Critical FEM parameters for Pauwels III femoral neck fracture fixation: A computational study with cadaveric...

Pauwels type III femoral neck fractures are biomechanically challenging because the steep fracture orientation produces dominant shear forces at the fracture interface. This study evaluated key finite element (FE) modeling parameters for simulating fracture fixation and validated numerical predictions against cadaveric experiments. Human femora were fractured at 70° and tested under physiological axial loading to measure construct stiffness and interfragmentary motion (IFM). Eight finite element (FE) configurations were generated by systematically varying bone material representation (homogeneous versus heterogeneous), the presence or absence of screw compression load (SCL), and screw modeling strategy (simplified versus explicit thread geometry). Heterogeneous FE models incorporating SCL predicted construct stiffness within 15% of experimental measurements, whereas homogeneous models overestimated stiffness by up to 120% and substantially underestimated IFM. Inclusion of SCL eliminated opening IFM at 700 N, consistent with experimental observations, while omission of SCL resulted in persistent interface opening. In contrast, explicit modeling of screw threads had negligible influence (<5%) on global stiffness, IFM, or maximum von Mises stress, although it affected local stress distribution near the threads. These findings indicate that bone heterogeneity and fixation-induced compression play a decisive role in improving FE prediction accuracy, whereas detailed thread geometry may be simplified to improve computational efficiency without compromising global mechanical outcomes.