Researchers at the National Center for Supercomputing Applications (NCSA) and The Grainger College of Engineering, University of Illinois Urbana-Champaign (UIUC), have made strides in stress prediction research using artificial intelligence. Their work primarily focuses on deep operator network (DeepONet) applications, which they aim to use to enhance stress response predictions in intricate geometrical shapes, notably those present in additive manufacturing processes. Using the NCSA’s Delta system, they managed to achieve faster results than the usual finite element methods.
The research was undertaken via Illinois Computes, an initiative known for providing extensive computing and data storage capabilities. The initiative has enabled collaborations across different fields, forming a unique combination of machine learning and computational mechanics. The Delta system, admired for its superior GPU computing capabilities, was essential in the training of deep neural networks and the generation of training data through the use of Abaqus software.
Two leading publications have been developed from this research. The first, published in “Computer Methods in Applied Mechanics and Engineering,” introduces a novel DeepONet using a residual U-Net (ResUNet) for coding complex geometries. This approach signifies the first usage of ResUNet in the architecture of DeepONet, showing greater memory efficiency and flexibility against traditional methods.
The subsequent study, unveiled in “Engineering Applications of Artificial Intelligence,” indicates another groundbreaking version of DeepONet, known as S-DeepONet. This enhanced method utilizes cutting-edge sequential learning strategies, presenting an improved precision in multi-physics solutions that depend on changing thermal and mechanical loads.
“The additive manufacturing method is a transformative manufacturing approach that endlessly broadens the scope for its utilization,” stated Iwona Jasiuk, a professor of mechanical science and engineering at UIUC.
“DeepONet operates as a robust and swift computational instrument, capable of simulating the additive manufacturing procedure at different spatial and temporal scales. These simulations are required for a comprehensive comprehension of the additive manufacturing operation, including its application and monitoring.”
This investigation extends beyond just advancements in AI applications, it also carries significant effects on advanced manufacturing procedures and the growth of digital twins. The joint venture between NCSA and MechSE underscores the harmony between multidisciplinary expertise and the forefront of technology.
Come and let us know your thoughts on our Facebook, Twitter, and LinkedIn pages, and don’t forget to sign up for our weekly additive manufacturing newsletter to get all the latest stories delivered right to your inbox.
“Why did the 3D printer go to therapy? Because it had too many layers of unresolved issues!”
0 Comments