Parameter Optimization of High-Frequency Floor Based on Semirigid Boundary Conditions and Its Effect on the Serviceability of Human-Induced Vibration

Abstract

The accurate representation of real floor behavior is crucial for assessing human-induced vibration serviceability. Many researchers focus on developing computational models and field testing to reflect actual serviceability conditions. Although traditional boundary conditions are commonly used, limited attention has been given to optimizing these conditions. To begin with, this paper introduced semirigid boundary conditions into a finite-element model of the floor and optimized physical parameters and boundary constraint stiffness using the simulated annealing-particle swarm optimization (SA-PSO) method, resulting in a more realistic computational model. Subsequently, a random crowd load model was established by combining the social force model (SFM) and the pedestrian load model, and a random crowd-floor mutual coupling calculation model was established based on the improved pedestrian biomechanical model. In addition, the mode shape functions of the floor with semirigid boundary conditions were obtained by extracting the optimized mode shape vector of the floor and using cubic spline interpolation. The dynamic response of the floor was computed using the modal analysis method, and the accuracy of the suggested method and model was verified by experiment. Furthermore, the effect on the vibration serviceability of the floor was analyzed under crowd walks randomly on the floor with different boundary conditions both before and after optimization. This study revealed: (1) achieving a computational model consistent with real floors requires optimization of relevant parameters and boundary conditions, reducing the error in floor frequency from 39.48% to 5.44% compared to measured results. (2) The human-induced vibration serviceability for existing floors may be misjudged by using traditional boundary conditions. The mean value of peak acceleration of floor with semirigid boundary conditions increased by 35.92%, with a 42% increase in the probability of serviceability problems compared to the SSCC boundary. (3) High-frequency floors also experience human-induced vibration serviceability issues. For a floor with a fundamental frequency of 10.3540 Hz, simulation and experimental results during five-person random walks indicate that peak acceleration at the middle position of the floor exceeds serviceability limits.