Abstract
The numerical framework of the open source CFD solver REEF3D is utilised to study the fluid–structure interaction of an open ocean aquaculture system and waves. The presence of the net is considered in the momentum equations of the fluid using a forcing term based on Lagrangian–Eulerian coupling and the hydrodynamic loads on the net. They are defined semi-empirically using a screen force model. Here, the hydrodynamic force coefficients are calculated from the net geometry and fluid velocity. The necessary force coefficients are predicted from a new simulation-based screen force model. Here, CFD simulations are performed to obtain the hydrodynamic loads on net panels for varying net geometries, angles of attack and velocities. Then, a Kriging metamodel is applied to fit a polynomial to the data. The proposed net model is validated against measurements for waves and current through rigid net panels and applied to simulate the dynamics of an open ocean aquaculture cage in waves. In current, the model predicts the drag forces and velocity reduction within a 10% error band, whereas it tends to under-predict the lift forces by up to 20%. In waves, the model tends to over-predict the crests with increasing wave height, but the deviations are also within a 10% error band.
Highlights
The Open Ocean Aquaculture (OOA) industry has seen strong growth over recent years due to an increasing demand for aquaculture products
Computational Fluid Dynamics (CFD) models can include the two-way coupling between the fluid and the structure
The validation process for current and wave cases shows that this approach is capable of predicting the forces on and velocity reduction behind rigid net panels of various geometry
Summary
The Open Ocean Aquaculture (OOA) industry has seen strong growth over recent years due to an increasing demand for aquaculture products. First CFD attempts were based on two-dimensional calculations and a porous medium approach to account for the effect of the net on the fluid [5–7] This kind of approximation is necessary due to the large length scale difference between the twines of a net and the size of the complete structure, which prevents the resolution of the net on the same numerical grid as the fluid domain. An incremental pressure-correction algorithm [20] is combined with a third-order accurate TVD Runge–Kutta scheme [21] to solve the system of equations explicitly For this purpose, a rectilinear grid is defined in the numerical domain, and the pressure and velocity fields are defined on the cell centres and faces, respectively. A ghost cell approach and the message passing interface (MPI) are implemented for the inter-processor communication
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
