Integration of disassembly operations during product design is an important issue today. As known, the number of possible disassembly sequences increases significantly with the number of parts in a product. Thus, the generation of proper disassembly sequences order is critical. Most of existing methods often require tremendous computational resources while, at the same time, they often fail to find realistic and optimal solutions for complex products disassembly.
Disassembly operations cover a broad range of the Product Life Cycle (PLC) regarding operations of disassembly during: product design, production process, product maintenance and finally at the end of PLC. It is estimated that at the earliest stages of product design, the cost of these operations is at almost 30% of its total cost. Modelling these operations requires a lot of geometrical, kinematical, technological and ergonomical data and their synthesis in order to reduce the algorithmic complexity of the disassembly simulation process. Nowadays, disassembly operation simulation of industrial products finds a strong interest in interactive simulations through immersive and real-time schemes.
In this context, in the first time, the aim of this thesis is to develop a new method for generating selective disassembly sequences. When disassembling, it is important to eliminate the components which are unrelated to the target components prior to sequence generation. In order to address this configuration, this thesis presents a method for generating the feasible disassembly sequences for selective disassembly. The method is based on the lowest levels of a disassembly product graph. Instead of considering the geometric constraints for each pair of components, the proposed method considers the geometric contact and collision relationships among the components in order to generate the so-called Disassembly Geometry Contacting Graph (DGCG). The latter is then used for disassembly sequence generation thus allowing the number of possible sequences to be reduced by ignoring any components which are unrelated to the target. The method is applied for automatic generation of selective disassembly sequences for mechanisms with different degrees of complexity. The disassembly simulations can be performed either from an automated or interactive point of view using standard computer equipment or through immersive and real-time simulation schemes. In order to address this diversity of configurations, a simulation framework was developed integrated in a Virtual reality environment thus allowing generating the minimum number of possible disassembly sequences.
The available disassembly evaluation methods today seldom make disassembly as the preferred end-of-life solution for the reuse of parts or components in an economically sustainable way for lower value products. In recent years Virtual Reality interfaces have been wildly used to simulate various processes and in particular assembly/disassembly operations during the initial stage of product design. In the second time, a method for disassembly operation evaluation by 3D geometric removability analysis in a Virtual environment is proposed. It is based on seven new criteria which are: visibility of a part, disassembly angles, number of tools’ changes, path orientation changing, sub-assembly stability, neck score and bending score. All criteria are presented by dimensionless coefficients automatically calculated thus allowing evaluating disassembly sequences complexity. For this purpose, a mixed virtual reality disassembly environment (VRDE) is developed based on Python programming language, utilizing VTK (Visualization Toolkit) and ODE (Open Dynamics Engine) libraries. The framework is based on STEP, WRL and STL exchange formats. The analysis results and findings demonstrate the feasibility of the proposed approach thus providing significant assistance for the evaluation of disassembly sequences during Product Development Process (PDP).
Further consequences of the present work consist in ranking the criteria according to their importance. For this purpose, moderation coefficients may be allocated to each of them thus allowing a more comprehensive evaluating method.
mise à jour le 4 septembre 2014