Biomechanical comparison of four patient-specific titanium implant designs for hemimandibular reconstruction: a finite element analysis

Mandibular defects remain difficult to reconstruct, and the optimal geometry of patient-specific 3D-printed titanium implants is uncertain. This study compared four implant designs for hemimandibular reconstruction under progressively increasing bilateral masseter loading. A three-dimensional mandibular model was reconstructed from cone-beam computed tomography of an adult male and segmented into cortical and cancellous bone. A hemimandibular defect (Jewer-Boyd lateral with condyle, not crossing the midline) was simulated. Four titanium implant designs were evaluated: Solid, Custom Tray, Lattice, and Reconstruction Plate. Each assembly was meshed in ANSYS Workbench using 3D tetrahedral elements, with final element sizes ranging from 0.55 to 1.0 mm and 1.08 to 1.21 million nodes per model. Static linear analysis was performed with masseter loading of 100 N, 300 N, and 500 N applied to each masseter on both sides. Both condyles were fully constrained, and occlusal contacts at incisors, canines, and molars were restricted vertically. Maximum principal stress was assessed in cortical bone. Equivalent von Mises stress was assessed in cancellous bone, the titanium implants, and the screws. All implant and screw stresses remained below the Ti-6Al-4 V yield strength of 830 MPa. At 500 N per side, peak implant stress was lowest in the Solid design (105.12 MPa) and highest in the Reconstruction Plate (279.19 MPa). Calculated cortical maximum principal stress was above the selected 115 MPa reference threshold in all four models: Custom Tray, 125 MPa; Reconstruction Plate, 130.75 MPa; Solid, 197.25 MPa; and Lattice, 215.03 MPa. The Custom Tray produced the lowest calculated peak cortical stress, whereas the Lattice produced the highest. Cancellous bone stresses remained below 32.4 MPa. Peak stresses were consistently localised around fixation sites adjacent to the resection. Within this static, single-anatomy, unvalidated model, the four designs showed distinct implant and peri-implant stress distributions. The Custom Tray produced the lowest calculated peak cortical stress, while the Solid design produced the lowest implant stress. Fixation regions were the most consistent sites of peak stress. These findings represent relative computational comparisons rather than evidence of clinical superiority.