A 3-dimensional computerized model of a human mandible with posterior teeth was generated. The bilateral sagittal split ramus osteotomy was virtually performed on this model. The separated model was assembled with 9 fixation methods: single screw, 2 screws one behind the other, 2 screws one below the other, 3 screws in an L configuration, 3 screws in an inverted backward L configuration, miniplate with 2 screws, miniplate with 4 screws, 2 parallel plates (upper + lower border), and square miniplate with 4 screws. Then, 75-, 135-, and 600-N vertical loads were applied on the posterior teeth of these models. The stress distribution on the screw sites on the buccal cortex was measured by the finite element method.
In this model all the fixation methods withstood forces between 75 and 135 N. However, the single-screw and the 2-hole miniplate models showed that the stress distributions in the configurations were intolerable when 600 N of posterior force was applied. The results of this study indicated that the inverted backward L configuration with 3 bicortical screws was the most stable.
Although this study indicated that the inverted backward L configuration with 3 bicortical screws was the most stable pattern, most of the patterns had adequate stability for clinical applications (mean, 125 N).