Biomedical Engineering Reference
In-Depth Information
11.3
FInIte element analySIS oF the IntaCt PelvIS
11.3.1 f
inite
e
lement
m
odelinG
and
a
nalySiS
The pelvis is a complex and irregular structure that cannot be subject to oversimplification. In some
previous studies, two-dimensional or quasi-three-dimensional models were constructed to tackle
clinical problems (Huiskes 1987). The over-simplified structure was unable to simulate complex
operations and provide accurate results. With the development of advanced medical imaging tech-
niques and high-performance computers, any human organ can be accurately modeled in three
dimensions (Liang et al. 2011; Cheung et al. 2005). A male cadaver (age 30 years; mass 65 kg;
height 172 cm) free from any pathological abnormalities, trauma, or deformity in the pelvis or lower
extremities was used in this study. Transverse CT images of this subject were acquired at 1.25-mm
intervals from the third lumbar vertebra to the middle tibia. A total of 960 images were acquired, of
which 620 were used to construct the FE model.
Model segmentation was carried out with MIMICS 12.0 (Materialise, Inc., Leuven, Belgium)
on the sacrum and bilateral iliac bones. The cortical and cancellous bones were also distinguished
effectively as different point clouds. The point cloud files were imported into Geomagic Studio 9.0
(Raindrop Geomagic, Research Triangle Park, North Carolina) and converted into polygonal sur-
face models through the point phase, polygon phase, and shape phase.
The solid models were then imported into ANSYS 12.0 (ANSYS Inc., Canonsburg, Pennsylvania),
in which a Boolean operation was used to produce cortical and cancellous models for various bones.
All bones were meshed with three-dimensional 10-node tetrahedron structural solid elements
(SOLID92). This element has a quadratic displacement behavior and is capable of modeling irregu-
lar meshes. To simplify the analysis, all tissues were idealized as homogeneous, linearly elastic,
and isotropic. The Young's modulus and Poisson ratio of all tissues are listed in Table 11.4. The
parameters were all adopted from Bodzay, Flóris, and Váradi (2011).
The symphysis pubica was modeled with three-dimensional link elements. The nodes at the
opposite iliac bones were selected first and exported as two groups with the space coordinates.
A C-language program was then used to search for a matching for each point in either group. A
matching was defined as two points between which the displacement was the shortest than any
other couples. This method has been previously used to simulate cartilage and soft tissues (Liang
et al. 2011).
The FE model of the intact pelvis is shown in Figure 11.5. The TI was fixed in six degrees of
freedom to simulate a sitting posture. A compressive force of 500 N was vertically loaded on the
lumbosacral disc to simulate normal body weight (Jia et al. 2008). The FE analysis was carried out
with ANSYS software. The von Mises stress of the intact model is detailed in Figure 11.6. Stress
concentrations occurred mainly around the arcuate line, acetabulum, isciadic ramus, SIJ, and TI.
The maximum von Mises stresses in the sacrum and right and left iliums were 5.40, 6.21, and
5.84 MPa, respectively. The stress was distributed symmetrically to a great extent. Considering
the physiological asymmetry, modeling, and receivable computational error, the computation was
encouraging.
table 11.4
material Properties of tissues in the Finite element models
young's modulus (mPa)
Poisson ratio
Cortical bone
17,000
0.3
Cancellous bone
400
0.2
Sacroiliac joint
68
0.2
Symphysis pubica
50
0.2
Search WWH ::
Custom Search