Research On Models Conversion And Its Application In MCNP Modeling

The thesis concentrates on researching & developing Computer Aided Modeling System of MCNP. It is a sub-project of "one hundred talents" project-"Digital Reactor". MCNP is a widely applied particle transport program around the world, but it is very difficult to construct its geometry model manually, so it is of great practical interest to research & develop the computer aided modeling system for it. Based on the survey of the related Computer Aided Modeling software, the thesis presents the research goal-developing a Computer Aided Modeling system based on CAD technology; and makes an intensive study on the related theory and algorithms. The thesis's main contents and novel points include:[1] By introducing the "Ditigal Reactor" system, the thesis points out that MCNP is an important partof the system, and shows the importance and necessity to research & develop Computer AidedModeling system for MCNP. Based on the intensive study on MCNP Geometry model and CADtechnology, the thesis concludes that the key problem of developing such computer-aided modingsystem is to research "BREP→Half-Space Representation" Conversion Algorithm.[2] Based on the intensive study on the related theory and previous work on "BREP→ CSG"Conversion, the thesis presents a new conversion algorithm: "BREP → Half Space Representation"Conversion based on Decomposition. The concept of "dominant surface" is presented; and it is usedto optimize the conversion algorithm.[3] "Decomposition surface selection algorithm based on heuristics" is presented in the thesis, and it isused to to optimize the conversion algorithm, too.[4] Based on the intensive study on 3D Feature Recognition technology, the thesis presents a new concept of "Conversion Feature", so that 3D Feature Recognition technology is introduced into Representation Conversion Algorithm for conversion algorithm optimization.[5] Based on "BREP→Half-Space Representation" Conversion, four "Conversion Features" are defined in the thesis. The "conversion features" are Singe Loop Protrusion Feature, Singe Loop Depression Feature, Double Loops Protrusion Feature, and Double Loops Depression Feature. The thesis also studies how to utilize these "conversion features" to optimize "BREP→Half-Space Representation" conversion algorithm.[6] The concept of EAAG (Extended Attributed Adjacent Graphic) is presented in the thesis; and the algorithm of "Conversion Feature" Recognition based on EAAG is also presented. The conversion algorithm is improved by utilizing "Conversion Feature"; and the new algorithm is called improved"BREP→Half-Space Representation" conversion algorithm.