Numerical Analysis and Experimental Assessment of the Dynamic Characteristics of Periodic Structures

Vibrations in the built environment result from road and railway traffic, construction, and industrial activities. In a frequency range between 1 and 80 Hz, these vibrations may cause annoyance to people or disturbance to sensitive equipment, while between 16 and 250 Hz they may lead to structure borne noise. Large buildings such as office, school and residential buildings often have a regular geometry consisting of repeated units. The present PhD project focuses on the accurate numerical prediction of railway-induced vibrations in periodic building structures. Two main challenges exist within the frame of this project. Firstly, the numerical prediction of vibrations in new built or existing structures in a frequency range up to 80 Hz or even 250 Hz requires a sufficiently fine mesh when using the conventional finite element method, leading to large systems of equations and a high computational cost. The second challenge is the difficulty in the calibration of numerical models based on experimental vibration data. Clusters of modes with similar mode shapes occur in periodic structures, complicating the comparison of numerical and experimental results. The objectives of this project are to study the behavior of regular structures, including the development of numerical methods which exploit the regular geometry, to obtain physical insight in the dynamic behavior of repetitive building structures, and to provide guidance for experimental design. The findings from the study will be verified based on in situ experiments on a building close to a railway line.