Use of non-linear regression to evaluate drag force and volume coefficient of structure of square cage

Abstract

Cage aquaculture is increasing in importance in fisheries. However, in past studies the relationship between drag force and volume co efficient of the cage bag, and four key factors affecting its efficiency, i.e. flow velocity, ratio of depth to width of the cage, twine diameter to bar length of netting (d/l), and weight of the cage has been limited to partial regression and qualitative discussion. This paper establishes two non-linear regressions of the drag force and volume coefficient of a cage bag with seven flow velocities, five ratios of depth to width of the cage, four d/l, and five weights of the cage. The experimental data are processed using the forward selection method in the stepwise regression selection procedure in order to prioritize each factor, and then establish non-linear regressions of the drag force and volume coefficient using multiregression analysis. The results show that factors affecting the drag force are, in order of importance, flow velocity, ratio of depth to width of the cage, d/l and weight. With respect to the volume coefficient, factors in order of importance are flow velocity, weight, ratio of depth to width of the cage, and d/L. The two regressions can predict drag force and volume coefficient more comprehensively than those in previous studies. Finally, the regressions are applied to the full-scale structure of the cage by Tauti’s modeling rules.