Research On The Key Techniques Of The Parallel Numerical Simulation For Fish School

With the continuous progress of bionics research,fish swimming provides new ideas for the propulsion of underwater vehicles.Investigations on the mechanism of the fish swimming and its interaction with the flow field will have great guiding significance for developing the new concept underwater bionic robot.In recent years,numerical simulation has developed to an important scientific method,but numerical study on the fish swimming mechanics is still in its infancy.In this thesis,aiming at the poor usability,low efficiency and narrow applicability faced by fish swimming simulation,we studies in depth the key techniques of parallel numerical simulation for fish swimming.The main contents and innovations include:1.The parallel numerical simulation framework for fish swimming is designed and implemented.(Chapter 2)We abstract the parallel numerical simulation framework for fish swimming and design an open hierarchical architecture.Based on the open-source software platform Open FOAM,we design the fish-fluid coupled solver for the complex flow simulation.Using the dynamic library mechanism,we define the uniform interface for the description of fish surface boundary.Based on the idea of grid scanning,we implement the eddy feature extraction algorithm.To accelerate the generation of initial grids,we design the body shape pretreatment tool.These research provide dedicated modeling platform for domain users and help promote the research progress of fish swimming mechanism.2.The maximum error based multi-selection greedy algorithm is proposed and the greedy point selection based RBF mesh deformation method is implemented.(Chapter3)To improve the efficiency of the point selection procedure,a parallel multi-selection greedy method has been developed in this theses.Multiple points are selected at each step to accelerate the convergence speed of the greedy algorithm.In addition,two strategies are presented to determine the specific selecting number.The parallelization of the greedy point selection is realized based on a master-slave model,and a hybrid decomposition algorithm is proposed to address the load imbalance problem.Numerical benchmarks show that both our multi-selection method and the parallelization could obviously improve the point selection efficiency.3.The RBF method is coupled with local adaptive(LA)method for the first time and the RBF-LA hybrid mesh deformation is designed and implemented for simulation with large-displacement moving boundary.(Chapter 4)We introduce the topology optimization method into RBF mesh deformation and design the RBF-LA hybrid method.Based on the current edge length and predefined growth factor,a hybrid mesh length scale calculation method is proposed.With this method a local refinement algorithm is designed.To optimize the Delaunay edge swap procedure,a mesh quality testing method is implemented.These research improve the robustness of the RBF mesh deformation and provide technical support to solve the problem with large-displacement moving boundary.4.The Kármán gaiting model is employed for the first time to numerically investigate the mechanism of fish exploiting vortices.(Chapter 5)To understand the mechanism of fish exploiting vortices,we employ the Kármán gaiting model for the first time to numerically investigate the hydrodynamics of fish swimming in the vortex street.The Kármán gaiting efficiency is also defined to evaluate the ability of fish extracting energy from vortices.Based on simulations over a wide range of controlling parameters,we discuss in detail the efficiency of fish exploiting vortices,the force coefficient and the hydrodynamic interactions between fish and vortices.Our numerical results show essential supplements to experiments and give scientific explanation of fish exploiting vortices.