Start Date

2018 5:40 PM

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.

Abstract

The advancement of computational power has afforded the resolving of more complex flows in three dimensional space. Several numerical codes have been developed and to single out which method to use is becoming a more complicated task. In this study we present two computational fluid dynamic (CFD) platforms, FLOW-3D and Xflow on 3D turbulent flows in a fish-bone type fishway. One of the basic yet most crucial step in a number of CFD codes is the creation of efficient computational meshes for real world applications. Regardless of the numerical method applied, the mesh quality will have pronounced effect on the final results. Owing to these effects, mesh based (FLOW-3D) and meshless (Xflow) CFD codes are evaluated. FLOW-3D solves the Reynold’s averaged Navier‑Stokes equations and the continuity equation with the aid of Fractional Area‑Volume Obstacle Representation (FAVOR), while Xflow is particle based fully Lagrangian based on the Lattice-Boltzmann method solver. For both codes, turbulence is treated using the large eddy simulation (LES) model. Numerical model accuracy is assessed comparing the complex flow field in the fishway with fish migration behaviour of fish‑swimming species in a laboratory experiment. In addition, computational times and the performance of the models at different discharge rates (0.016 and 0.075 m3/s), classified as low and medium, respectively, is assessed. Generally, the findings indicated that Flow-3D better estimated the flow structure, velocity distribution and flow depth compared to Xflow, while Xflow was relatively faster in terms of computational time despite compromised accuracies.

Share

COinS
 
May 16th, 5:40 PM

Performance Assessment of FLOW-3D and XFlow in the Numerical Modelling of Fish-bone Type Fishway Hydraulics

The advancement of computational power has afforded the resolving of more complex flows in three dimensional space. Several numerical codes have been developed and to single out which method to use is becoming a more complicated task. In this study we present two computational fluid dynamic (CFD) platforms, FLOW-3D and Xflow on 3D turbulent flows in a fish-bone type fishway. One of the basic yet most crucial step in a number of CFD codes is the creation of efficient computational meshes for real world applications. Regardless of the numerical method applied, the mesh quality will have pronounced effect on the final results. Owing to these effects, mesh based (FLOW-3D) and meshless (Xflow) CFD codes are evaluated. FLOW-3D solves the Reynold’s averaged Navier‑Stokes equations and the continuity equation with the aid of Fractional Area‑Volume Obstacle Representation (FAVOR), while Xflow is particle based fully Lagrangian based on the Lattice-Boltzmann method solver. For both codes, turbulence is treated using the large eddy simulation (LES) model. Numerical model accuracy is assessed comparing the complex flow field in the fishway with fish migration behaviour of fish‑swimming species in a laboratory experiment. In addition, computational times and the performance of the models at different discharge rates (0.016 and 0.075 m3/s), classified as low and medium, respectively, is assessed. Generally, the findings indicated that Flow-3D better estimated the flow structure, velocity distribution and flow depth compared to Xflow, while Xflow was relatively faster in terms of computational time despite compromised accuracies.