Assembly accuracy prediction for over-constrained interference fits by coupling surface error and part deformation

Over-constrained interference fits are widely used in high-precision mechanical products such as aerospace structures, precision machine tools, etc. Its assembly accuracy is critical for the product performance, which is determined by the coupled effects of interference, manufacturing deviations, and over-constraint, and is highly challenging to achieve accurate prediction. Thus, this study proposes a novel method for assembly accuracy prediction of over-constrained interference fits, accounting for coupling effects of surface error and part deformation. First, non-uniform rational B-splines interpolation is employed to reconstruct the non-ideal mating surfaces, which are then fused with the ideal computer-aided design (CAD) geometry to generate a realistic CAD model of the over-constrained interference-fit parts. Second, based on this model, a surface-constrained registration approach is proposed to determine the initial mating state of the assembly. Finally, considering the structural complexity of the CAD model, a U-spline-based isogeometric analysis model is constructed to perform deformation analysis, and the results are extracted to enable high-precision prediction of assembly accuracy. Validation is conducted by practical experiments using representative over-constrained structures, showing agreement between predictions and experiments. These results strongly demonstrate the effectiveness of the proposed method for accurate assembly accuracy prediction for over-constrained interference fit, considering the coupling effects of surface error and part deformation.