Study On Flow Field Characteristics Of Tandem Double Cylinders Based On Mode Decomposition Methods

Experimental and numerical methods are currently common methods to solve the flow field system.The experiment has the advantages of strong intuition and higher reliability,but the research cost and the implementation conditions are high;Although the research cost of numerical simulation is low,its calculation time is long,The data of numerical simulation is jumbled,and the processing efficiency is low.Considering the advantages and disadvantages of experiment and numerical simulation,many experts focus on data-driven as a new technical support to deal with complex flow field.Based on data-driven background,this study focuses on tandem flow fields of two cylinders,and solves the flow field through CFD.With the help of data driven technology,three kinds of reduction methods are used to deal with the results of CFD calculation to find the internal information extraction and vortex shedding characteristic of the object flow field.The specific research contents are listde as follows:The numerical model of flow field around two tandem cylinders is established.The corresponding conditions are formulated for different spacing between upstream and downstream cylinders and different Reynolds numbers.The numerical calculation of each condition is carried out and the effectiveness of the calculation results is verified.By collecting the numerical results of each time step,the flow field snapshot is generated,and the data set used in the algorithm training is provided.This paper introduces the theoretical basis of the three modal decomposition algorithms,introduces the simulated flow field to carry out identification analysis,and deeply analyzes the use of the three reduced order algorithms.The flow field around a tandem double cylinder with Reynolds number Re = 200 and spacing L / D= 4 under typical conditions is studied.The object flow field is divided into three stages: unstable equilibrium stage,transition stage and stable limit cycle stage.The order reduction of three stages is carried out by using POD and DMD,respectively.The vortex structure and stability of the three stages are compared and analyzed.Also,this paper analyzes flow field characteristics,such as the proportion of modal energy and stability.The sparse identification algorithm combined with the low dimensional Galerkin method is used to analyze the system,and the quadratic nonlinear equations are obtained to describe the stability of the system.Three order reduction algorithms are used to analyze the influence of two factors,i.e.the distance between the upper and lower cylinders and the Reynolds number of the incoming flow,on the flow characteristics of the object.Based on POD and DMD,the energy proportion and coherent structure of each mode are analyzed and compared,Also,the velocity,pressure and vorticity characteristics of the flow field at a certain time are reconstructed.The stability of the modes of each order is analyzed by DMD algorithm.The low order nonlinear equation describing the system characteristics is obtained by using SINDy algorithm.The applicability of each working condition in SINDy algorithm is analyzed based on the SNR of the system.The applicability of three modes decomposition methods to deal with the characteristics of vortex shedding mode and control stability is analyzed and contrasted.