AAU Student Projects - visit Aalborg University's student projects portal
A master's thesis from Aalborg University
Book cover


Implementation and comparison of two numerical models for trawl cod-ends

Author

Term

4. term

Publication year

2018

Submitted on

Pages

90

Abstract

Den bageste del af et trawl, posen, samler fangsten under bugsering. Dens form bestemmer maskeåbningen og påvirker dermed selektiviteten, altså hvilke fisk der tilbageholdes. Projektet undersøger pose-deformation med to numeriske modeller: en aksialsymmetrisk model (forudsætter rotationssymmetri) og en 3D finit‑element‑model, der opdeler nettet i mange små trekantede elementer. Begge modeller beskrives og implementeres i C# med principper fra objektorienteret programmering. Vi afprøver forskellige løsninger til at forbedre konvergensen af Newton–Raphson‑metoden, som typisk bruges til at beregne posens ligevægtsform. De numeriske resultater fra de to modeller sammenlignes, og der udføres også et forsøg i en strømningskanal (flume tank) med et bevægelsessporingssystem for at måle posens 3D‑form. Resultaterne peger på god overensstemmelse mellem modellerne for småskalaposer med tynde nettråde.

The back of a trawl net, the cod‑end, collects the catch while the net is towed. Its shape controls how wide the meshes open and therefore which fish are retained (selectivity). This project studies cod‑end deformation using two numerical models: an axisymmetric model that assumes rotational symmetry, and a 3D finite element model that divides the net into many small triangular elements. Both models are described and implemented in C# using object‑oriented programming. We test several ways to improve convergence of the Newton–Raphson method, which is commonly used to compute the equilibrium shape. We compare numerical results from the two models, and we also run a flume tank experiment with a motion‑tracking system to capture cod‑end shape in 3D. The results indicate good agreement between the models for small‑scale cod‑ends made with thin net twines.

[This abstract was generated with the help of AI]