Abstract
KEYWORDS – Transmission Architecture, Drivetrain Modeling, Sensitivity Analysis, Resonant Mode, Shift Control
ABSTRACT – Drivetrain NVH characteristics are very important to vehicle drive quality, and if not controlled properly, undesired drive quality issues may be perceived by the driver through audible sound, drive torque and chassis, etc. To date, the drivetrain NVH-related studies have been mostly focused on the driving conditions with the transmission in gear, and the excitation forces mainly come from the exogenous inputs like the engine or brakes. In these studies, the transmission is usually over simplified by some lumped inertias and ratios, and the resulting models are not adequate to study the drivetrain dynamic processes involving transmission components or those during gear shifting with clutches as possible system inputs. On the other hand, the structure models used in the transmission community are commonly inertia-based with transmission dynamic components (e.g., stiffnesses) neglected. Such models are sufficient for energy related analysis and control synthesis, but incapable to predict drivetrain dynamic behaviors during the shift. As the transmission system is trending towards more gear steps, the vehicle drive with the transmission being through gear shifting will grow and account for a significant amount of time on a driving cycle. Therefore, it becomes important to understand how the drivetrain structures response to a clutch input in order to achieve smooth shift quality. Although there are ample studies on both drivetrain dynamics and transmission shift controls, neither has the adequate model and proper methods to address the unique drivetrain dynamic processes while the transmission is through gear shifting. Until now, there has been little fundamental study on the drivetrain structures, dynamic characteristics and their influences on the shift controls in the literature. Through the years, rules of thumb have been generated by transmission engineers on certain drivetrain structures and the control approaches with the risk of shift vibrations. However, questions like what is the physics behind a shift vibration issue; why some shifts are inherently more difficult to control than others; and why the same controls works well on this transmission but not on similar shifts on the other transmission, remain unclear to a large degree. In this paper, we fill the gap between the normal practices used for the study of the drivetrain system and the need for more generic models and methods for the drivetrain system during gear shifting. The modeling framework and the novel sensitivity analysis method developed allow us, for the first time, to gain deep understandings of the drivetrain dynamics involving the transmission components. Generic drivetrain structures have been identified including the key contributing components to the relevant resonant modes. The results can clearly reveal how the system can possibly interact to induce vibrations. The method developed here is generic and can be used for transmission control design, architecture selection, damper design, drivetrain dynamic analysis and troubleshooting, etc. It can also benefit the study on vehicle dynamics and other NVH related analysis.