Experimental and Simulation Investigation of an Adaptive Pendulum-Tuned Mass Damper for Engineering Structures

Abstract

Slender civil structures remain vulnerable to dynamic actions, making robust, implementation-ready damping strategies a design priority. Building on our earlier feasibility study, which introduced an active, liquid-based pendulum-tuned mass damper with mass redistribution for real-time retuning in skyscrapers, the present work generalizes the concept to an adaptive pendulum-tuned mass damper in which variable pendulum length is the primary mechanism of adaptation — independent of any specific actuation technology. We developed a compact model of the coupled flexible frame–adaptive pendulum-tuned mass damper system with explicitly  linearized  descriptors  suited  for  tuning  and  design. Then, we validated it experimentally on a shake-table using a flexible frame equipped with an adaptive pendulum-tuned mass damper. The quantitative comparison considers tuned and detuned settings under identical operating conditions. In numerical studies, adaptive length control at resonance achieves up to 85% suppression relative to a fixed-length tuned mass damper, confirming the performance margin available from geometric retuning. In shake-table experiments, the undamped reference shows a peak floor acceleration  of  4.8 m/s². When  optimally  tuned  (pendulum length 2 cm), the peak reduces to 1.8–2.0 m/s², i.e., by ≈ 60–63%. Detuned settings yield 3.7 m/s² (5 cm) and 3.2 m/s² (7 cm) at 5 Hz, corresponding to ≈ 23% and ≈ 33% reductions, respectively. These measured values, together with the model–experiment error analysis, substantiate the predictive accuracy of the proposed framework and quantify the sensitivity to mistuning. The novelty of this study lies in the generalization and experimental validation of an adaptive pendulum-tuned mass damper with variable pendulum length as a practical solution for vibration control in slender engineering structures. Unlike earlier feasibility studies focused on a specific liquid-based implementation, the proposed compact modeling framework and shake-table experiments confirm superior mitigation efficiency (up to 80–85% at resonance) and provide directly applicable design guidelines for real structures.