Research On Remote Interactive Visualization Technologies For CAE Data

With complexity of CAE analysis and continuous improvement of computational accuracy, the scale of computational data was getting larger. The large-scale data visualization need high-performance software and hardware. Computational data visualization can construct public service platform serving in various areas. The platform integrated various visualization tools and high-performance computational hardware, providing various data visualization services for users by online and offline. In this way users played out technological advantages of professional institutions to solve post-processing visualization problem in current large-scale computational data visualization.FEA data and data acquired by other method can be regarded as discrete data if lack of topological information. Discrete data not only had better adaptability, but also can be processed according to different resolutions. So it was used widely in large-scale computational data visualization.Based on the research of computational data remote visualization system and post-processing visualization technologies, remote interactive visualization platform based on shared interfaces was constructed in the way of video&command transmission and a series of visualization algorithms were proposed, which provided an effective way for high-performance data visualization.Firstly, research on remotely real-time interactive visualization technologies. Based on remotely interactive visualization technology which realized shared interfaces by video&command transmission, a tool for remotely interactive visualization operation was provided for users in clients, which provided a mapping window for visualization software running in server and in this window uses can operate computational data. Based on multi-task management and scheduling technologies, platform can support real-time visualization operation for multi-user. Based on servlet and socket, communications between platform client and management module were realized.Secondly, because current visual technology mainly faced to FEA data and involved less in discrete data, a series of visualization algorithms for discrete data were proposed. First, discrete data space was dividing by binary tree and the underlying nodes of binary tree were coding. Then discrete points were distributed to the outermost layer bounding box according to their coordinates. In this way, the neighbor domain structure of discrete data space was constructed. Based on discrete data preprocessing, a series of visualization algorithms were proposed, including streamline searching, gradient curve searching, nephogram drawing and isosurface extraction, and the efficiency of these algorithms were validated using aerodynamic performance analysis of the airflow around aircrafts as an example.Finally, according to platform target and functional requirement, the structure and work processing of platform were established. And on this basis computational data remote visualization platform software was developed and the functions of this software were validated.