Feature Tracking And Visual Analysis Of Unsteady Flow Field

For complex numerical simulation of unsteady flow fields,visualization is an effec-tive way for scientists to understand the physical properties of flow fields and improve experiments and simulations.In traditional visualization methods,numerical simula-tions of all time steps need to be stored on disk and then read from disk during subse-quent visual calculations.With the improvement of numerical simulation accuracy,data access overhead will lead to the failure of real-time storage of flow field data,slowing down the development cycle.In addition,due to the complexity of the flow field,the adoption of a direct visualization will lead to the problem that the results are difficult to analyze.Considering that when scientists analyze the flow field,they mainly focus on the regions of interest,including the time series feature regions with large changes in the flow field and the spatial feature regions with representative significance.There-fore,timing and spatial feature extraction and tracking of flow field data is an effective way to solve the above problems.Firstly,for the data storage problem,parallel rendering is carried out at the same time of numerical simulation,and the volume data is output in the form of visualiza-tion.With the visualization result,scientists are assisted to understand the numerical simulation situation in real time,and the storage of time series feature regions is an ef-fective data reduction method.However,when the data scale increases,due to the huge communication overhead during the image synthesis phase,the overall performance of traditional parallel rendering will reduce.Therefore,a parallel rendering method with-out composition was designed in this paper,which avoids the image synthesis stage in the traditional parallel rendering method and achieves efficient visualization of flow field data under the condition of increasing data scale.In order to further analyze the time series feature regions of flow field,due to the complexity of the flow field,directly visualization will make it difficult to analyse due to the complexity of the data.Therefore,this paper proposed a flow field visualization method based on spatial features.Firstly,Q criterion is used to extract the spatial features of the flow field.Subsequently,this paper designed a feature tracking method based on graph optimization.This method uses a weighted bipartite Graph to model the matching relationship of the feature regions of adjacent time steps,which has better applicability and accuracy compared with the traditional feature tracking method.Based on feature extraction and tracking results,the evolution of flow field feature regions over time can be clearly displayed.The above feature extraction and tracking of flow field data focus on global data statistics,which cannot directly reflect individual feature attributes and evolution of flow field.Aiming at this problem,a visual analysis system for unsteady flow field UFFVis was designed in this paper,which can display morphological changes of feature regions during the evolution,and explore the evolution of the feature regions through a vari-ety of interactive means.Finally,two case studies are organized to demonstrate the effectiveness of the system in exploring the evolution of flow field.