Version HTML de base
45
The long-term exposure of the human body to vibrations
may lead to mechanical fatigue [18-20] and low-back prob-
lems due to microfractures in bones (cortical and cancel-
lous), on endplates and microlesions in the intervertebral
discs [21]. Even if the dynamic loading amplitude can
appear as low, such a repetitive loading induces mechanical
stresses that may cause microfractures in bone and low-
back pain after a long term exposure with a large number
of excitation cycles. The risk of damage by mechanical
fatigue may be considered not significant for young driv-
ers due to the regeneration effect, but it can be critical
for old drivers due to the loss of the mechanical proper-
ties of bones (Young’s modulus, density, ultimate stress)
and intervertebral discs (damping) [22-25].
- The first experiment on fatigue fracture of the lumbar was
reported by Hardy et al [26]. The authors examined the
effect of cyclic axial compression and cyclic axial transverse
bending on ten fresh and five embalmed human lumbar
spines (five lumbar vertebrae and intervertebral discs).
The compressive cyclic load ranged between 500 and
4500 N at a loading rate of 2 Hz. Compression fractures
of vertebrae occurred after 1 290 000 cycles. Annulus
injury was not observed in this loading mode.
- Adams et al [27] performed cyclic flexion fatigue tests
at a frequency of 0.67 Hz on 41 human cadaveric lumbar
intervertebral joints. The mean peak load applied to the
segments was 3076 N, individual loads being based on
the age, sex, and body weight of the respective cadaver.
27% of the specimens failed either with endplate fractures
or anterior crush within 9600 cycles. They also observed
formation of poster lateral radial fissures in the annulus
of segments with degenerated discs.
- Liu et al [28] tested eleven human lumbar interverte-
bral joints. Cyclic axial load at a frequency of 0.5 Hz was
applied between 37% and 80% of the ultimate compres-
sive strength of vertebral bodies. The experiment was
conducted at room temperature. The authors noted in
five specimens, loaded between 60% and 80% of the
mean ultimate compressive strength, an abrupt increase
in the maximum compressive displacement at load cycle
numbers lower than 2000. This increase in compressive
displacement, equivalent to an irreversible height loss of
the specimens, was interpreted as a sign of compression
fracture. Fractures were observed in trabecular bone and
in endplate.
- Hansson et al. [14] exposed 17 lumbar motion segments
from eight spines to a 0.5 Hz sinusoidal dynamic compres-
sive loading regime. Specimens were aged between 37
and 82 years. Testing was performed at room temper-
ature. The applied load varied between 60% and 100%
of the ultimate compressive strength. Failures occurred
between 1 and 950 cycles. Their observed failure damage
was Schmorl’s node and central endplate fracture – i.e.
primarily related to failure of the trabecular bone beneath
the endplate.
- Brinckmann et al [13] carried out an extensive study on 70
lumbar motion segments exposed to a 2s rise time triangu-
lar (0.25Hz) compressive regime. Testing was performed
at 37Cº. The applied load varied between 20% and 70%
of the ultimate compressive strength. Tested failure was
recorded when a step occurred in the deformation–time
curve. This usually resulted in extrusion of bone marrow.
Most of the fractured specimens had damages at the
endplate. They found that at loads less than 30% static
strength, failure was rare and they argued that a normal-
ised stress of 30% could be regarded as an endurance
limit for in vivo exposure.
- Gallagher et al [12] tested thirty-six human lumbar motion
segments. Fatigue tested using spinal compressive and
shearloads that simulated lifting a 9 kg weight in three
torso flexion angles (0°, 22.5° and 45°). The equivalent
loads ranging 25% to 60% the ultimate compressive load.
Motion segments were creep loaded for 15 min and then
cyclically loaded at 0.33 Hz until failure or the maximum
number of cycles (10000) was completed. 25 of the 36
segments failed via fatigue prior to the 10 000 cycle maxi-
mum under applied loads ranging from 40% to the 60%.
These specimens were visually inspected and dissected
so that the mode of failure could be determined. Failure
modes included endplate fractures, vertebral body frac-
tures, and/or zygapophysial joint disruption. A new classi-
fication scheme characterizing the nature of the endplate
was developed in this investigation. The classification
scheme of the endplate fracture completes those identi-
fied by Brinckmann et al. [13].
The vertebral endplate appears to be the tissue most likely
to experience initial failures in tests of the both ultimate
compressive strength and fatigue failure [12]. Vertebral
body fractures and disruption of the zygapophysial joints
are also observed in compressive loading. However, fail-
ures of the intervertebral discs are less frequent. Discs
may fail when compressed in flexion [12, 27] or combined
with an axial twist, or as the result of the cascade initiated
by the endplate failure. Research using porcine models
has suggested that disc herniation can also be caused by
repeated flexions and extensions under moderate compres-
sive loads. Table 1 summarized the fatigue test specifica-
tions found in literature and describes the various parts
of lumbar spine that have been affected.
On the other hand, several authors have developed numer-
ical models to describe the fatigue behaviour of corti -
cal bone [29-35] and the fatigue behaviour of trabecular
bone. Guo et al [34] has modelled trabecular bone as
an idealised two-dimensional honeycomb- like structure
made up of an array of hexagonal cells. Each trabecula
was modelled as a linear elastic beam with the prop-
erties of cortical bone. Taylor et al [35] has simulated
the fatigue behaviour of cancellous bone based on the
assumption that the fatigue behaviour of trabecular
bone is similar to that of cortical bone using continuous
damage mechanics (CDM), accounted for both modulus
degradation and the accumulation of permanent strain,
with a FE approach.
On a global level, the fatigue behaviour of whole vertebrae
depends on the interaction of behaviour of all components
(cortical and trabecular bone, endplate, intervertebral
disc). Due to the complexity of vertebrae, few numerical
models were developed in order to simulate the fatigue
behaviours of whole vertebrae. Since data on the endur-
ance limits of vertebrae in the high cycle range are not
available; therefore, results of testing other bones (corti-
cal and trabecular bone, cartilage, and intervertebrale
disc) were used to simulate the fatigue behaviour of whole
lumbar vertebrae. Then the main objective in this study is
to complete the endurance fatigue behaviour information
of lumbar vertebrae exposed to vertical vibration.