3D model reconstruction of damaged parts in an automated remanufacturing process using additive manufacturing technology

• 主要作者: Aprilia
• 其他作者: -
• 格式: Thesis-Doctor of Philosophy
• 語言: English
• 出版: Nanyang Technological University 2021
• 主題: Engineering::Manufacturing::CAD/CAM systems; Engineering::Manufacturing::Metrology
• 在線閱讀: https://hdl.handle.net/10356/152266
• 機構: Nanyang Technological University

Automated remanufacturing process is of significant interest in many sectors of industry as current manual remanufacturing process is prone to human error, inconsistent in quality and costly to process. Current advancements in additive manufacturing and 3D scanning technology supports opportunities for realizing automated remanufacturing. However, to date, a fully automated remanufacturing process has not yet been achieved. Current limitations were identified to the absence of effective damage detection algorithm, nominal model reconstruction algorithm, and automated machine toolpath generation process. The existing shortfall in capabilities required time consuming human input in the assessment of damage(s) and in defining the appropriate remanufacturing toolpath profiles. For widespread acceptance of “automation of remanufacturing process”, all steps should be carried out automatically by software with no human interaction except for the final verification. Automated flow should be seamless with information obtained from scan data to the final machining processes. This thesis aims to work towards achieving a fully automated remanufacturing process by establishing and proposing an automated remanufacturing process framework exploiting current state-of-the-art technologies, such as directed energy deposition technology and automated 3D scanning technology. This framework was designed with the practical considerations where, the scan input obtained from the automated scanning process is in the form of partial scan data. The object’s nominal model is often unavailable or not suitable for use. There are also multiple damages located at different locations, and the need of generating machine toolpaths directly and automatically from the scan data and the reconstructed nominal shape. Various existing frameworks have tackled some of these considerations, however, a framework incorporating all considerations have not yet been proposed. Having all considerations, incorporated, would enable a fully automated remanufacturing process for single or multiple damages located at different locations. A nominal model reconstruction algorithm for multiple damages at different locations and a direct derivation methodology of the machine toolpath shapes from the scan data have been developed to support the realisation of this framework. A case study has been carried out and it has demonstrated the feasibility of using the proposed process framework, the developed algorithm and methodology for automated remanufacturing process. The scanning boundary of the partial scan data has been successfully detected and the edge damage type has been correctly identified. Consequently, the nominal shape of the damage region has been successfully reconstructed according to the identified damage type, and the cavity cutting shape dimensions have been successfully derived from the scan data.