Abstract: BACKGROUND: The stability of the pelvis is mainly determined by the posterior pelvic ring and the sacroiliac joint. The posterior pelvic ring injury and the dislocation of the sacroiliac joint caused by high energy impacts such as car accidents increase year by year. Surgical treatment is the best method, and there are many kinds of endophytorepair methods in clinical practice, but which treatment method has the best biomechanical properties is still controversial.  
OBJECTIVE: To compare the biomechanical properties of three kinds of internal implants: anterior double plates, posterior bridging plate and tension nail in the repair of unilateral unstable pelvic posterior ring injury, to provide a reference for the clinical treatment and development of a new pelvic tension screw.
METHODS: (1) Finite element simulation: Mimics, Wrap and SolidWorks were used to establish normal pelvic model, unilateral injured pelvis model, and three kinds of internal implant repaired models (anterior double plates, posterior bridging plate and tension nail). Ansys was used to analyze the stress and deformation of the models. (2) Biomechanical test: A total of 15 intact pelvic specimens were randomly grouped into five groups, normal pelvic model, unilateral injured pelvis model, anterior double plates, posterior bridging plate and tension nail groups. The mechanical test was performed using an Instron E10000 testing machine.  
RESULTS AND CONCLUSION: (1) Simulation: In the normal pelvic model, the average displacement of the sacrum was 0.174 mm, and the maximum stress of the sacral iliac bone was 10.51 MPa, and the stress distribution was uniform. The mean sacral displacement of the unilateral injured pelvis model was 0.267 mm, and the stress concentration of the model was obvious. The mean displacement of the sacrum in the three repaired models was close to that in the normal pelvic model, and the stress distribution of the sacral iliac bone in the tension nail repaired model was uniform. (2) Mechanical test: The stiffness of the normal pelvic model was (226.38±4.18) N/mm, and that of the unilateral unstable pelvic model was the smallest (130.02±2.19) N/mm. The deviation of the normal pelvic model stiffness and the three repaired models’ stiffness were all within (±10%), and the repair effect was obvious. (3) The simulation results were in agreement with the experimental results. (4) The biomechanics of the tension nail repaired model was the most similar to that of the normal pelvis, and this method was the best. The repairing stiffness of the anterior double plate was too large, and the stress shielding effect was more significant. The posterior bridging plate repair could not solve the compensatory effect of the normal side soft tissue and had defects. This study provides an optimal basis for clinical surgery. (5) The new type of pelvic tension nail should be improved from the point of view of the tension nail to retain the good biomechanical properties of the tension nail, while adding other advantages, such as being used for the osteoporotic pelvis.

Key words: unstable pelvic injury, internal implant, tension nail, finite element analysis, biomechanics