Human-induced Stochastic Vibration And Optimal Control For Long-span Footbridges And Floors

With the widespread application of high-performance materials in modern civil engineering, such as the high-strength steel and fiber reinforced polymer, more and more structures with light mass, flexibility and low damping appear in construction. Correspondingly, the crowd-induced resonance, resulting from the coincidence of human-excitation frequency spectrum and structure natural frequency, may lead to the excessive vibration which compromises pedestrians'or residents'vibration comfort. However, the current research regarding human-induced resonance and its control is far from success, e.g. lack of systematic vibration assessment procedure of dynamic behavior of such structures. This thesis investigates the vibration characteristics of long-span flexible structures induced by various crowd random activities using stochastic vibration modelling, and further presents the corresponding stochastic optimization procedure of multiple tuned mass dampers (MTMD) in suppressing crowd-induced vibration. The main concerns are as follows:1. Based on traffic flow theory, multi-point loading procedure put forward by Ellingwood and Tallin, standard single foot force model developed by Chen Yu, and pseudo excitation method proposed by Lin Jiahao, the crowd-footbridge stochastic vibration model, in which pedestrians are modeled as a crowd flow characterized with the average time headway, is developed to consider the worst vibration state of footbridge. In this stochastic vibration model, an analytic formulation is developed to calculate the acceleration power spectral density in arbitrary positions of footbridge with arbitrary span layout. Resonant effect is observed as the footbridge natural frequencies fall within the frequency bandwidth of crowd excitation.2. To suppress the excessive acceleration under human walking, a MTMD system is used to improve the footbridge dynamic characteristics. According to the crowd-induced footbridge stochastic vibration model, an optimization procedure, based on the minimization of maximum root-mean-square acceleration in the footbridge, is introduced to determine the optimal design parameters of MTMD system. Numerical analysis shows that the proposed MTMD system is more effective and reliable than a single TMD in reducing dynamic response during crowd-footbridge resonance, and the proper frequency spacing enlargement will effectively reduce the off-tuning effect due to the variation of footbridge natural frequency.3. By using finite element analysis (FEA), a standard procedure is proposed to check the vibration comfort of floor under pedestrian excitation. Moreover, regarding the rhythmic excitation, a stochastic vibration model is developed to assess the floor resonance. This model is not only suitable for the single-mode resonance but also justified by the multi-mode resonance for floors with modal frequencies closely spaced. When calculating the maximum root-mean-square acceleration envelope for floors, this model is capable of considering the various loading regions and the entire excitation frequency bandwidth of human rhythmic activity.4. Based on the loading characteristic of human walking and rhythmic activity and genetic algorithm, a stochastic optimization procedure is introduced to quantify the optimal design parameters of MTMD vibration-suppressing system. This procedure accounts for various loading regions and the variability of excitation frequency, and furthermore has the potential to avoid the ineffectiveness of vibration suppression for the floor with closely-spaced modal frequencies, which is the prominent disadvantage of traditional TMD design methodology.5. The stochastic vibration models, together with the optimization procedure of MTMD design stated above are extended to the simpler manual calculation procedure for checking human vibration comfort requirements, and vibration-suppressing methods accounting for various human-induced excitation and structure categories.