Semi-Active Control of Magnetorheological Elastomer Engine Mounts: Diagnosing the Limits of Transferring Skyhook Control

This paper implements and evaluates closed-loop semi-active control for a magnetorheological elastomer (MRE) engine mount, going beyond passive-state comparisons. An on-off Skyhook law proposed for an MRE seat suspension is adapted to a reduced-order engine–chassis model, with the engine identified as the controlled mass, and compared with a model-based clipped LQR controller. Both laws receive comparable tuning effort and are tested under static, time-varying, broadband random and engine-excitation inputs, using engine heave, chassis heave, engine acceleration, transmitted force and coil power as metrics. Performance is strongly input-dependent: under deterministic inputs, neither law improves appreciably on the best passive state; under broadband random excitation, Skyhook gives a small and unreliable reduction in engine heave that depends on the plant assumptions, and the tuned clipped LQR gives a considerably larger one that persists across them; and under the engine’s own excitation, neither law improves the response and the Skyhook laws fall back on the stiff passive state, which transmits the most force. In the model, the heave reductions come with higher transmitted force than the soft passive state, a penalty that depends strongly on the plant, and the clipped LQR benefit is roughly halved by an actuator lag or delay of ten milliseconds.