Porous cylinder arrays for optimal wake and drag characteristics

TL;DR

This study investigates how the arrangement of porous cylinder arrays, modeled after mangrove roots, affects wake and drag characteristics using experiments, simulations, and optimization to identify configurations that optimize flow properties.

Contribution

The paper introduces a combined experimental and numerical approach to optimize porous array configurations for desired wake and drag behaviors, highlighting the importance of geometry over simple porosity metrics.

Findings

Optimal arrangements significantly influence wake and drag characteristics.

Enstrophy is more sensitive to cylinder placement than drag.

High-drag arrays disrupt flow and reduce particle deposition.

Abstract

The root systems of mangroves, a tree species found in intertidal tropical and sub-tropical coastal zones, provide a natural barrier that dissipates wave energy effectively and reduces sediment erosion. In this work, we use a combination of experiments and numerical simulations to examine the wake and drag characteristics of porous arrays of cylinders, which serve as simplified models of mangrove root networks. Optimal arrangements of the arrays are obtained by coupling Navier-Stokes simulations with a multi-objective optimization algorithm, which seeks configurations that minimize wake enstrophy and maximize drag acting on the porous structure. These optimal configurations are investigated using Particle Image Velocimetry, and the internal and external flow around the porous arrays are analyzed using a combination of Proper Orthogonal Decomposition and Lagrangian particle tracking. Large variations in drag and enstrophy are observed by varying the relative positions of the cylinders, which indicates that the geometrical arrangement of porous arrays plays a prominent role in determining wake and drag characteristics. A sensitivity analysis suggests that enstrophy is more sensitive than drag to specific cylinder placement, and depends on distinctive flow patterns that develop in the interior due to interactions among individual cylinders. Arrays with higher drag are found to have higher average flow speeds in the wake and give rise to time-varying Lagrangian Coherent Structures, making them unfavorable for particle deposition and erosion. Overall, the results indicate that high-level metrics such as array porosity may not be sufficient on their own for predicting wake and drag characteristics.