The objective of this paper is to improve the performance estimation model of the internal flow field of a torque converter. Compared with performance experiment results, the converter based on the one-dimensional model does not satisfy the performance requirements demanded in practice. Therefore, we need to develop more predictable and reliable performance estimation models. In order to obtain shape information on three-dimensional blade geometry, a process of reverse engineering conducts a torque converter assembly, impeller, turbine and stator. In addition, a CFD simulation including mesh generation and post-processing was carried out to extract equivalent parameters from the internal flow field. The internal flow field can be explained by analyze the correlation between a performance estimation model and CFD analysis. The equivalent performance model adopts the variation of energy loss coefficients for a given operating condition according to the application of a changing energy loss coefficient by the least mean squares method. The estimated equivalent model improves the agreement in performance between experiments and the theoretical model. This model can reduce the error to within about 3 percent. Furthermore, this procedure for predicted performance achieves eminence in the estimation of the capacity factor.