Dr. Chengkun Huang, Los Alamos National Laboratory
Abstract:
Over the past century, particle accelerators have undergone remarkable development and evolution, pushing the boundaries of particle and nuclear physics. Today, their influence extends far beyond probing the universe鈥檚 fundamental mysteries. These sophisticated machines drive innovation in materials science, biology, and energy research by powering advanced radiation and light sources, and transform fields such as medicine, industry, and national security. With breakthroughs like high-energy, high-intensity lasers, next-generation accelerators are on the horizon, promising to deliver cost-effective, compact solutions that could revolutionize science and technology even further. In this presentation, I will trace the evolution of particle accelerators and highlight how advances in computation and data-driven science are enabling us to meet the growing complexity of next-generation systems. In particular, I will introduce a promising framework for the development of digital twins of particle accelerators, combining high-fidelity, multi-scale, multi-physics computational modeling with surrogate models enhanced by Scientific Machine-Learning (Sci-ML) techniques that embed physical principles at their core. Such an approach opens the door to future particle accelerators that are not only more powerful, but also more automated and deeply integrated into the scientific discovery process.