Physics-governed executable modelling of triboelectric nanogenerators

Triboelectric nanogenerator (TENG) research has historically relied on a fragmented collection of analytical theories and numerical solvers. These disparate tools often operate in isolation, making it difficult to maintain physical consistency across different device geometries or to reproduce simulation results. The lack of a unified, executable workflow forces researchers to manually coordinate geometry, material parameters, and boundary conditions, which often obscures the underlying physical assumptions and limits the scalability of TENG design studies.

To address these challenges, the authors developed TENG-CLAW, a platform that treats TENG modeling as a charge-defined electrostatic closure. The framework establishes a self-consistent hierarchy where triboelectric, pre-charging, and compensating charges serve as the primary state variables. This approach allows the system to bridge the gap between the infinite-plate analytical limit—suitable for near-uniform fields—and finite-geometry numerical formulations required for edge-dominated devices. By using the Method of Moments to solve the coupled integral equations, the platform captures complex phenomena like edge-localized charge accumulation and field distortion without relying on empirical corrections.

Beyond the physical model, TENG-CLAW introduces an executable workflow that converts open-ended research requests into structured, physically admissible tasks. The platform compiles user intent into a typed intermediate representation (TENG-IR) and subjects it to a deterministic preflight check. This governance layer evaluates unit consistency, solver compatibility, and model validity before execution. By storing results in an experiment graph, the platform ensures that every output is linked to its specific solver route, assumptions, and input parameters, facilitating reproducible research and long-term design reuse.