| Abstract Scope |
Directed Energy Deposition (DED) is highly sensitive to thermal instability, which strongly influences melt pool behavior, heat accumulation, microstructural evolution, and final mechanical properties. However, existing studies primarily focus on melt pool analysis alone, providing limited understanding of overall thermal distribution and penetration behavior during multilayer deposition. This study leverages multiple cameras for comprehensive thermal monitoring and prediction in DED. A top-view infrared camera captures melt pool geometry and thermal distribution, while a side-view camera monitors heat penetration and thermal cycling behavior. Instead of analyzing only the melt pool region, isothermal segmentation at hot zone is introduced to characterize thermal distribution with improved consistency and physical interpretability. Furthermore, a data-driven feedforward prediction framework is developed to predict thermal distribution under varying process conditions, establishing deeper relationships between process parameters and spatial thermal behavior in DED. |