Significant dynamic loads can be generated by crowds of people such as spectators at a sporting event or a rock music concert, and revelers at a discotheque. In the U.S. building codes, the recommended design values are considered to be deterministic statically equivalent values; thus, the variability of the loads is ignored, and a probabilistic design is not possible. Also, since the frequency components o f the loads are not codified, the designer has no means of checking serviceability criteria such as excessive vibration.

We used a large suspended-type force platform to measure dynamic loads of individuals and small groups. We measured both steady state and transient dynamic loads for individuals and groups of two and four performing seven prescribed maneuvers. Periodic loads of individuals were modeled using the Fourier series, and transient loads are composed of one or more exponentially decaying functions. The descriptive parameters obtained from the modeling processes are functions of several independent random variables; we performed multivariate regression analysis using these parameters to yield a model for a person's load history. Shifting individual load-time histories and superimposing the results yielded group loads. The phase angles were exponentially distributed.

Using Monte Carlo simulation, we randomly generated periodic and transient loads for groups of four. The simulated results compared closely with the corresponding laboratory-measured data. The acceptability of the results for periodic loading were further checked by comparing the simulated and measured frequency spectra. The latter revealed that, in addition to a dominant harmonic corresponding to the load fundamental frequency, the periodic loading also has secondary dominant harmonics that must not be ignored.

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £78 +P&P)