Study and discussion on computational efficiency of ice-structure interaction by peridynamic

Zhang, Yuan and Zhang, Guoyang and Tao, Longbin and Wang, Chao and Ye, Liyu and Sun, Shuai and Han, Kang (2023) Study and discussion on computational efficiency of ice-structure interaction by peridynamic. Journal of Marine Science and Engineering, 11 (6). 1154. ISSN 2077-1312 (https://doi.org/10.3390/jmse11061154)

[thumbnail of Zhang-etal-JMSE-2023-Study-and-discussion-on-computational-efficiency-of-ice-structure-interaction]
Preview
 Text. Filename: Zhang_etal_JMSE_2023_Study_and_discussion_on_computational_efficiency_of_ice_structure_interaction.pdf
Final Published Version
License: Creative Commons Attribution 4.0 logo
Download (1MB)| Preview

Abstract

The peridynamic (PD) theory is based on nonlocal mechanics and employs particle discretization in its computational domain, making it advantageous for simulating cracks. Consequently, PD has been applied to simulate ice damage and ice–structure interaction under various conditions. However, the calculation efficiency of PD, similar to other meshless methods, is constrained by the number of particles and the inherent limitations of the method itself. These constraints hinder its potential for further development in the field of ice−structure interaction. This study aims to explore the computational efficiency of various methods that can be employed to improve the computational cost of PD in ice–structure interactions. Specifically, we analyze the computational efficiency of three different methods (the MPI parallelization, the updated link−list search method, and the particle−pair method) and their collaborative calculation efficiency to reduce simulation time. These methods are employed to calculate ice–ship interaction, and their coupled efficiency is studied. Furthermore, this study discusses the computation strategy to improve efficiency on using the PD method to calculate ice–structure interaction. The present work provides scholars who employ PD to calculate ice–structure interaction or ice damage with a referential discussion plan to achieve an efficient numerical computation process.

CORE (COnnecting REpositories)