On an accurate A-posteriori error estimator and adaptive time stepping for the implicit and explicit composite time integration algorithms

Yazhou Wang, Tong Zhang, Xuelin Zhang, Shengwei Mei, Ningning Xie, Xiaodai Xue, Kumar Tamma

Research output: Contribution to journalArticlepeer-review

7 Scopus citations

Abstract

This paper focuses on the design of an accurate and versatile A-posteriori error estimation and adaptive time stepping for general composite schemes and typical of the implicit ρ-Bathe and explicit Noh-Bathe composite methods for time integration in second-order transient systems. Two novel error estimators, Disp/P3 and Disp/P4, are newly proposed based on the generalized polynomials. They have generalized compact single-step representations and work for both implicit and explicit, dissipative and non-dissipative composite schemes. Validation tests of linear and nonlinear problems show that the estimated local error reaches excellent agreement with the exact local error, and the third-order convergence rates of the local error are obtained. Besides, the Disp/P4 is more accurate than Disp/P3 as the effectivity index is almost identical to unity when random time steps are taken into consideration, and hence the Disp/P4 is highly recommended to be merged to the adaptive time stepping procedure. Three examples encompassing the wave propagation, damped/undamped spring-pendulum, and nonlinear spring-mass system are solved by implicit and explicit composite methods via the adaptive time stepping. Numerical results show the advantage of adaptive time stepping based on the proposed error estimator in contrast to constant time stepping regarding the CPU cost.

Original languageEnglish (US)
Article number106789
JournalComputers and Structures
Volume266
DOIs
StatePublished - Jul 1 2022

Bibliographical note

Funding Information:
This work is supported by The Science and Technology Project of China Three Gorges Corporation (Grant No. 202103385) and The Major Science and Technology Project of Inner Mongolia Autonomous Region (Grant No. 2021ZD0032). Acknowledgment is due to Professor Tamma’s computational mechanics research lab at the University of Minnesota.

Publisher Copyright:
© 2022 Elsevier Ltd

Keywords

  • Adaptive time stepping
  • Composite time integration
  • Error estimation
  • Second-order transient systems

Fingerprint

Dive into the research topics of 'On an accurate A-posteriori error estimator and adaptive time stepping for the implicit and explicit composite time integration algorithms'. Together they form a unique fingerprint.

Cite this