Multi-timestep integration in computational dynamics.

From MaRDI portal





The article is devoted to construction and investigation of stability and robustness of algorithms for nonlinear dynamics. Subcycling in such algorithms can greatly reduce the computation needed to solve an impact problem. Robust algorithms need high-frequency energy dissipation and major timestep update dependent on all minor timestep states to ensure stability. Here partial velocity approach is used for subcycling an explicit generalized alpha method. The algorithm that holds velocity constant tends to be more robust. The central difference method is extended by partial velocity subcycling, and the need to store additional quantities can be avoided by accumulating impulses associated with different element timesteps. High-frequency damping has been effectively added with the use of constant velocity on a timestep interface. Two corrector cycles are used to remove the problem of statistical stability. The application of provided algorithms are mechanics of rigid and flexible multibody systems.NEWLINENEWLINEFor the entire collection see [Zbl 0972.00063].











This page was built for publication: Multi-timestep integration in computational dynamics.

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2768000)