Enhanced weakly-compressible MPS method for violent free-surface flows: Role of particle regularization techniques

TL;DR

This paper introduces an enhanced particle method for violent free-surface flows, improving particle regularization techniques to achieve more accurate, stable, and volume-conserving simulations of complex fluid behaviors.

Contribution

It proposes novel enhancements to pairwise collision and particle shifting techniques, including a dynamic PC method and a consistent PS algorithm, for better accuracy and stability in free-surface flow simulations.

Findings

Dynamic PC reduces pressure noise with low energy dissipation.

Consistent PS conserves volume during extreme deformations.

Enhanced methods improve accuracy and stability in benchmark tests.

Abstract

This paper develops a consistent particle method for capturing the highly non-linear behavior of violent free-surface flows, based on an Enhanced Weakly Compressible Moving Particle Semi-implicit (EWC-MPS) method. It pays special attention to the evaluation and improvement of two particle regularization techniques, namely, pairwise particle collision (PC) and particle shifting (PS). To improve the effectiveness of PC in removing noisy pressure field, and volume conservation issue of PS, we propose and evaluate several enhancements to these techniques, including a novel dynamic PC technique, and a consistent PS algorithm with new boundary treatments and additional terms (in the continuity and momentum equations). Besides, we introduce modified higher-order and anti-symmetric operators for the diffusive and shear force terms. Evaluation of the proposed developments for violent free-surface flow benchmark cases (2D dam-break, 3D water sloshing, and 3D dam-break with an obstacle) confirms an accurate prediction of the flow evolution and rigid body impact, as well as long-term stability of the simulations. The dynamic PC reduces pressure noises with low energy dissipation, and the consistent PS conserves the volume even for extreme deformations. Comparing the role of these new particle regularization techniques demonstrates the effectiveness of both in assuring the uniformity of the particle distribution and pressure fields; nevertheless, the implementation of PS is found to be more complex and time-consuming, mainly due to its need for free surface detection and boundary treatment with many tuning parameters.