Biomedical Engineering Reference
In-Depth Information
Fig. 7.54
Obtained trabecular architecture for the selected apparent density control values
The essential and natural boundary conditions are suggested in the work of
Chou et al. [
28
]. The implant system is submitted to an occlusal load F
0
= 100 N
applied directly in the crow, inclined 11 in relation to the implant longitudinal
axis. A second load condition is suggested: an uniform distributed pressure along
the outer surface of the cortical bone, q
0
¼ 500 kPa, intended to simulate the effect
of the mandibular flexure [
28
] and a more realistic overall boundary condition
[
27
]. Regarding the essential boundary conditions, the model is constrained in the
basis, along x and y directions. In Fig.
7.53
c it is possible to observe the referred
boundary conditions.
Three distinct medium bone density control values were assumed: q
control
app
¼
0
:
90 g/cm
3
;
q
control
app
app
¼ 0
:
40 g/cm
3
. Notice that the process
stops when the medium bone density reaches the control value.
The obtained trabecular distributions for each one of the control values are
presented in Fig.
7.54
and the respective von Mises effective stress distributions is
presented in Fig.
7.55
. A permanent cortical bone perimeter with a 0.5 mm-
0.8 mm thickness was considered as suggested in the literature [
27
,
28
].
The results show that the proposed remodelling algorithm, combined with the
NNRPIM, is capable of reproducing trabecular distributions very close to the FEM
results, Fig.
7.56
. The remodelling algorithm was able to predict the same higher
apparent density regions on the mandible patch. Under the implant it is possible to
observe high-density horizontally oriented regions connecting the cortical layers
on the bone periphery. Also the bone resorption, immediately below the implant,
and the diagonal trabeculae connecting the implant structure to the peripheric
cortical layers are predicted by the bone tissue remodelling algorithm. A closer
¼ 0
:
65 g/cm
3
and q
control
