Rapid Visualization Method For Large-Scale CAD Model With Repeated Structure

Large-scale CAD model visualization is a problem that has not yet been solved completely.At present,there are many such CAD models which are consisted of large number of regularly arranged objects.For such CAD models,this dissertation research on fast visualization methods by taking use of their features,and the methods have been incorporated into self-developed software SuperMC/MCAM(Multi-physics Coupling Analysis Modeling program).The main work of this dissertation is as the following:1)Hybrid cross-section visualization method based on pixel threshold evaluation:All basic objects are converted into 3D texture during pre-processing stage;select the visual object by the visual clipping algorithm;and during the visualization stage,determine cross-section visualization method for each object by evaluating its projection on screen.The result demonstrates that the presented method can maintain visual accuracy,and at the same time it is fast enough for interactive interaction.2)3D visualization method based on reference grid model mapping:All the basic object is meshed as the reference grid,and then the other objects are visualized by the mapping transformation of the reference grid;before the visualization stage,select the visual object with the hierarchical clipping algorithm.This method can effectively meet the requirements of 3D visualization analysis of large-scale CAD model,and achieve the expected design goal.3)The presented algorithm has been researched on the SuperMC developed by the FDS team of Chinese Academy of Science.Both HM(Hoogenboom Martin,entities number is 6 362 400)reactor core model and ADS(Accelerator Driven Sub-critical System,entities number is 1114 384)reactor model have been used for testing,and the test results have been verified the reliability of the proposed method.