Architecture Development of Digital Twin-Based Wire Arc Directed Energy Deposition
Chang Jong Kim, Sung Han, Bo Wook Seo, Inyeong Bae, Duckbong Kim, Seok Kim, Young Tae Cho
Abstract
Abstract This study proposes a comprehensive digital twin (DT)-based architecture for wire arc directed energy deposition (WA-DED) to address critical challenges in digital manufacturing, particularly focusing on interoperability, real-time process control, and defect monitoring. Built upon the ISO-23247 framework, the architecture integrates standardized approaches and advanced technologies to enable systematic anomaly detection and precise process management. Key components include flexible robot trajectory generation and path strategy modules, enhancing design flexibility and manufacturing efficiency. Bidirectional data flow is facilitated through an open-source platform using socket communication, ensuring seamless interoperability between industrial robots and various applications. Real-time anomaly detection and prediction are implemented using YOLOv10 and LabVIEW, enabling the identification of deposition defects through 2D data analysis and real-time prediction of contact tip to work distance (CTWD) using 1D data. These capabilities improve process monitoring and provide critical insights for quality control. Although full cloud integration is yet to be realized, future developments aim to establish continuous feedback loops through advanced cloud-based data storage and network systems. This will further enhance autonomous deposition processes by enabling dynamic responses to real-time changes in heat input, temperature, and deformation. The proposed DT-based WA-DED system represents a significant advancement in intelligent, cyber-physical manufacturing environments. It lays the foundation for enhanced process optimization, paving the way for future research and industrial applications in high-performance metal additive manufacturing.