Augmented sensing for geared drivetrains

The knowledge of the contact forces and resulting strains gears and bearings is of paramount importance to gain insights into the state and the performance of complex transmissions system. Despite its high relevance, the correct evaluation of the contact forces full strain fields is rarely achieved with today’s available sensors. This means that simulation tools such as Finite Elements (FE) or (flexible) Multibody Dynamics (MBD)simulations are used to derive these quantities – although only in an approximative way. However, due to the complexity of the contact phenomena in gears and bearings, these local quantities are often inaccurately predicted, which forces engineers to oversize their designs. As Early Stage Researcher in the Eco Drive project, my research focuses on the development and validation of an estimation-based augmented sensing technique that makes it possible to accurately assess contact forces and strain fields (e.g., at the tooth root or on the bearing raceways) by combining a few well-positioned and easily accessible sensors (e.g., strain gauges and accelerometer) with advanced numerical gear-contact models. Key innovations beyond the state-of-the-art aim at: (i) the development of a novel augmented sensing strategy that can deal with the challenges of non-smooth state estimation. No such framework of methods that allows this seems to be readily available and applicable to contact problems; (ii) the creation of algorithms for optimal sensor layout for non-smooth mechanics; (iii) the prototyping of a tool for online/offline monitoring of drivetrain systems that include both gears and bearings as critical components.