Ilenia Battiato - Pushing Multiscale Modeling of Battery Systems through Symbolic-Numeric Computing скачать в хорошем качестве

Ilenia Battiato - Pushing Multiscale Modeling of Battery Systems through Symbolic-Numeric Computing

Recorded 07 October 2025. Ilenia Battiato of Stanford University presents "Upscaling and Automation: Pushing the Boundaries of Multiscale Modeling of Battery Systems through Symbolic-Numeric Computing" at IPAM's Bridging Scales from Atomistic to Continuum in Electrochemical Systems Workshop. Abstract: Batteries are at the heart of both the electrification revolution in the transportation sector as well as Battery Energy Storage Systems (BESS). BESS will play a central role in powering AI while enabling the ongoing decarbonization of electricity networks. While current research is primarily focused on pushing batteries to operate at their physically permissible limits to enhance their market penetration, their current optimization, design and control-based operations primarily rely on semiempirical reduced-order-models (ROMs) that often lack theoretical rigor to accurately and predictively handle the complex multiscale interactions intrinsic to battery systems from the sub-pore all the way to the rack-scale. Such inaccuracies can lead not only to suboptimal design and operations, but also to severe safety risks. In this talk, we first show how mathematical coarse-graining theories can be used to define an integrated and theoretically consistent workflow to connect scales, by formally developing effective/macroscopic (ROM) models with a priori guaranteed accuracy, identifying the dynamic conditions under which such models are predictive, and defining (not fitting) parameters based on fine-scale information. We then demonstrate how automated symbolic deduction can be exploited to advance accurate multiscale model development and discovery, while reducing the time to rigorously derive coarse-grained models from months/years to minutes/seconds. Finally, we demonstrate both the theoretical framework as well as its integration in novel symbolic-numeric codes for two case: electrochemical transport from the pore- to the layer-scale and thermal runaway from the cell to the pack scale. We will also discuss ongoing work focused on consistently integrating DFT with pore-scale models. Learn more online at: https://www.ipam.ucla.edu/programs/wo...