Control-Oriented Modeling of a Vehicle Drivetrain for Shuffle and Clunk Mitigation
To be published on April 2, 2019 by SAE International in United States
Flexibility and backlash of vehicle drivelines typically cause unwanted oscillations and noise, known as shuffle and clunk, during tip-in and tip-out events. Computationally efficient and accurate driveline models are necessary for the design and evaluation of torque shaping strategies to mitigate this shuffle and clunk. To accomplish these goals, this paper develops two control-oriented models, i.e., a full-order physics-based model and a reduced-order model, which capture the main dynamics that influence the shuffle and clunk phenomena. The full-order model comprises several components, including the engine as a torque generator, backlash elements as discontinuities, and propeller and axle shafts as compliant elements. This model is experimentally validated using the data collected from a Ford truck. The validation results indicate less than 1% error between the model and measured shuffle oscillation frequencies. The reduced-order model is derived by lumping 24 inertia elements into 2 elements, 4 stiffness and damping elements into 2 elements, and 2 backlashes into 1 element. As part of the reduced-order model development, the paper investigates: (i) the effect of lumping transmission and final drive backlashes; (ii) the effect of different tire models; and (iii) observability, controllability, and computational complexity. Simulation results show that the reduced-order model replicates the behavior of the full-order model with less than 5% error in predicting shuffle frequency, thus making it suitable for the design of torque shaping controllers to mitigate shuffle and clunk.
- Prithvi Reddy - Michigan Technological Univ.
- Kaushal Darokar - Michigan Technological Univ.
- Darrell Robinette - Michigan Technological Univ.
- Mahdi Shahbakhti - Michigan Technological Univ.
- Jason Blough - Michigan Technological Univ.
- Maruthi Ravichandran - Ford Research & Advanced Engineering
- Mary Farmer - Ford Research & Advanced Engineering
- Jeff Doering - Ford Research & Advanced Engineering