Dynamic analysis and mitigation strategies for torsional oscillations in rotating systems

<p dir="ltr">Rotating systems are central to many engineering applications, including automotive, industrial machinery, and aerospace systems. A persistent challenge in such systems is torsional vibration, which arises from the mechanical and geometric characteristics of the components. In automotive drivelines, torsional vibrations not only accelerate fatigue failure of key components but also contribute to driver discomfort and drowsiness. While extensive research has examined driveline vibrations, existing models generally lack the fidelity required to capture nonlinear excitations or to conduct component-level modal analysis within complex assemblies. This thesis addresses these gaps by developing high-fidelity models and control strategies to analyze and mitigate torsional vibrations in automotive drivelines.</p><p dir="ltr">The work identifies four primary excitation sources: (i) combustion-induced torsional vibrations from internal combustion engines (ICE), (ii) friction-induced vibrations (FIV) from wet clutches, (iii) parametric excitations from universal joints (U-joints), and (iv) clearance-induced torsional excitations in U-joints. A novel mathematical model is proposed to capture these nonlinear effects while enabling component-level modal analysis to identify the most affected driveline elements. Validation is carried out against experimental data collected from two Class 8 medium-duty vehicles equipped with 6-speed automatic transmissions powered by a Cummins 6.7L diesel engine.</p><p dir="ltr">Experimental results reveal resonances at 30 Hz and 100 Hz, coinciding with engine orders at 2000 rpm, which excite high-amplitude torsional oscillations in the planetary carrier and input clutch hub, ultimately leading to fatigue failure. The simulation model successfully reproduces these findings and identifies 100 Hz as a natural mode. Modal analysis confirms that excitations at these frequencies result in significant torsional instabilities at the input clutch hub and planetary carrier. To address this, design modifications are suggested to adjust turbine shaft inertia and stiffness, thereby shifting natural frequencies outside the operating range. Additionally, an active gear-shifting control strategy is developed to prevent resonance in gears 4 and 5, with feasibility demonstrated experimentally in highway driving conditions.</p><p dir="ltr">The thesis also addresses friction-induced vibrations in wet clutches, a less explored but critical source of driveline torsional instability. During clutch engagement, self-excited oscillations driven by the negative slope of the friction–velocity characteristic give rise to vibro-acoustic phenomena such as squawk. To mitigate these effects, a novel constraint sliding mode control with reaching law (SMC+RL) is proposed. This control framework enforces physical constraints on clamping force and slip-speed dynamics, achieving robust suppression of FIVs while ensuring accurate slip tracking (≤ 3 rad/s deviation) and timely clutch lock-up (≤ 0.1 s deviation). Benchmarking demonstrates superior performance and computational efficiency compared to existing advanced controllers. The approach is experimentally validated through processor-in-the-loop testing on an automotive-grade microcontroller, confirming real-time feasibility.</p><p dir="ltr">Finally, the work advances understanding of U-joint dynamics. Using Floquet theory, parametric resonances are identified in two-U-joint drivelines at small joint angles $(≤ 5^{\circ})$, where subharmonic and combination resonances can occur. Furthermore, clearance-induced torsional excitations, which remain largely unexplored in the literature, are investigated through both piecewise-smooth and non-smooth multibody models. The novel set-valued formulation developed here captures unilateral frictional impacts between the crosspiece and yoke, revealing rich nonlinear behaviors including periodic, quasi-periodic, and chaotic dynamics.</p><p dir="ltr">Overall, this thesis makes significant contributions by identifying key sources of torsional vibrations in automotive drivelines, proposing high-fidelity modeling frameworks, and developing robust passive and active control strategies. The findings advance both the theoretical understanding and practical mitigation of driveline torsional vibrations, with potential applications across broader rotating system domains.</p>