Human-structure dynamic interaction between building floors and walking occupants in vertical direction

Abstract

While modern building floors feature lightweight materials and slender structural elements, their dynamic interaction with walking occupants has not been quantified. This is despite the proven and significant influence of this interaction on human-induced vibration levels of other types of lightweight structures, such as footbridges. This work presents an experimental study to quantify the effect of walking pedestrians on the frequency response functions (FRFs), which are dependant on the corresponding modal properties, of two floors, a relatively light floor with low fundamental frequency and a heavier floor with higher fundamental frequency. It also proposes an improved methodology to take into account the interaction between walking pedestrians and supporting floors in the response calculation of human-induced vibrations. Instead of the conventional mass-spring-damper or inverted-pendulum models, the proposed model utilises two experimentally-driven transfer functions, related to the dynamics of walking individuals, over a range of frequencies between 1 Hz and 10 Hz, to mathematically describe the dynamics of this interaction. Hence, the proposed model is relevant to floors with fundamental frequency less than 10 Hz (i.e. low-frequency floors). The results show that walking occupants can cause significant reduction in the amplitudes of the FRFs. This reduction ranges from 44% and 62% for a floor occupied by two or six walking pedestrians, respectively, to 10% for a heavier floor with a higher fundamental frequency occupied by six walking pedestrians. This implies that ignoring this phenomenon in the design can result in an overestimation of the predicted vibration levels. This is especially the case for floors with relatively low fundamental frequency and modal mass. Furthermore, the derived transfer functions related to the dynamics of walking individuals indicated the existence of three whole-body modes of vibration with frequency less than 10 Hz. The performance of the proposed humanstructure interaction model is verified with experimental measurements of vibration responses related to individual occupants walking on three floors. The simulated vibration levels are consistent with their measured counterparts indicating the applicability of the proposed model.