Multiphysics Modeling of Gearbox NVH in Electric Drivetrains: Methods, Tools, and Trends

In modern electric vehicles (EVs), where the absence of a combustion engine reveals new acoustic challenges, gear and gearbox noise—especially tonal “whine”—has emerged as a prominent NVH (Noise, Vibration, and Harshness) concern. This review investigates the state-of-the-art multiphysics simulation workflows capable of predicting NVH from root excitation through structural vibration and up to radiated airborne noise. Emphasis is placed on software ecosystems developed between 2015 and 2025, including Romax, AVL EXCITE, Siemens Simcenter, SMT MASTA, MSC Adams/Nastran/Actran, KISSsoft + RecurDyn, and COMSOL Multiphysics. The review explores simulation layers ranging from analytic torsional models to coupled flexible multibody dynamics (MBD), finite-element structural response, and acoustic FEM/BEM methods. Recent trends such as per-tooth microgeometry definition, flank waviness modelling, use of measured topography (e.g., CMM data), and digital twin concepts are discussed in depth. Furthermore, the review highlights validation challenges—especially the limited system-level correlation between predicted and measured noise—and identifies research gaps regarding EV-specific excitations, manufacturing variation modeling, and NVH-oriented design optimization. This work aims to give engineers and researchers a structured overview of integrated CAE methods to “front-load” gearbox NVH prediction in electrified drivetrains, thereby improving design cycles and acoustic performance.