Collaborative Research: Elements: Unifying Framework for Multiphysics Spray Simulation with AI-Enhanced Usability

Sprays are found everywhere in modern life, from fuel injection into engines and medicine delivery by inhalers, to painting and coating processes on consumer goods and the manufacturing of powders in food and pharmaceutical industries. The performance of these systems, in terms of energy use, emissions, product quality, and manufacturing cost, is directly governed by the spray. The central challenge is predicting the spray itself, which involves many physical processes that unfold at once and across vastly different sizes: a liquid sheet first tears into thin threads and then into a fine mist of droplets, which then disperse, evaporate, react, or form solid particles. Because no single predictive tool captures all these effects, engineers and scientists have long relied on simplified trial-and-error models. The most capable simulation methods are scattered across separate tools usable only by a small number of experts. This project addresses this critical gap by unifying these advanced capabilities into a single open-source framework. A built-in artificial intelligence (AI) assistant lowers the expertise barrier and shortens the time needed to set up a realistic simulation from weeks to hours. By making predictive spray simulation practical and widely usable, the project advances the national interest through improved energy efficiency and reduced emissions in transportation and power, better manufacturing of medicines, foods, and coatings, and stronger competitiveness for American industry. The project also develops students and the broader workforce through undergraduate and graduate course modules and through workshops held at the leading professional conference for liquid atomization and spray systems. This project develops a unified, open-source, high-performance simulation framework for predictive multiphysics spray modeling that brings together capabilities currently available only in separate, specialized codes. Within a single discretely conservative formulation, the framework couples sharp capture of the liquid-gas interface, sub-grid models for the breakup of thin films and threads into droplets, tracking of individual droplets and particles as they disperse through a turbulent gas, multicomponent evaporation and phase change, and automatically reduced chemical kinetics, spanning both low-speed and compressible flow regimes. The framework integrates these individually mature components into a production-grade, scalable code with documented, reproducible benchmarks and measurable performance. To make the resulting tool broadly usable, a retrieval-augmented AI assistant is deployed that draws on documentation to perform parameter checks and generate simulation cases from natural-language descriptions, reducing setup time from weeks to hours. The framework is validated against partner-provided experimental datasets across a set of flagship applications, including spray combustion, spray drying, surface coating, and droplet impact and erosion. The work delivers a reproducible, openly available research solver, validated benchmarks, and an accessible interface, providing shared infrastructure that enables new multiphysics research and engineering design beyond the duration of the award. This award by the Office of Advanced Cyberinfrastructure is supported by the Section of Civil, Mechanical, and Manufacturing Innovation, and the Section of Chemical, Bioengineering, Environmental and Transport Systems in the Directorate for Engineering. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.