Sea Loads on Closed Flexible Fish Cages

• Main Author: Strand, Ida M.
• Other Authors: Sørensen, Asgeir J., Lader, Pål Furuset, Faltinsen, Odd Magnus
• Format: Doctoral or Postdoctoral Thesis
• Language: English
• Published: NTNU 2018
• Subjects: VDP::Technology: 500::Marine technology: 580; Norway; Atlantic salmon
• Online Access: http://hdl.handle.net/11250/2492628

Norway has become the world's largest producer of Atlantic salmon through the use of open net structures in the sea. The aquaculture facilities have grown in both size and number. Currently, the industry faces increased attention on environmental challenges related to fish escapes, sea lice, diseases, and pollution. A possible solution is to use a closed flexible fish cage (CFFC) with impermeable membrane material instead of nets used in conventional aquaculture cages. Compared to a net-based structure the response to sea loads of the new membrane-based system changes completely. Few ocean structures exist with large, compliant submerged components, and even fewer with a free surface. It is therefore presently limited existing knowledge about how CFFCs will respond to sea loads. It is therefore a need for the development of fundamental knowledge and understandings of the physics of the CFFC. From this understanding we will develop mathematical models of the sealoads and the response of the CFFC. Model experiments of scaled models of the CFFC in current for different filling levels have been conducted in multiple rounds to build knowledge of the system. In current, the response of the CFFC has been found to be highly dependent on the filling-level. For underfilled CFFC, a large deformation shaped like a hemispherical cup, comparable to a parachute, appear at the front facing the current. This deformation lead to a change in the geometry resulting in an increase in the drag force related to a full CFFC. From the experiments, it was observed that the problem of a CFFC in current can be characterised by a complex interaction between the membrane, the fluid masses within the CFFC and the outside fluid flow. In order to reduce the complexity of the problem, it was decided to model the system in 2D. In addition, it was chosen to shift the focus from current to waves allowing for the use of potential flow theory in the load modelling. To develop theory and understanding of the membrane structure, and the coupling between structural response and internal water motions, the response of a 2D rectangular sloshing tank with a fabric membrane side wall subject to forced sway motion was analysed. It was found that the response and the eigenfrequencies of the coupled system relied heavily on both the membrane structure through the tension, the membrane length, and the hydrodynamic pressure. A mathematical model of a 2D CFFC in waves was then developed. In order to analyse the CFFC in waves, equations for the geometry and initial tension of the membrane of the CFFC were found. Based on the found geometry and static tension, the wave response of a 2D CFFC in sway, heave and roll was studied. It was found that the rigid body wave induced motion responses of a CFFC with membrane in sway, heave and roll are significantly different from the responses of a rigid CFFC. Very large ratios between free-surface elevation amplitudes and incident wave amplitude are predicted inside the tank at the first and third natural sloshing frequencies. It implies that nonlinear free surface effects must be accounted for inside the tank in realistic sea conditions, as well known from other marine sloshing applications. Within linear structural theory we require that the dynamic tension in the membrane of the CFFC is smaller than the static tension. For the analysed case with given dimensions, for significant wave heights larger than 0.5 meter, the most probable largest dynamic tension is larger than the static tension. For negative total tensions the structural model is not valid. Therefore, a higher order structural model should be used. The effect of scaling of elasticity on the rigid body motion have also been investigated. The response of the CFFC using an elasticity available in model scale have been compared to the response of the CFFC using the elasticity for full scale. These responses where found to deviate to a large extent. This raises severe questions of the direct use of results from model scale experiments for the CFFC.