FE modeling of the foot and implant constructs
A previously validated three-dimensional FE model [26, 27] was modified to investigate the impact of different implant material combinations on the stress distribution of the implant and the foot after TAR. The foot model included thirty bones and enveloped soft tissue. All bones were connected by 134 major ligaments and a plantar fascia. The ligaments were represented by spring elements with a ‘no compression’ option. The 3D geometries of the plantar fascia and Achilles tendon were constructed. The other five muscle tendons, namely tibialis posterior (TIBP), flexor hallucis longus (FHL), flexor digitorum longus (FDL), peroneus brevis (PB), and peroneus longus (PL), were also incorporated into the model using bar elements, at their corresponding anatomical attachment sites. Articular surfaces were modeled in the foot, with surface-to-surface contact elements.
The plantar soft tissue was modeled by an incompressible Ogden hyperelastic material [26]. The strain energy function U of the first-order Ogden model was defined by:
Table 1 Material property and element type of the foot model
|
Elastic modulus (MPa)
|
Poisson ratio
|
Cross-sectional area (mm2)
|
Element type
|
Bone
|
7300
|
0.3
|
-
|
4-node tet
|
Cartilage
|
1.01
|
0.4
|
-
|
4-node tet
|
Ligament
|
260
|
0.4
|
18.4
|
Spring
|
Plantar Fascia
|
350
|
0.4
|
-
|
4-node tet
|
Achilles tendon
|
816
|
0.3
|
-
|
4-node tet
|
Flexor tendon
|
450
|
0.3
|
12.5
|
Connector
|
Plantar soft tissue
|
1st Ogden incompressive hyperelastic model
=0.0375 MPa, =5.5
|
4-node tet
|
Ground
|
Rigid
|
8-node hex
|
TAR implant modeling and implantation
In this study, we chose to replace a FE foot model with the INBONE II Total Ankle System (Wright Medical Technology, Inc., Memphis, TN, US) (see Fig. 1), the only ankle implant system currently used in China. It has a medullar segmented stem, assembled with a tibial tray, and a symmetric fixed bearing with full conformity to the talus component [34]. To the best of the authors’ knowledge, no studies have previously investigated the biomechanical performance of this ankle system.
To match the shape of the current foot model, we chose and reverse engineered a size two long tibial component, a size two polyethylene insert with 6 mm thickness, and a size two talar component of the implant system. The materials of the tibial component, bearing, and talar component were Ti6Al4V, UHMWPE, and CrCoMo, respectively. Tibia and talus were resected with the protection of medial malleolus, and this ankle system was implanted following its operative guideline [35] from the manufacturer. The implantation procedure was guided and checked by two senior foot and ankle surgeons (XM and XW). The mesh size of the implant on contact biomechanics was determined by mesh convergence test (Supplementary Table S1). 1 mm mesh size at the articular surface, and 1.5 mm at the rest of the implant were chosen for model accuracy. UHMWPE was modeled as an isotropic elastic-plastic material [36] with a yield stress of 10.86 MPa and a Poisson’s ratio of 0.46. The stress-strain curve of the UHMWPE was presented in Fig. 2. The rest of the implant materials were modeled as isotropic elastic materials [10, 37, 38], and their mechanical properties were listed in Table 2. In this study, four types of materials were used to model tibial and talar components (Ceramic, Co-Cr-Mo, Ti6Al4V, and CFR-PEEK) while three types of materials were used to model bearing insert (CFR-PEEK, PEEK, and UHMWPE).
Table 2 Material property of the INBONE II Total Ankle System
Implant Material
|
Elastic modulus (MPa)
|
Poison ratio (ν)
|
Ceramic
|
350,000
|
0.26
|
Co-Cr-Mo
|
210,000
|
0.3
|
Ti6Al4V
|
114,000
|
0.342
|
CFR-PEEK
|
18,000
|
0.4
|
PEEK
|
4,100
|
0.36
|
UHMWPE
|
Stress-strain relationship [36]
|
0.46
|
Loading and boundary condition
Frictional contact interaction among the plantar surface and the ground was defined using a coefficient of friction (COF) of 0.5 [26]. The interaction between bone and implant was defined as a tie condition to simulate in-bone growth effects. The tibial tray and bearing were tied since its fix-bearing design. The bearing and talus interface was defined as frictionless [23].
We analyzed the biomechanical characteristics of the TAR implant at a time during the gait cycle when the ankle force reached a peak, namely the second peak of the ground reaction forces (GRF) associated with walking. A rigid plate under the foot was used to model the ground support. The ground was constrained to allow movement in a vertical direction only. The loading condition has previously been established in our foot FE model [26, 27]. In brief, a targeted maximum vertical GRF (623.1 N for the subject with a bodyweight of 60 kg) was generated solely by contracting plantar flexors corresponding to the push-off phase in gait. The obtained convergent solution of the muscle forces that maintain the second metatarsal shaft orientation was 1620 N of GS complex, 267 N of the TIBP muscle, 130 N of the FHL muscle, 81 N of the FDL muscle, 91N of the PB muscle, and 193N of the PL muscle. Muscle forces of major plantar flexors were applied via the tendons attached. To only investigate the impact of implantation and implant material, we assumed all models shared the same as the boundary and loading conditions. The complete model was illustrated in Fig. 3.
Data Analysis
All FE models were solved in ABAQUS (Simulia Corp., Providence, RI, USA). The von Mises stress distribution on each bone, plantar surface pressure, von Mises stress distribution at the articular surface, and the resected surface of the talus were evaluated for each material combination.