An accurate assembly eccentricity prediction method for tenon and mortise based on multi-contact surface registration algorithm

Abstract

Tenon and mortise assembly in aero-engine turbine disks significantly influences performance and service life. Machining errors introduce geometric deviations on contact surfaces, leading to eccentricity and unbalance during high-speed rotation. However, predicting eccentricity under stable operation remains challenging due to the complex shapes of tenon and mortise. Current virtual assembly methods primarily address face-to-face contact, while tenon and mortise assembly involves multi-contact surfaces. This paper proposes an accurate eccentricity prediction method based on a multi-contact surface registration algorithm, implemented in three steps. First, a geometric distribution error (GDE) model is established using high-precision point cloud data from a Coordinate Measuring Machine (CMM). Then, a multi-contact surface registration algorithm is developed to simulate virtual assembly under stable operating conditions, and the assembly pose prediction model is established. Finally, the eccentricity of tenon and mortise is numerically calculated to establish an assembly eccentricity prediction model. Experimental validation on actual parts confirms the GDE model’s micron-level precision. Tests on aero-engine simulation components demonstrate strong agreement between predicted and theoretical eccentricity values. The proposed method, integrating GDE modeling and multi-contact surface registration, effectively predicts stable-operation eccentricity and provides practical guidance for real-world assembly processes.