Unsteady Incompressible Flow Solver On Moving Hybrid Grids And Numerical Study Of Flow Mechanism For Fish Swimming

As the most popular aquatic animals, all fishes have excellent swimming ability toadapt the aquatic environment. Their geometry and locomotion manner washypothesized to be optimized because of the evolution of billions of years. Thecomparative biomechanics and physiology of moving through water has long attractedthe attention of both biologists and engineers, and recent decades have witnessedconsiderable growth in the study of aquatic animal locomotion.The purpose of this paper is to develop a numerical method which is suitable forthe study of biology-fluid, including an efficient/robust Dynamic Hybrid MeshGeneration Technique and Unsteady Incompressible Flow Solver. Based on thenumerical method, the undulation motion of a three-dimensional fish, as well as theschooling of fishes, has been studied numerically. The fluid mechanism of thegeneration of “thrust” and the interaction between fishes were analyzed.There are totally eight chapters in this dissertation.The first chapter is the introduction. In this chapter, the research background wasintroduced briefly. Then the main progress in recent years was reviewed, including theunsteady flow mechanism, experimental techniques, and numerical methods. Finally,the work of this paper was discussed.The dynamic hybrid mesh generation method was presented in Chapter2. Firstly,the static hybrid mesh generation method for complex configurations was introduced.Secondly, several mesh deforming techniques for prismatic elements or tetrahedralelements, such as the spring relaxation and Delaunay background grid mapping algorithm, were specified. Then, the dynamic hybrid mesh generation method used inthis paper was developed with integrating the local re-meshing strategy to improve theapplicability for large morphing or displacement of complex geometries. Finally, thedynamic hybrid grids for some moving boundary cases were shown, which illustratesthe robustness and capability of the present method.The unsteady incompressible flow solver was developed in Chapter3. Theartificial compressibility method coupled with the well-known dual-time steppingalgorithm was adopted in this paper. To improve the convergence history in thesub-iterations, an efficient block Lower-upper Symmetric Gauss-Seidel (BLU-SGS)implicit method wad employed. In order to deal with large-scale problems includingmillions of cells, the parallelization strategies based on geometric domaindecomposition technique was developed also. Some typical steady/unsteadyincompressible test cases were simulated in Cheaper4, and compared withexperimental data and other numerical results. The numerical results demonstrated the accuracy and efficiency of the unsteady incompressible flow solver in this paper.In Chapter5, the undulating motion of a tunny-shaped model was simulated.Influence of Reynolds number, turbulence flow, as well as the caudal fin shapes, wasanalyzed. Numerical results show that the efficiency for tunny-shaped model cruisingin lower Reynolds number is relatively poor, illustrating that the tunny-shape is onlysuitable for larger ocean fishes when swimming with higher Reynolds number. Thecomparison between the laminar and the turbulence flows show that the turbulenceflow can reduce the flow separation behind the spindle-shaped fish body, which isuseful to reduce the drag of body and to increase the swimming efficiency. Thecomparison between several caudal fin models show that the standard crescent-shapedcaudal fin is the most efficient for cruising, because of lower lateral disturbance onwater and less power losing.In Chapter6and7, the fish schooling in two-dimensional and three-dimensionalcases was studied respectively. The inline, trianguler and diamond shaped arrayswere simulated. Numerical results of two-dimensional arrays show that the aft-fishmust keep its undulating phase angle same as the fore-fish in the in-line array.However, when the aft-fish lies between the fore-fishes, such as in the triangular or diamond arrays, the aft-fish need not keep its undulating phase angle, and it wouldget a better interaction at some fore-aft-distance. In three-dimensional cases, theconclusion is different from that of two-dimensional cases. The favorable interactionhas not been found in this work. The possible reason is the undulating manner adoptedin this paper. The interactions between fishes when schooling are very complicate,further study should be carried out in the future.Chapter8concluded the study in this paper, and the future work was discussedalso.