Technology
材料研发
Positive
Northeastern University UQ-KGAT Framework: AI Knowledge Graph Directly Designs Printable Alloys — KG-AMS Nickel-Based Alloy Achieves 1307 MPa Room-Temperature Strength
2026-09-03
·
Impact score: 18
A research team at Northeastern University has published the UQ-KGAT framework in Nature Communications, encoding alloy composition, printing process, solidification behavior, and performance into a knowledge graph. Using only 142 L-DED samples to jointly search composition and process parameters, they designed the KG-AMS nickel-based alloy, which exhibits a room-temperature tensile strength of 1307 MPa with 22.2% elongation, retains 655 MPa at 900°C, and achieves a CT porosity as low as 0.038%. Using the same framework with 163 literature samples, they designed the LPBF aluminum alloy KG-AMAA. This work elevates AI from optimizing print parameters to directly designing printable alloys, with general applicability across alloy systems and processing routes.
📰 News Source
This article was translated from a Chinese source.
Source URL not available
Show Chinese original
东北大学UQ-KGAT框架：AI知识图谱直接设计可打印合金，KG-AMS镍基合金室温强度1307MPa
东北大学团队在《自然·通讯》发表UQ-KGAT框架，将合金成分、打印工艺、凝固行为与性能编码为知识图谱，仅用142个L-DED样品联合搜索成分与参数，设计出KG-AMS镍基合金（室温抗拉强度1307MPa、延伸率22.2%，900°C仍保持655MPa，CT孔隙率最低0.038%）；同框架163个文献样品设计LPBF铝合金KG-AMAA。AI从优化打印参数升级为直接设计可打印合金，跨合金体系、跨工艺通用。