Multi-Physics Numerical Modeling and Design Optimization of a Wet Clutch

Abstract

A wet clutch is a complex hydrodynamic device which couples the rotating elements in the powertrain system to transfer torque from the engine and/or electric motor to the wheels. The clutch plays an important role in the next generation propulsion systems for enabling gear shifting in automatic transmissions, motor speed reduction in electric vehicles (EV), and mode-switching in hybrid electric vehicles (HEV). Typically, the design of clutch applications is evaluated on a component test bench, and the process relies heavily on manual tuning, making it an expensive and time-consuming process. Numerical optimization tools are highly desirable for upfront design improvement, but their development has been difficult due to the complexities in simulating wet clutch dynamics. Therefore, modeling the physics which governs the clutch engagement process is crucial for the development of clutch design and control.

A high-fidelity multi-physics model is developed for predicting the clutch engagement behaviors for complex geometries and operating conditions, with a focus on HEV application. An iterative scheme is used to simulate realistic clutch engagement while balancing the externally applied actuator force against the viscous and mechanical contact forces developed in the clutch interface. The viscous stress during the squeeze film process is modeled using the Navier-Stokes equations. The friction contact is simulated using an empirical model based on the detailed characterization of the friction material as well as the examination and modeling of the associated contact mechanics. The engagement model captures the effects of rotation, squeeze, flow through grooves, mechanical contact, heat transfer, porous media flow, air-entrainment, and cavitation.

In addition, an experimental study is conducted on a programmable engagement test bench to understand the complex dependency of clutch behavior on the design and the operating conditions. Further, the characteristics of the friction material, such as elasticity, oil penetration time, real contact area and surface roughness are analyzed. The experimental results play a vital role in the development of the multi-physics numerical model and the optimization framework. Moreover, the data is valuable for validation of the numerical models.

Based on the wet clutch engagement model, an optimization methodology is developed to evaluate various aspects of clutch design. Parameters such as groove depth, width, angle, and shape are studied under realistic operating conditions and clutch controls. The clutch design optimization framework is developed with the aim of achieving faster squeeze of fluid from the clutch interface. This ensures faster transition from the hydrodynamic to the contact phase of the engagement process, and reduces the non-linearity and hysteresis in the clutch transfer function, i.e. the relationship between clutch torque and applied actuation pressure. The optimization methodology can be generalized to include the evaluation of other design parameters and objective functions. Optimization is performed using gradient-based and the evolutionary algorithms. Various scaling options for the objective function are evaluated. Accordingly, one-parameter, two-parameters and multi-objective optimization cases are studied. An example case for the groove shape optimization is performed to demonstrate the potential of the optimization framework. The development of the present design optimization method paves a path for a systematic analysis of wet clutches during the conceptualization phase of the powertrain design process instead of the trial-and-error methods currently used in practice.