Research On Sheet Metal Digitized Unfolding And Optimal Nesting Technologies With Process Constraints

Sheet metal processing is an important part of the mechanical manufacturing and sheet metal parts are widely used in various industry branches such as the aerospace industry, shipbuilding industry, automotive industry and so on. Take the aerospace industry for example, sheet metal parts constitute the main part of a modern aircraft body and account for over 50% of the whole aircraft parts. As the aircraft sheet metal component is usually manufactured with a complex structure together with a high geometry accuracy and it always takes the mode of multi-varieties and small-batch production, its manufacturing time occupies 15% of the total working hour and it directly affects the whole production process of an aircraft. In the whole manufacturing process, sheet metal unfolding and nesting are two important working procedures. The widely used aircraft parts such as the ribs of airfoil or bulkheads of airframe make the conventional graphic solution or the analytical method no longer suitable for the modern aircraft design. Meanwhile, the production mode of multi-varieties and small-batch in aircraft area needs an optimal nesting technology to enhance the economy of cutting process. Based on the above characteristics and the present situation of the aircraft sheet metal manufacturing, we will choose some typical parts in this area as the research object. With the goal of enhancing the algorithm‘s generality, efficiency and automation, we will make a research on the digitized unfolding and optimal nesting technologies for the aerospace and shipbuilding areas. The result of this dissertation will lay a solid foundation for the digital manufacturing in the above areas.One of the prominent features of the aircraft sheet metal components is that it is of a wide variety, which means various unfolding methods need to be developed to handle different types of parts. In order to achieve a good generality, we choose the triangulated surface as the unfolding model in this thesis. We evaluate the mesh surface developability by using the Gaussian curvature and divide them into developable surface and complex surface. As for those surfaces with a good developability, we adopt the energy-based flattening method to unfold them. To overcome the low flattening efficiency of the traditional energy-based method, we develop an innovative energy relaxation process based on variable steps. According to its special flattening mechanism, we adopt a hierarchy flattening method to enhance its robustness. The later process treating can make amendments to the automated flattened result, which further enhances the flattening accuracy. We choose the typical ribs and frame parts in aircraft industry as the test model and the selected parts with different types of flanges are unfolded for verification. The experimental result shows that there are almost no area and length distortions after the unfolding process when the developability value of the chosen part is between 0.9985 and 1. Compared with the existing methods, our method can avoid a great deal of manual operations such as feature selection, cross-section generation and so on, which largely enhances the efficiency, generality and automation of the unfolding method.As for the flat pattern development of sheet metal components with complex surfaces, all the existing methods have their own application limitations since they are greatly influenced by the geometric shapes of deformed surfaces. Therefore, we propose a newly emerging mesh parameterization approach named As-Rigid-As-Possible method to calculate the blank shape of the complex sheet metal parts. After the introduction of the general term of the deformation energy, we make a comprehensive study on the crucial steps such as the calculation of the local rigid matrix, the minimization of the global deformation energy and so on. Numerous experiments verify that the ARAP method can effectively solve the quasi-vertical wall, long-shaped structure, notch problem and so on. The ARAP method can hardly be affected by the part geometry and it can steadily provide a reliable initial solution for a further blank shape optimization.As for the sheet metal nesting problem, we start from the rectangular packing and we choose the best-fit(BF) heuristic as our basic packing algorithm in this thesis. As for the possible ?towers‘ problem while implementing best-fit heuristic, we adopt the genetic algorithm to optimize the top packing result. Using the local optimization method can obtain a reasonable rectangular packing result in affordable time. On this basis, we choose the no-fit polygon as the intersection test tool. After adding the shrinking and manual filling operation, we make our fundamental algorithm also fit for the irregular packing problem, which improves its generality. After a full consideration of the main process constraints such as the plate fiber direction, cutting loose, edge-sharing cutting and so on, we transfer the above constraints into the corresponding nesting rules and integrate them into our already developed fundamental nesting algorithm. Our proposed nesting algorithm can handle the rectangular and irregular packing problems in aerospace and shipbuilding areas while taking the process constraints into consideration.Based on the above research results, we have developed a sheet metal digitized unfolding software for lightweight model(Surface Para) and an optimal nesting software driven by production plans respectively. By choosing some typical sheet metal components and nesting examples as test models, both of them are effectively validated in application.