Footfall Induced Vibration Response of a Base-Isolated Steel Building

Footfall induced vibrations can occur in elevated walking surfaces in structures due to normal human activities such as walking and running. If the vibrations are too high, they can cause occupant discomfort. Being able to reliably estimate the dynamic properties of the supporting structural system is a key part of designing a structure to limit such vibrations. However, currently, methods to assess the floor vibration response in realistic structures is not well known. This paper describes walking tests at an office building in the Christchurch CBD designed for earthquake both before and after fitout to determine properties and response. Sensitivity of floors to vibration response parameters such as the individual’s weight, the walking frequency and the presence of non-structural elements is considered. The results are then compared to numerical analyses which were conducted using OASYS, a commercial footfall analysis package, and SeismoStruct, a frame analysis software package. It was found from experimental tests dimensionless response factors (R-values) were generally less than 5 even for pre fit-out indicating that the floor considered is likely to be comfortable for occupants. The effect of fit-out tended to decrease the R-value by a factor of 1.2, but there was considerable scatter. The numerical OASYS and SeismoStruct models had similar stiffness and damping. However, they tended to estimate R-values greater about 1.0 larger than that found from walking field tests, which could be due to the fact that a larger damping value of the actual floor were calculated. These were 7 % and 10 % for pre fit-out and post fit-out respectively. In addition, in the field walking tests, the frequency range of 1.8-2.2 Hz had a negligible effect on the response. There was generally positive correlation between walker mass and R-value.