed the ultimate strength of titanium alloys used in our sim-
ulation (800900MPa) thus suggesting that the construct for all
Please cite this article in press as: Ramlee MH, et al. Finite element analysis of three commonly used external xation devices for treating
Type III pilon fractures. Med Eng Phys (2014), http://dx.doi.org/10.1016/j.medengphy.2014.05.015
ARTICLE IN PRESS
G Model
JJBE-2506; No. of Pages 9
8 M.H. Ramlee et al. / Medical Engineering & Physics xxx (2014) xxxxxx
xations appears to provide adequate stability with minimal risk
of implant failure [78].
In treating type III pilon fractures, initial considerations such
as pin placement and the type of external xator must be made
properly before the surgery can be conducted. Previous clinical
reports have mentioned that misplacing the pins or improper use
of these devices can lead to a high incidence of complications, with
pin infection and loosening in up to 50% of cases and malunion
rates of up to 45%[24,26,79]. In the simulations conducted here, we
only attempted to show the comparison of ankle external xator
in terms of the biomechanical properties. Although the Unilateral
frame showed larger displacements and relative micromovements
compared to the other two constructs, stability could be better
achieved by placing the proximal pin closer to the fractured seg-
ments. The literature onpinplacement is fairly limited, thus further
study is necessary to assess the biomechanical effects of different
pin orientations.
As with any study, there are limitations that need to be consid-
ered so as to not to overstate the ndings. In the present study,
several assumptions and simplications were made which may
have resultedinalterations to the predicteddata. These limitations,
which are inherently particular to computer modelling involving
the reconstruction of complex joints such as the ankle, will be
present and are unfortunately unavoidable. First, it should be noted
that the model was simplied using isotropic material properties
in this simulation, thus excluding all other important factors that
may inuence the prediction of fracture stability. Although more
complex modelling would yield more realistic outcomes, omitting
such details appears to be an acceptable practice as demonstrated
by many previous studies [39,40,50,51]. Further simplications of
ligaments and plantar fascias modelling using linear links again
may have diverted the results from demonstrating true, real life
outcomes. The elements used to simulate the ligaments were not
modied, and therefore were allowed to resist both tension and
compression. Though this may not mimic the actual behaviour of
ligaments, the simplied properties of a spring have been used by
others with acceptable accuracy [4043]. The insertion points of
the ligaments were estimated based on a reference book by Netter
(2003) [44] and conrmed by an orthopaedic specialist. This was
done as there are no published reports regarding the geometrical
details of all 37 ligaments modelled in our study. Considering the
existence of subject-specic variation in ligament insertion points,
we believe that these estimations were valid. The determination
of the region of interest for the analysis is another limitation of
our study. Due to constraints in computing resources, we modelled
only the distal half of the tibia and bula. Nevertheless the FE anal-
ysis still valid as similar region of interest has been used by others
[31,32].
5. Conclusion
The results of nite element predictions suggest that the Delta
frame provides better stability and generates lower construct
stresses as compared to Mitkovic and Unilateral external xators.
Our data therefore suggest that Delta xators are superior for treat-
ing this type of fracture. However, further studies are required to
validate the outcome of these simulated studies, such as those of
cadaveric or clinical studies.
Ethical approval
Access to the CT images of the right lower limbusedinthis study
was grantedbyDr. ZainunBt. A. Rahman, Headof Department Diag-
nostic ImagingandDr. Ghazali Ismail, Chairmanof Clinical Research
Centre, Hospital Tengku Ampuan Afzan, 25100 Kuantan, Pahang
Darul Makmur, Malaysia.
Acknowledgement
Theworkhas beencarriedout usingtheresearchgrants received
fromeScienceFund, Ministryof Science, TechnologyandInnovation
Malaysia, FRGS Ministry of Education Malaysia, and UTMResearch
University Grants. More than one of the authors of this paper was
supported under University of Malaya HIR-MOHE research grant.
Conict of interest
None declared.
References
[1] Kapukaya A, Subasi M, Arslan H, Tuzuner T. Non-reducible, open tibial plafond
fractures treated with a circular external xator (is the current classication
sufcient for identifying fractures in this area). Injury 2005;36:14807.
[2] Piper KJ, Won HY, Ellis AM. Hybrid external xation in complex tibial plateau
and plafond fractures: an Australian audit of outcomes. Injury 2005;36.
[3] Marsh JL, Nepola JV, Wuest TK, Osteen D, Cox K, Oppenheim W. Unilateral
external xation until healing with the dynamic axial xator for severe open
tibial fractures. J Orthop Trauma 1991;5:3418.
[4] Bonar SK, Marsh JL. Tibial plafond fractures: changing principles of treatment.
J AmAcad Orthop Surg 1994;2:297305.
[5] Kapukaya A, Subasi M, Arslan H. Management of comminuted closed tibial pla-
fond fractures usingcircular external xators. ActaOrthopBelg2005;71:5829.
[6] Mittal R, Matthews SJ, Zavras DT, Giannoudis PV. Management of ipsilat-
eral pilon and calcaneal fractures: a report of 2 cases. J Foot Ankle Surg
2004;43:12330.
[7] Prayson MJ, Moon BS. Stabilization of the fractured tibial plafond. Oper Tech
Orthop 1999;9:21628.
[8] Borrelli JJ, Ellis E. Pilon fractures: assessment and treatment. Orthop Clin N Am
2002;33:23145.
[9] Spinosa FA. Chapter 19classication of fractures and dislocations. Foot Ankle
Radiol 2003:41551.
[10] Bartolozzi P, Lavini F. Fractures of the tibia pilon. Milan: Springer; 2004.
[11] Anwar R, Tuson KWR, Khan SA. Classication and diagnosis in orthopaedic
trauma. NewYork: Cambridge University Press; 2008.
[12] Hopton BP, Harris NJ. Fractures of the foot and ankle. Surgery 2010;28:5027.
[13] Ruedi T, Matter P, Allgower M. Intra-articular fractures of the distal tibial end.
Hel Chir Acta 1968;35:55682.
[14] Ruedi T. The treatment of displaced metaphyseal fractures with screws and
wiring systems. Orthopedics 1989;12:559.
[15] Ruedi T, Allgower M. The operative treatment of intra-articular fractures of the
lower end of the tibia. Clin Orthop Relat Res 1979;138:10510.
[16] Ruedi T. Fractures of the lower end of the tibia into the ankle joint: results 9
years after opne reduction and internal xation. Injury 1973;5:1304.
[17] Liporace FA, Yoon RS. An adjunct to percutaneous plate insertion to obtain
optimal sagittal planealignment inthetreatment of pilonfractures. J Foot Ankle
Surg 2012;51:2757.
[18] Dudko S, Kusz D, Wojciechowski P, Stoltny T. Operative treatment of ankles
fractures using internal osteosynthesis by a minimal surgical approach. Foot
2004;14:18591.
[19] Mauffrey C, Vasario G, Battiston B, Lewis C, Beazley J, Seligson D. Tibial pilon
fractures: a reviewof incidence, diagnosis, treatment, and complications. Acta
Orthop Belg 2011;77:43240.
[20] Pellegrini M, Cuchacovich N, Lagos L, Henriquez H, Carcuro G, Bastias C.
Minimally-invasive alternatives in the treatment of distal articular tibial frac-
tures. Fur Sprunggelenk 2012;10:3745.
[21] Wall OR, Pinder R, Faraj AA. Enders nail xation of tibial pilon fracturesa safe,
minimally invasive approach for high risk patients in a small district general
hospital. Inj Extra 2007;38:8.
[22] Etter C, Ganz R. Long-termresults of tibial plafond fractures treated with open
reduction and internal xation. Arch Orthop Trauma Surg 1991;110:27783.
[23] Picanz J. Poor results mark ORIF of tibial plafond fractures. Orthop Today
1990;10:12.
[24] Teeny S, Wiss DA, Hathaway R, Sarmiento A. Tibial plafond fractures:
errors, complications, and pitfalls in operative treatment. Orthop Trans
1990;14:26571.
[25] Rammelt S, Marti RK, Raaymakers ELFB, Grass R, Zwipp H. Joint preserving
reconstruction of malunited pilon fractures. Fur Sprunggelenk 2012;10:6272.
[26] Sirkin M, Sanders R, DiPasquale T, Herscovici D. Astaged protocol for soft tissue
management in the treatment of complex pilon fractures. J Orthop Trauma
1999;13:7884.
[27] Cheema GS, Arora S, Sabat D, Singla J, Goel N, Maini L. The results of two-staged
operative management of pilon fracturesa review of 25 cases. J Clin Orthop
Trauma 2011;2:1048.
Please cite this article in press as: Ramlee MH, et al. Finite element analysis of three commonly used external xation devices for treating
Type III pilon fractures. Med Eng Phys (2014), http://dx.doi.org/10.1016/j.medengphy.2014.05.015
ARTICLE IN PRESS
G Model
JJBE-2506; No. of Pages 9
M.H. Ramlee et al. / Medical Engineering & Physics xxx (2014) xxxxxx 9
[28] Barbieri R, Schenk R, Koval K, Aurori K, Aurori B. Hybrid external xation in the
treatment of tibial plafond fractures. Clin Orthop Relat Res 1996;332:1622.
[29] Bone L, Stegemann P, McNamara K, Seibel R. External xation of severely com-
minutedandopentibial pilonfractures. ClinOrthopRelat Res 1993;292:1017.
[30] Mitkovic MB, Bumbasirevic MZ, Lesic A, Golubovic Z. Dynamic external xa-
tion of comminuted intra-articular fractures of the distal tibia (type C pilon
fractures). Acta Orthop Belg 2002;68:50814.
[31] Cheung JT-M, Zhang M, An KN. Effect of Achilles tendon loading on plantar
fascia tension in the standing foot. Clin Biomech 2006;21:194203.
[32] Wang CL, Cheng CK, Chen CW, Lu CM, Hang YS, Liu TK. Contact areas and
pressure distributions in the subtalar joint. J Biomech 1994;28:26979.
[33] Yosibash Z, Trabelsi N, Milgrom C. Realiable simulations of the human proxi-
mal femur by high-order nite element analysis validated by the experimental
observations. J Biomech 2007;40:368899.
[34] Chen WP, Tai CL, Shih CH, Hsieh PH, Leou MC, Lee MS. Selection of xation
devices in proximal femur rotational osteotomy: clinical complications and
nite element analysis. Clin Biomech 2004;19:25562.
[35] Millington SA, Grabner M, Wozelka R, Anderson DD, Hurwitz SR, Crandall JR.
Quantication of ankle articular cartilage topography and thickness using a
high resolution stereophotography system. Osteoarthr Cartil 2007;15:20511.
[36] Akiyamat K, Sakai T, SugimotoN, Yoshikawa H, SugamotoK. Three-dimensional
distribution of articular cartilage thickness in the elderly talus and cal-
caneus analyzing the subchondral bone plate density. Osteoarthr Cartil
2012;20:296304.
[37] Brown CP, Nguyen TC, Moody HR, Crawford RW, Oloyede A. Assess-
ment of common hyperelastic constitutive equations for describing normal
and osteoarthritis articular cartilage. Proc Inst Mech Eng H: J Eng Med
2009;223:64352.
[38] Li Z, KimJ-E, DavidsonJS, Etheridge BS, AlonsoJE, Eberhardt AW. Biomechanical
response of the pubic symphysis in lateral pelvic impacts: a nite element
study. J Biomech 2007;40:275866.
[39] Gislason MK, Stanseld B, Nash DH. Finite element model creation and stability
considerations of complex biological articulation: the human wrist joint. Med
Eng Phys 2010;32:52331.
[40] Bajuri MN, Abdul Kadir Mohammed Raq, Murali Malliga Raman, Kamarul T.
Mechanical and functional assessment of the wrist affected by rheumatoid
arhtritis: a nite element analysis. Med Eng Phys 2012;34:1294302.
[41] Ezquerro F, Jimenez S, Perez A, Prado M, de Diego G, Simon A. The inuence
of wire positioning upon the initial stability of scaphoid fractures xed using
Kirschner wires: a nite element study. Med Eng Phys 2007;29:65260.
[42] Tao K, Wang D, Wang C, Wang X, Liu A, Nester CJ, et al. An in vivo exper-
imental validation of a computational model of human foot. J Bionic Eng
2009;6:38797.
[43] Yu J, Cheung JT-M, Fan Y, Zhang Y, Leung AKL, Zhang M. Development of a
nite element model of female foot for high-heeled shoe design. Clin Biomech
2008;23. S31-S8.
[44] Netter FH. Atlas of human anatomy. 3rd ed NewYork City, USA: ICONLearning
System; 2003.
[45] Pfaefe HJ, Tomaino MM, Grewal R, Xu J, Boardman ND, Woo SL, et al. Tensile
properties of the interosseous membrane of the human forearm. J Orthop Res
1996;14:8425.
[46] Siegler S, Block J, Schneck CD. The mechanical characteristics of the collateral
ligaments of the human ankle joint. Foot Ankle 1988;8:23442.
[47] Beumar A, vanHemert WL, Swierstra BA, Jasper LE, Belkoff SM. Abiomechanical
evaluation of the tibiobular and tibiotalar ligaments of the ankle. Foot Ankle
Int 2003;24:4269.
[48] Liacouras PC, Wayne JS. Computational modeling to predict mechanical func-
tion of joints: application to the lower leg simulation of two cadaver studies. J
Biomech Eng 2007;129:8117.
[49] Iaquinto JM, Wayne JS. Computational model of the lower leg and foot/ankle
complex: application to arch stability. J Biomech Eng 2010:2010.
[50] Nakamura S, Crowninshield RD, Cooper RR. An analysis of soft tissue loading in
the foota preliminary report. Bull Prosthet Res 1981;18:2734.
[51] QiuTX, TeoEC, YanYB, Lei W. Finiteelement modelingof a3D coupledfoot-boot
model. Med Eng Phys 2011;33:122833.
[52] KimH-J, KimS-H, Chang S-H. Bio-mechanical analysis of a fractures tibia with
composite bone plates accordingto the diaphyseal oblique fracture angle. Com-
pos B: Eng 2011;42:66674.
[53] FanY, XiuK, DuanH, Zhang M. Biomechanical andhistological evaluationof the
application of biodegradable poly-l-lactic cushion to the plate internal xation
for bone fracture healing. Clin Biomech 2008;23:S716.
[54] Benli S, Aksoy S, Havitcioglu H, Kucuk M. Evaluation of bone plate with low-
stiffness material in terms of stress distribution. J Biomech 2008;14:322935.
[55] Cordey J, Borgeaud M, Perren SM. Force transfer between the plate and the
bone: relative importance of the bending stiffness of the screws andthe friction
between plate and bone. Injury 2000;31:218.
[56] Tajdari M, Javadi M. A new experimental procedure of evaluating the
friction coefcient in elastic and plastic region. J Mater Process Technol
2006;13:24750.
[57] Heintz S, Gutierrez-Farewik EM. Static optimization of muscle forces dur-
ing gait in comparison to EMG-to-force processing approcach. Gait Posture
2007;26:27988.
[58] Anderson FC, Pandy MG. Static and dynamic optimization solutions for gait are
practically equivalent. J Biomech 2001;34:15361.
[59] Kim S-H, Chang S-H, Son D-S. Finite element analysis of the effect of bending
stiffness and contact condition of composite bone plates with simple rectan-
gular cross-section on the bio-mechanical behaviour of fractures long bones.
Compos Part B 2011;42:17318.
[60] Cheung JT-M, Zhang M, Leung AKL, Fan YB. Three-dimensional nite element
analysis of the foot during standinga material sensitivity study. J Biomech
2005;38:104554.
[61] Simkin A. Structural analysis of the human foot in standing posture. Tel Aviv:
Tel Aviv University; 1982.
[62] Aarnes GA, Steen H, Kristiansen LP, Festo E, Ludvigsen P. Optimum load-
ing mode for axial stiffness testing in limb lengthening. J Orthop Res
2006;24:34854.
[63] Chandran P, Puttaswamaiah R, Dhillon MS, Gill SS. Management of complex
open fracture injuries of the midfoot with external xation. J Foot Ankle Surg
2006;45:30815.
[64] Zgonis T, Roukis TS, Polyzois V, Wukich DK. Surgical management of the unsta-
ble diabetic Charcot deformity using the Taylor spatial frame. Oper TechOrthop
2006;16:107.
[65] Izaham RMAR, Kadir MRA, Rashid AHA, Hossain MG, Kamarul T. Finite ele-
ment analysis of Puddu and Tomox plate xation for open wedge high tibial
osteotomy. Injury 2012;43:898902.
[66] Bergmann G, Deuretzbacher G, Heller M, Graichen F, Rohlmann A, Straus J,
et al. Hip contact forces and gait patterns from routine activities. J Biomech
2001;34:85971.
[67] Donaldson FE, Pankaj P, Simpson AHRW. Bone properties affect loosen-
ing of half-pin external xators at the pin-bone interface. Injury 2012;43:
176470.
[68] Brianza S, Brighenti V, Lansdowne JL, Schwieger K, Boure L. Finite element anal-
ysis of a novel pin-sleeve system for external xation of distal limb fractures
in horses. Vet J 2011;190:2607.
[69] Gardner TN, Mishra S. The biomechanical environment of a bone fracture
and its inuence upon the morphology of healing. Med Eng Phys 2003;25:
45564.
[70] Nash RA, Nunamaker DM, Boston R. Evaluation of a tapere-sleeve transcortical
pin to reduce stress at the bone-pin interface in metacarpal bones obtained
from horses. AmJ Vet Res 2001;62:95560.
[71] Richardson DW, Nunamaker DM, Sigafoos RD. Use of an external skeletal xa-
tion device and bone graft for arthrodesis of the metacarpophalangeal joint in
horses. J AmVet Med Assoc 1987;191:31621.
[72] Williams TM, Nepola JV, DeCoster TA, Hurwitz SR, Dirschl DR, Marsh JL.
Factors affecting outcome in tibial plafond fractures. Clin Orthop Relat Res
2004;423:938.
[73] Claes LE, Wilke HJ, Augat P, Rubenacker S, Margevicius KJ. Effect of dynamiza-
tion on gap healing of diaphyseal fractures under external xation. Clin
Biomech 1995;10:22734.
[74] Goodship AE, Kenwright J. The inuence of induced micromovement upon
healing of experimental tibial fractures. J Bone Jt Surg 1985;67B:6505.
[75] Chen G, Schmutz B, Wullschleger M, Pearcy MJ, Schuetz MA. Computational
investigations of mechanical failures of internal plate xation. Proc Inst Mech
Eng H: J Eng Med 2010;224:11926.
[76] Ebrahimi H, Rabinovich M, Vuleta V, Zalcman D, Shah S, Dubov A, et al.
Biomechanical properties of an intact, injured, repaired, and healed femur:
an experimental and computational study. J Mech Behav Biomed Mater
2012;16:12135.
[77] Moazen M, Mak JH, Etchels LW, Jin Z, Wilcox RK, Jones AC, et al. The effect of
fracture stability on the performance of locking plate xation in periprosthetic
femoral fractures. J Arthroplast 2013;28:158995.
[78] Gorsse S, Miracle DB. Mechanical properties of Ti6Al4V/TiB compos-
ites with randomly oriented and aligned TiB reinforcement. Acta Mater
2003;51:242742.
[79] RedfernDJ, SyedSU, Davies SJM. Fractures of the distal tibia: minimallyinvasive
plate osteosynthesis. Injury 2004;35:61520.