VIBRATION ANALYSIS OF A FLOOR SYSTEM BASED ON PROBABILISTIC HUMAN-INDUCED FORCE MODEL

In order to evaluate the vibration performance of actual floor system under probabilistic human activity excitations, dynamic characteristics and probabilistic responses are investigated by in-situ dynamic measurements and numerical simulations. Firstly, frequency response function based shaker modal testing is conducted and the actual modal parameters are identified. Then, probabilistic multi-harmonic walking force models, both for single person walking and for crowd walking, are established, in which the variability of the walking individual and random walking paths are included. The multi-mode responses are calculated under probabilistic walking loads based on the identified modal parameters. Statistical analysis for obtained response samples is performed to get the response distribution and the probability of certain vibration level. Results show that the human-induced vibration response approximately follows normal distribution, and walking frequency plays a dominant role to the response distribution. In addition, probabilistic evaluation for human-induced vibration response can be conducted based on the obtained response distribution characteristics. In the process of crowd walking response analysis, the free decay response analysis for tested acceleration response is conducted to obtain the variation of modal properties of floor under different response levels. Results show that the variation of natural frequency with different response levels is small, but the variation of modal damping ratios is relatively obvious. The results can provide a reference for probabilistic evaluation of human-induced vibration of the similar structure.