Abstract
Children with cerebral palsy often experience gait impairments that reduce mobility and functional independence. This study presents a preliminary computational comparison of four manually developed PEEK-based hip-joint configurations intended for a pediatric lower-limb rehabilitation exoskeleton. A planar kinematic-static model was used to calculate reaction forces and hip-joint moments for equivalent static load-direction angles ranging from 0° to 44°. The four configurations were then evaluated by finite element analysis using a homogeneous, isotropic, and linearly elastic PEEK model. The calculated hip-joint moment decreased from 392.66 N·m at 0° to 153.13 N·m at 44°. Under the highest-moment static case, Model 3 produced the lowest resultant displacement of 2.992 mm, whereas Model 4 produced the lowest calculated maximum von Mises stress and equivalent strain, with values of 100.1 MPa and 0.01509, respectively. The stress calculated for Model 4 corresponded to a yield ratio of approximately 1.10 relative to the assumed PEEK yield strength of 110 MPa. This small numerical margin does not establish structural safety. The results are limited by the absence of mesh-convergence verification, dynamic and CP-specific loading, manufacturing anisotropy, fatigue analysis, and experimental validation. Model 4 is therefore reported only as the configuration with the lowest calculated peak stress under the adopted numerical conditions. Further numerical verification and mechanical testing are required before practical pediatric use can be considered.