Flexible Multibody System Dynamics And Its Design Sensitivity Analysis

In this paper, the flexible multibody system (FMS) dynamics and its design sensitivity analysis are investigated based on the Isogeometric Analysis (IGA) and Absolute Nodal Coordinate Formulation (ANCF).IGA is a newly proposed finite element method (FEM). Different from traditional FEMs, IGA employs the same discription framework for geometric modeling and finite element analysis (FEA). It can obtain the exact FEA model by directly meshing the NURBS geometry, which is hard to be done by traditional FEMs. To extend IGA into FMS, a Green-Lagrange strain tensor based method is proposed to solve the large rotation and deformation coupled problems. Furthermore, this thesis also proposes a continuum mechanics based IGA method to analyze FMS. All the components of the dynamic equations are deduced in detail and their numerical evaluation methods are reviewed. The feasibility of this formula-tion is verified by numerical example studies. The proposed approach has great importance in the extension of FMS modeling methods, improvements of the pre and post processing efficiency and the numerical calculation precision.ANCF is another new FEM, whose most distinguished feature is the use of global slopes as nodal variables. In this thesis, ANCF is extended into FMS, especially in large rotation and deformation problems. The results are verified by numerical experiments and a comparison between ANCF and IGA is made. Differences between these two methods are featured out, which is valuable for the selection of appropriate geometric modeling method in different cases for FEA.Design sensitivity analysis of FMS is investigated based on dynamic analysis of FMS. Finite difference method, direct differentiation method and adjoint variable method are systematically studied and their characteristics are analysed. An ANCF and continuum mechanics based method is proposed to assemble the system sensitivity equations, which avoids the sophisticated symbolic differentiation process and greatly improves the efficiency of construction and numerical integration of system sensitivity equations. Finite difference method and the revised direct differentiation and adjoint variable methods are compared through numerical examples. Results show that the revised direct differentiation method is better than the other two methods in the aspects of system equation construction, numerical integration and accuracy.All these improvements greatly enhance the application value of the direct differentiation method in the engineering optimization of the ANCF-based FMS.In addition, the system dynamic equations and numerical integration methods are researched in detail. Six kinds of implicit integration methods are compared from the aspects of global convergence, energy preservation, kinematic constraint drift and numerical efficiency, which is valuable for the choice of integration methods in different circumstances.