An optimal family of controllable numerically dissipative time integration algorithms for structural dynamics

X. Zhou, K. K. Tamma, D. Sha

Research output: Contribution to journalConference articlepeer-review

1 Scopus citations

Abstract

Based on the generalized single-step representations within the scope of the classical linear multi-step (LMS) algorithms, an optimal family of controllable numerically dissipative time integration algorithms are developed which are fundamentally useful for structural dynamics computations. The optimality of the algorithms are in the sense of achieving optimal algorithmic properties in all aspects within the limit of the Dahlquist theorem which pertains to only a special class referred to as the so-called linear multi-step (LMS) methods with attention to: unconditionally stable, second-order accurate, zero-order displacement and velocity overshoot, minimal dissipation and dispersion with respect to the selected magnitude of the principal roots in the high-frequency limit. The comparisons of the optimal family of algorithms with the currently available controllable numerically dissipative algorithms are shown which demonstrate the superiority of most of the algorithmic properties. An illustrative elasto-plastic large deformation structural dynamic problem is also presented from an application viewpoint.

Original languageEnglish (US)
Pages (from-to)3050-3060
Number of pages11
JournalCollection of Technical Papers - AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics and Materials Conference
Volume5
StatePublished - Jan 1 2001
Event42nd AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conference and Exhibit Technical Papers - Seattle, WA, United States
Duration: Apr 16 2001Apr 19 2001

Keywords

  • Optimal algorithmic properties
  • Structural dynamics
  • Time integration algorithms

Fingerprint Dive into the research topics of 'An optimal family of controllable numerically dissipative time integration algorithms for structural dynamics'. Together they form a unique fingerprint.

Cite this