According to the Arizona State University (ASU), researchers use artificial intelligence to significantly improve the metal -3d pressure, which makes it faster, more precise and less wasteful. Professors Aviral Shrivastava and Ashif Iquebal lead the load through their project “Compam: Activating Computer Additive Manufacturing”.
The team's challenge was to print 3D a five-axis 316l stainless steel-marine propeller with sophisticated geometry and performance requirements and at the same time to be carefully controlled the microstructure of the metal. Their goal was to reach grain sizes under a micron – smaller than the silk strand of a spider – which drastically improves the material properties.
“When we perform metal pressure, the quality of the metal actually depends on the cooling curve,” said Shrivastava. Conventional methods require either months of simulations for supercomputers or expensive experiments and errors. The team aims to cut the drastically by developing a physical, Ki -operated system that forms in real time during pressure in real time. Instead of brute force simulations, your model only intelligently focuses on critical zones and skip parts that remain stable. This not only lowers the simulation time, but also increases the accuracy.
“Physics is only a number of rules obeyed in the real world,” said Shrivastava. By combining these rules with data -controlled learning, the AI adapts without relying on massive data records.
“The real value of this work is the ability to bridge research and industrial needs,” said Iquebal. In sectors such as aerospace or defense, in which the material performance is not negotiable, the ability to predict and optimize material properties in advance is a game channel.
With the state-of-the-art 3D printer from ASU-equipped with lasers and a robot arm with six axes, the team is predicted with the actual microstructures in a printed propeller. Your results are evaluated against conventional methods and the tools you have developed are created open source.
These efforts not only accelerate the advanced production, but also weaves the AI directly in engineering training and industry applications – and transforms theory into a tangible, metallic reality.