Finite element simulation of cement-bone interface micromechanics: a comparison to experimental results

Abstract

Recently, experiments were performed to determine the micromechanical behavior of the cement-bone interface under tension-compression loading conditions. These experiments were simulated using finite element analysis (FEA) to test whether the micromechanical response of the interface could be captured in micromodels. Models were created of experimental specimens based upon microcomputed tomography data, including the complex interdigitated bone-cement morphology and simulated frictional contact at the interface. The models were subjected to a fully reversed tension-compression load, mimicking the experimental protocol. Similar to what was found experimentally, the simulated interface was stiffer in compression than in tension, and the majority of displacement was localized to the cement-bone interface. A weak correlation was found between the FEA-predicted stiffness and the stiffness found experimentally, with average errors of 8 and 30% in tension and compression, respectively. The hysteresis behavior found experimentally was partially reproduced in the simulation by including friction at the cement-bone interface. Furthermore, stress analysis suggested that cement was more at risk of fatigue failure than bone, concurring with the experimental observation that more cracks were formed in the cement than in the bone. The current study provides information that may help explain the load transfer mechanisms taking place at the cement-bone interface.

Figure 1

a) Image of an experimental specimen in which 12 DIC sampling points are indicated b) μCT slice and c) an FEA model of the cement-bone interface specimen. The boundary conditions as indicated are simplified with respect to the actual applied conditions.

Figure 2

Bubble plot for measurement of gaps at the cement-bone interface from μCT scan data and FEA mesh. Bubble size is proportional to the number of measurements. Data are based on 32 sampling points per specimen using a stereology approach.

Figure 3

Experimental and computational force-displacement at the contact interface of a single specimen. The arrows indicate the loading direction. For illustrative purposes, the experimental and computational curves have been shifted to make them pass the point of zero stress and displacement.

Figure 4

Experimental versus FEA predicted deformation of the bulk cement and bone. The deformations are expressed relative to the total deformation of the cement-bone interface specimens.

Figure 5

Maximal principal stress distribution in an FEA model with little (left) and a lot of interdigitation (right), under maximal tension load. The enlargements show bone and cement structures that experienced high stress levels, indicating that these structures contributed to load transfer over the cement-bone interface.

Figure 6

Stress distribution in the bone and cement during maximum tension (top) and compression (bottom). The error bars indicate the standard deviation of the data for each stress level.