Prolonged exposure to whole body vibration (WBV) has been associated with prevalence of spinal disorders among operators of vibrating mobile machinery. The study of biodynamic responses of body segments is thus pertinent for our understanding of potential injury mechanisms and designs of interventions. This study concerns seated body biodynamic responses to vertical vibration through measurements at the drivingpoint and at body segments, and development of an analytical model for prediction of global and localised responses. Experiments were undertaken to simultaneously measure driving-point apparent mass (APMS) and body segment acceleration transmissibility of 12 adult subjects under random vertical vibration in the 0.5-20 Hz frequency range. Measurements were taken at the C7, T5, T12, L3 and L5 vertebral locations along the fore-aft and vertical axes using skin-mounted micro-accelerometers, and at the scalp using a light-weight head strap with a micro-accelerometer. The study involved four sitting postures realised with different combinations of hands position (on the lap or on the steering wheel) and back support (none or a vertical support), and three excitation magnitudes (0.25, 0.5 and 1 m/s² RMS). Mathematical correction methods were employed to account for skin effects, sensor misalignments, and seat inertia effects. The corrected body-segment responses of the twelve subjects depicted a clear dependence on the back support condition (p<0.01), while the influences of hand position and vibration magnitude were also significant but relatively weaker.

Owing to the significant influences of the postural parameters, it was concluded that different support-specific datasets would be necessary to describe the WBV responses and identification of biodynamic models. A 19 degrees-of-freedom anthropometric multi-body biodynamic (MBD) model of the 50th percentile male subjects was formulated on the basis of the known anthropometric inertial and joint properties to simulate sagittal plane motions of the body under vertical WBV. The visco-elastic parameters of various joints were identified through minimisation of a set of error functions derived from different combinations of target responses using the Genetic Algorithm. The minimisation of an error function based on measured vertical head, and fore-aft head and C7 vibration provided an acceptable convergence in primary resonance peaks in both the APMS and the segmental vibration responses. Eigen analysis of the resulting model revealed the presence of 4 significant modes at frequencies below 15 Hz, including two modes near the primary resonant frequency of 5 Hz (4.76 Hz and 5.71 Hz), corresponding to vertical movement of the whole body and pelvic rotation. The model was subsequently applied to estimate vibratory power absorbed within different joints and the total body. The total absorbed power of the model agreed reasonably well with the measured total power. The study revealed that a large portion of the power was absorbed at the body-seat interface, primarily by the buttock tissue. However, significant energy dissipation also occurred at the abdominal viscera and the lower lumbar joint (L5). The L5 was the only joint that showed relatively higher energy dissipation in translation as well as pitch rotation, which may be associated with the most widely reported location of pain and spinal injury