**1. Introduction**

Coastal structures are typically employed to reduce wave energy so as to mitigate coastal hazards for protecting the local residence [1] and coastal species [2]. On designing the structure, not only providing strong protection for the shore but also involving environmentally-friendly consideration should be balanced. In recent years, submerged-type structures have been extensively considered as alternative choices [3,4] to enhance water exchange and retain natural coastal landscape for a recreational purpose. On the other hand, coastal structures may have permeable parts, which leads to attenuate additional wave energy through viscous dissipation within the porous media [5,6]. A typical permeable structure mostly consisted of rubble-mounted elements. However, permeable objects can be built using several impermeable parts with slots to vary the porosity, which is known as screen-type barriers [7,8]. The classic type of barrier feature is thin, rigid, vertical, perforated, and surface-piercing, which is beneficial to account for economic and environmental concerns.

As reviewed in Huang et al. [9], most available studies in the literature have focused on evaluating the hydraulic performance in terms of wave reflection (R), transmission (T), and dissipation (D) coefficients, where surface-piercing-type barriers received more attention than those of submerged-type ones. Wu and Hsiao [7] numerically investigated solitary waves over a submerged dual-slotted-barrier system using a well-validated wave model based on the Reynolds-Averaged Navier-Stokes equations (RANS) by providing a simple empirical formula for estimating RTD coefficients, where wave conditions and porosities of each barrier are considered as the primary parameters for the estimations. However,

the flow fields of wave interactions with slotted barriers were studied sparsely. Although several numerical studies have provided simulated flow fields around slotted barriers [8,10], flow separation is one of the complicated phenomena in fluid mechanics and may not be able to be resolved accurately using numerical models unless the model has been rigorously validated through detailed model-data comparisons [11]. Using the particle image velocimetry (PIV), Liu and Al-Banaa [12] studied non-breaking solitary waves runup on a vertical surface-piercing barrier, and Wu et al. [13] investigated breaking solitary waves over a submerged bottom-mounted barrier. However, the flow fields due to the interaction of solitary waves and a submerged slotted barrier were not understood.

In practical applications, the elements of slotted barrier can be installed either horizontally or vertically, and, thus, the problem to be solved results in two-dimensional and three-dimensional setup for horizontal and vertical slotted barriers, respectively. Thomson [14] stated that the orientations of slotted barriers had an influence on transmitted waves based on experimental observation, where the horizontal slotted barrier appeared to be more effective in reducing wave transmission. In addition, choosing the shape of slotted barriers is one of the factors affecting the hydraulic performance. Krishnakumar et al. [15] stated that the slotted barrier with sharp edge elements such as square, rectangle, and triangle result in lower wave transmission but higher wave reflection than those consisting of circular shape elements. Additionally, Huang et al. [9] indicated that the rectangular element of perforated barriers may help generate more energy dissipation due to flow separation around the sharp edge elements of slotted barriers. Therefore, based on statements mentioned in available literature, the horizontal slotted barrier with rectangular elements may be the optimized setup as effective coastal structures, which can be considered a two-dimensional (2D) problem.

In this study, the primary aim is to investigate and understand the flow fields of solitary waves interacting with a submerged slotted barrier experimentally and numerically. A new experiment is performed in a laboratory-scale wave flume to measure the free surface displacement time series, the ensemble-averaged flow velocities, and the turbulent kinetic energy. Numerical simulation is carried out based on the RANS equations for the mean flow fields and the non-linear *k*-<sup>ε</sup> turbulence closure model to approximate the Reynolds stresses [16,17]. Detailed flow fields are addressed based on laboratory observations. Model-data comparisons in terms of the free surface elevation time series, the mean velocities, and the turbulent kinetic energy are performed to examine the accuracy of the numerical model and point out the limitation of numerical simulations.

#### **2. Research Methods**
