Date of Award

2014

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Graduate Group

Mechanical Engineering & Applied Mechanics

First Advisor

Dawn M. Elliott

Abstract

Finite element models are advantageous in the study of intervertebral disc mechanics as the stress-strain distributions can be determined throughout the tissue and the applied loading and material properties can be controlled and modified. However, the complicated nature of the disc presents a challenge in developing an accurate and predictive disc model, which has led to limitations in finite element geometries, material constitutive models and properties, and model validation. The objective of this dissertation is to develop a new finite element model of the intervertebral disc, to validate the model's nonlinear and time-dependent responses without tuning or calibration, and to evaluate the effect of changes in nucleus pulposus and cartilaginous endplate material properties on the disc mechanical response. This was accomplished through a cohesive series of studies. First, structural hyperelastic constitutive models were used in conjunction with biphasic-swelling theory to obtain material parameters for the disc tissues from recent tissue tests. A new disc finite element model was then constructed utilizing an analytically-based geometry created from the mean shape of human L4/L5 discs, measured from high-resolution 3D MR images and averaged using signed distance functions. The full disc model was then validated against experimental intervertebral disc loading datasets for compressive slow loading ramp, creep, and stress-relaxation simulations, and finally the new disc model was used to investigate the role of each individual disc tissue. The significance of this new disc model is threefold. First, an extensive validation was performed using the full nonlinear response of the intervertebral disc in three different loading modalities. The finite element predictions fit within the experimental range (mean ±95% confidence interval) of the nonlinear response. Second, the validation was predictive; no material parameters were determined using fits to any motion-segment data. All parameters were obtained from fits to the individual tissue responses. Furthermore, the loading mechanisms tested at the tissue level (confined compression, uniaxial tension) were different than those implemented at the full disc scale (quasi-static slow ramp, creep, stress-relaxation). Lastly, model validation was accomplished without any "tuning" or adjustment of the material parameters in order to force agreement between the FE and experimental responses.

Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS