Rapid and Automated Alloy Design with Graph Neural Network-Powered LLM-Driven Multi-Agent Systems

Alireza Ghafarollahi,Markus J. Buehler

2024-10-18

Abstract:A multi-agent AI model is used to automate the discovery of new metallic alloys, integrating multimodal data and external knowledge including insights from physics via atomistic simulations. Our multi-agent system features three key components: (a) a suite of LLMs responsible for tasks such as reasoning and planning, (b) a group of AI agents with distinct roles and expertise that dynamically collaborate, and (c) a newly developed graph neural network (GNN) model for rapid retrieval of key physical properties. A set of LLM-driven AI agents collaborate to automate the exploration of the vast design space of MPEAs, guided by predictions from the GNN. We focus on the NbMoTa family of body-centered cubic (bcc) alloys, modeled using an ML-based interatomic potential, and target two key properties: the Peierls barrier and solute/screw dislocation interaction energy. Our GNN model accurately predicts these atomic-scale properties, providing a faster alternative to costly brute-force calculations and reducing the computational burden on multi-agent systems for physics retrieval. This AI system revolutionizes materials discovery by reducing reliance on human expertise and overcoming the limitations of direct all-atom simulations. By synergizing the predictive power of GNNs with the dynamic collaboration of LLM-based agents, the system autonomously navigates vast alloy design spaces, identifying trends in atomic-scale material properties and predicting macro-scale mechanical strength, as demonstrated by several computational experiments. This approach accelerates the discovery of advanced alloys and holds promise for broader applications in other complex systems, marking a significant step forward in automated materials design.

Materials Science,Disordered Systems and Neural Networks,Mesoscale and Nanoscale Physics,Artificial Intelligence,Multiagent Systems

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the challenges in the design of multi - principal - element alloys (MPEAs), especially how to efficiently explore and discover new alloys with optimized or desired properties in the vast compositional space. Specifically, the paper focuses on body - centered cubic (BCC) alloys in the NbMoTa family, aiming to predict two key properties: the Peierls barrier and the solute/screw dislocation interaction energy. These two properties are crucial for understanding the relationship between the microstructure and the macroscopic mechanical properties of alloys, but traditional atomic simulation methods are very costly for calculating these properties.

To overcome this challenge, the paper proposes a multi - agent system based on graph neural networks (GNN), which can quickly and accurately predict the above - mentioned key physical properties, thereby significantly reducing the dependence on expensive atomic simulations. By combining the prediction ability of GNN and the dynamic collaboration of the multi - agent system, this system can autonomously navigate in the vast alloy design space, identify trends in atomic - scale material properties, and predict macroscopic mechanical strength, accelerating the discovery process of advanced alloys.

In summary, the main objectives of the paper are:
1. **Reduce the dependence on human expertise**: Through automated and intelligent means, reduce the need for expert knowledge in the process of new material discovery.
2. **Overcome the limitations of direct all - atom simulations**: Utilize machine learning and deep learning technologies, especially GNN, to replace traditional computational methods and improve computational efficiency.
3. **Accelerate the alloy design process**: Through the collaborative work of the multi - agent system, quickly explore and evaluate the design schemes of different alloy compositions, thereby speeding up the development of new materials.

The achievement of these objectives marks an important progress in the field of material design, which not only accelerates the discovery of alloy materials, but also provides new ideas for the automated design of other complex systems.