Comparison of simplified bone-screw interface models in materially nonlinear μFE simulations.

Area of Science:

• Biomechanics
• Computational modeling
• Materials science

Background:

• Micro finite-element (μFE) simulations are vital for assessing bone screw anchorage.
• Accurate modeling of the bone-screw interface at the microscale is challenging.
• Current gold-standard methods use computationally intensive physical contact models.

Purpose of the Study:

• To compare the accuracy and efficiency of simplified interface models against a gold-standard general contact model in μFE simulations.
• To evaluate the predictive capabilities of fully-bonded, tensionally strained element deletion (TED), and TED-M interface approaches.
• To assess the impact of these models on predicting deformations, stiffness, maximum force, and damage patterns.

Main Methods:

• Nonlinear μFE models were created using micro-CT images of human radii.
• A screw was virtually inserted, and models were subjected to tension, compression, and shear loads.
• Three simplified interface models (fully-bonded, TED, TED-M) were compared to a general contact (gold-standard) interface in Abaqus Explicit.

Main Results:

• The fully-bonded interface overestimated stiffness (19%) and maximum force (26%), with inaccurate damage patterns.
• Both TED and TED-M reduced errors in stiffness and force, improving damage prediction accuracy.
• TED yielded superior whole-construct stiffness predictions (1% error), while TED-M excelled in maximum force prediction (1% error).

Conclusions:

• TED and TED-M provide computationally efficient and accurate alternatives to physical contact models for bone-screw interface simulation.
• While fully-bonded interfaces may suffice for some applications, TED and TED-M offer enhanced predictive capabilities.
• These simplified models can significantly reduce computational cost without substantial loss of accuracy in biomechanical assessments.