Two model domains were developed using integrated 3D bathymetric surfaces for low- and high-relief sites (Pomeroy and others, 2022). Each domain was defined in a longshore uniform, two-dimensional vertical (2DV) coordinate system (x, z), with x pointing shoreward, z pointing upward, and the origin at the still water level. The initial mesh was 55 m long and 6.5 m high, with a uniform resolution of 0.25 m. The bathymetry surfaces were snapped to the model domain with snappyHexMesh tool using three levels of grid refinement (min. cell size of 0.125, 0.06, and 0.03 m) in the free surface region and the fine-scale areas of the model domains. The bed and atmosphere in the models were treated with zeroGradient and inletOutlet boundary conditions, respectively, with walls set as empty (non-computational) boundaries. At the model inlet, outlet, and along the bed, the fixedFluxPressure boundary condition was applied to the pressure (hydrostatic) field to adjust the pressure gradient so that the boundary flux matched the velocity boundary condition. Turbulence parameters used respective wall functions to model boundary layer effects near the bathymetry. Time-averaged values of the dimensionless wall distance z-plus ranged from 35 to 120, where z-plus between 30 and 300 defines the log-law layer where wall functions are applicable. To minimize numerical dissipation in the models, a second-order unbounded numerical scheme was used for gradients, second-order bounded central differencing schemes for divergence, and an unbounded second-order limited scheme was used for the Laplacian surface normal gradients. Wave boundary conditions were handled by the IHFOAM toolbox (Higuera and others, 2013). A series of models were developed for both the low- and high-relief domains to simulate hydrodynamics across each bathymetry. Four water depths (h = 1, 2, 3, 4 m), five significant wave heights (Hs = 0.4, 0.8, 1.2, 1.6, 2 m), and five peak wave periods (Tp = 4, 8, 12, 16, 20 s) were selected to span the range of recorded conditions. Combinations of conditions that were physically unreasonable (in other words, those with excessive wave steepness or those above the theoretical breaking limit) were eliminated, resulting in 70 cases per domain and thus 140 total cases. A summary of the model scenarios is provided in the accompanying CSV file. The scenario models were set up by varying the vertical position of the bathymetry to create domains with four different water depths. Different wave generation theories according to Le Mehaute (1967) for each combination of h, Hs, and Tp were applied at the inlet boundary to drive wave flows through the model domains. Each model was executed for a total time of 2Tp plus 10Tp to generate two wave cycles during spin-up followed by 10 wave cycles that were used for subsequent data analysis. To improve runtime efficiency, three models were run simultaneously on six distributed parallel processors on an 18-core Intel i9-10980XE CPU at 3.75 GHz with 64 GB of RAM. Execution times varied by model scenario and ranged from approximately 12 to 160 hours. The bed and atmosphere in the models were treated with zeroGradient and inletOutlet boundary conditions, respectively, with walls set as empty (non-computational) boundaries. At the model inlet, outlet, and along the bed, the fixedFluxPressure boundary condition was applied to the pressure (hydrostatic) field to adjust the pressure gradient so that the boundary flux matched the velocity boundary condition. Turbulence parameters used respective wall functions to model boundary layer effects near the bathymetry. Time-averaged values of the dimensionless wall distance z-plus ranged from 35 to 120, where z-plus is between 30 and 300, which defines the log-law layer where wall functions are applicable. To minimize numerical dissipation in the models, a second-order unbounded numerical scheme was used for gradients, second-order bounded central differencing schemes for divergence, and an unbounded second-order limited scheme was used for the Laplacian surface normal gradients. Wave boundary conditions were handled by the IHFOAM toolbox (Higuera and others, 2013). Each model file has the following file structure, where each folder corresponds to a different component of the OpenFOAM model. This file structure is a modified version taken from the IHFOAM toolbox found in the OpenFOAM tutorials folder. The 0 and 0.org folders contain the model boundary conditions. The 0 folder is a copy of the 0.org folder as a backup of the initial settings. The constant folder contains the model constants as well as the model mesh (in the polyMesh subdirectory). The system folder contains the model settings. For more details, we refer the reader to the OpenFOAM User guide (
https://cfd.direct/openfoam/user-guide/) and the IHFOAM wiki site (
https://openfoamwiki.net/index.php/Contrib/IHFOAM).
