Floor live load survey and modelling by house VR viewing and public registration information

Abstract

Reliable live load value is a prerequisite for the safety design of civil structures, and the quantity of samples is the key data foundation for live load modelling. The traditional walk-in survey approach has encountered challenges such as low efficiency, high costs, and potential privacy concerns, resulting in a limited collection of live load samples and a halt in load surveys over the past three to four decades. To address these issues, this paper proposes a new online survey approach for floor live load sampling. The new approach uses a panoramic VR (virtual reality) house viewing system, available and open-access on many real estate trading websites, to collect load amplitude samples. Additionally, it utilizes public information systems, including real estate transactions and national enterprise credit system, to collect live load duration samples. The practical implementation of this novel approach yields a substantial volume of samples, with 1883 and 4676 load amplitude samples, and 121,561 and 46,353 duration samples obtained for office and residential buildings, respectively, constituting the largest live load sample set reported to date. Statistical analysis reveals that design live load values derived from the new samples are higher than the recommended values in the current code, due to a notable increase in the standard deviation of load amplitude and a significant decrease in the load duration. Overall, the proposed approach presents a novel, efficient, and convenient method for live load modelling.