Homogenized finite element models can accurately predict screw pull-out in continuum materials, but not in porous materials.

BACKGROUND AND OBJECTIVE

Bone screw fixation can be estimated with several test methods such as insertion torque, pull-out, push-in and bending tests. A basic understanding of the relationship between screw fixation and bone microstructure is still lacking. Computational models can help clarify this relationship. The objective of the paper is to evaluate homogenized finite element (hFE) models of bone screw pull-out.

METHODS

Experimental pull-out tests were performed on three materials: two polyurethane (PU) foams having a porous microstructure, and a high density polyethylene (HDPE) which is a continuum material. Forty-five titanium pedicle screws were inserted to 10, 20, and 30 mm in equally sized blocks of all three materials (N = 5/group). Pull-out characteristics i.e. stiffness (S), yield force (F), peak pull-out force (F) and displacement at F (d) were measured. hFE models were created replicating the experiments. The screw was modeled as a rigid body and 5 mm axial displacement was applied to the head of the screw. Simulations were performed evaluating two different conditions at the bone-screw interface; once in which the screw fitted the pilot hole exactly ("free-stressed") and once in which interface stresses resulting from the insertion process were taken into account ("pre-stressed").

RESULTS

The simulations representing the pre-stressed condition in HDPE matched the experimental data well; S, F, and F differed less than 11%, 2% and 0.5% from the experimental data, respectively, whereas d differed less than 16%. The free-stressed simulations were less accurate, especially stiffness (158% higher than the pre-stressed condition) and d (30% lower than pre-stressed condition) were affected. The simulations representing PU did not match the experiments well. For the 20 mm insertion depth, S, F and F differed by more than 104%, 89% and 66%, respectively from the experimental values. Agreement did not improve for 10 and 30 mm insertion depths.

CONCLUSIONS

We found that hFE models can accurately quantify screw pull-out in continuum materials such as HDPE, but not in materials with a porous structure, such as PU. Pre-stresses in the bone induced by the insertion process cannot be neglected and need to be included in the hFE simulations.