| Aircraft structure contains a large number of flexible parts. The stress deformation of flexible parts in assembly process is a key issue of the dimensional quality control in aircraft manufacturing. From the viewpoint of error sources, the shape error induced by part forming is a main contributor to the assembly variation, and the shape error of part is not changed randomly or independently, instead, exhibits certain correlation and regularity. Meanwhile, both clamping force required for shape closure in assembly and spring back deformation of assembly are influenced by the material error of parts. The dimensional variation in flexible assembly is the result of multiple error resources coupling with geometric errors and material errors of parts, and it is challenging to model the dimensional variation in flexible assembly. Up to now, the relevant researches mainly focus on the effect of geometric errors on the assembly variation, without consideration of the influence of material errors. With the support of National Natural Science Foundation of China (51275236), beginning with the piecewise linear elastic mechanics model of flexible assembly, this thesis investigates the mechanism and effect of shape and geometrical errors on the assembly variation. Furthermore, a methodology of dimensional variation analysis in flexible assembly is proposed.The main research contents and results are as follows:A three-dimensional geometric covariance model of part surface errors is established. On frequency scale, shape errors of part are firstly divided into two main types, i.e. warping and waviness. Subsequently, Legendre polynomials and Sinusoidal polynomials are respectively applied to fit the warping and waviness. Based on these mixed polynomials, the expression of geometric shape error of part is deduced. Furthermore, the covariance model is established, which could characterize the geometric correlation and regularity of shape error accurately.A piecewise linear elastic mechanics model of riveting process is proposed. According to the characteristics of riveting process in aircraft assembly, the riveting process is decomposed into four stages including placing, clamping, fastening and releasing. Combined with the linear elastic, small deformation assumption and the FEA method, the relationship is analyzed between the assembly forces and the geometric errors of key characteristics involving clamping points, joining points and measuring points for each assembly stage, and the mechanics model for dimensional variation analysis in flexible assemble is established. An aluminum alloy roll parts assembly experiment is designed to verify the effectiveness of piecewise linear elastic mechanics analysis model of flexible parts assembling. Finally, an assembly experiment with proportional specimens of skins is conducted to verify the proposed model.A methodology is developed for statistical variation analysis in flexible assembly considering both geometrical errors and material errors. Based on the piecewise linear elastic mechanics model of flexible assembly, a sensitivity matrix is deduced to depict the relationship between geometric errors of parts and assembly variation. With the perturbation theory, the sensitivity matrix is expanded by a first order Taylor series. In this way, the material errors of elastic modulus and poisson’s ratio are taken into consideration in the stiffness analysis of flexible assembly system. Furthermore, combing the piecewise linear elastic mechanical model with statistical principle, an approach on the statistical analysis in flexible assembly is proposed, and the statistical expressions of assembly forces and assembly variation are deduced. Finally, combining with two assembly experiments, i.e. sheet metal assembly and aircraft horizontal tail leading edge assembly, the proposed approach is compared with the Monte-Carlo simulation method based on FEA model of assembly process, and the results shows the validity and high efficiency of the proposed method. |
PDF Full Text Request