In this work, we discuss the numerical challenges involved in the computation of the complex eigenvalues of damped multi-flexible-body problems. Aiming at the highest generality, the candidate method must be able to deal with arbitrary rigid body modes (free–free mechanisms), arbitrary algebraic constraints, and must be able to exploit the sparsity pattern of Jacobians of large systems. We propose a custom implementation of the Krylov–Schur method, proving its robustness and its accuracy in a variety of different complex test cases.

While an accurate computation of the tangent stiffness matrix of multibody systems is usually of critical importance in various numerical analyses, one contribution often neglected in dynamics is represented by the tangent stiffness matrix of constraints . This term, resulting from the change of joints reactions with respect to the coordinates of the connected bodies, is indeed easily discarded for the sake of performance due to its slight effect on dynamic simulation results. On the contrary, it plays a critical role when static or eigenvalue analyses are required, especially for those cases featuring free motions. While the topic of holonomic constraint equations and their respective Jacobian matrix is not new in literature, this article aims to provide a general and unified formulation based on quaternion parametrization together with a consistent analytical expression of the tangent stiffness matrix derived through linearization. The article includes also the rheonomic contribution to holonomic constraints. The formulations presented in this work are built on a mixed-basis formulation, in use in many engineering applications , and allow to easily derive specialized versions (e.g. revolute, cylindrical, prismatic joints, etc.) from the same equation set. Examples demonstrate the doubly-positive effect of this additional stiffness term: first, static analyses are able to converge to the actual equilibrium position and second, the eigenvalues analyses proved to be more consistent. For this latter set of tests the non-linear dynamic results, observed in the frequency domain, are compared against those coming from eigenvalue analysis, in order to prove the augmented accuracy of the results.

In this work we discuss the numerical challenges involved in the computation of the complex eigenvalues of damped multi-flexible-body problems. Aiming at the highest generality, the candidate method must be able to deal with arbitrary rigid body modes (free-free mechanisms), arbitrary algebraic constraints, and must be able to exploit the sparsity pattern of Jacobians of large systems. We propose a custom implementation of the Krylov-Schur method, proving its robustness and its accuracy in a variety of different complex test cases.