Introduction: Whole-body vibration, commonly experienced in the workplace, may increase the prevalence of low back pain. Finite element (FE) models are supposed to point out the coherence between vibrations and damages, but they need to be improved and validated. The aim was to assess shear strength, dynamic shear and compressive stiffness, and compressive fatigue strength of human functional spinal units (FSUs) with relation to age, posture and individual characteristics. The findings were incorporated in an individualized FE model of FSUs.

Characterisation: Specimens from age group 20-44 yrs (Young) and 48-64 yrs (Old) were characterised by anthropometric data of donors and by CT based parameters, such as endplate area (AREA) and bone mineral density (BMD). Potting of FSUs was done in 0° (Neutral) and 10° (Flexed) posture.

Shear strength: L2-L3 FSUs were sheared anteriorly, while a physiological compressive load was applied. The testing groups were: Young-Neutral, Young-Flexed, Young-Creep (additional creep for 1 h) and Old-Neutral. Specimens in flexion tended to have higher failure strength. Specimens tested in neutral posture exhibited increased shear strength with increasing BMD.

Dynamic stiffness: L4-L5 FSUs were exposed to loading varying in frequency (up to 12 Hz), offset loads and amplitudes. The directions were anterior-posterior shear and axial compression. The groups were Young-Neutral, Young-Flexed and Old-Neutral. The energy ratio and the linearised stiffness in the quasi-static compression tests was the highest for Young-Flexion. Shear pre-load decreased the axial stiffness. The shear quasi-static stiffness however increased with increasing axial compression preload. Axial and shear stiffness increased with increasing test frequency and axial stiffness decreased with increasing amplitude.

Modelling: A partially individualised FE model of L4-L5 FSUs was created. It incorporates 23 individual geometric parameters derived from CT data. The annulus of the intervertebral disc was modelled as a ground substance with pre-stressed fibres and the nucleus as a viscoelastic liquid-filled cavity. The individualisation improved the modelling, but the behaviour shown in the experiments cannot be completely described yet. The results indicate that individual material parameters and individual descriptions of the facet joints should be taken into account as well.

Fatigue: The three groups of L4-L5 FSUs were additionally loaded in compression for 300,000 sinusoidal cycles (5 Hz) to initiate fatigue failure. The fatigue strength of young donor's specimens was unexpectedly high. None of them failed in neutral posture. Four specimens from older donors with low BMD failed if exposed to high physiological loads; there was an exponential relationship between cycles to failure and the product of AREA and BMD. Only two of Young-Flexed failed. One had low BMD and was in line with the relationship derived for the specimens from older donors, whereas the second with normal BMD did not. Age and individual characteristics should be considered in the analysis of fatigue strength and thus of whole-body vibration injuries.

Conclusions: Fundamental data that improve knowledge of spinal behaviour under various loading conditions were delivered. Individualisation of FE models is an important step to generate improved assessment and ultimately improve workplace safety.