An evolutionary power spectral density model of thunderstorm outflows consistent with real-scale time-history records.

This paper presents a new approach for modelling the evolutionary power spectral density (EPSD) of thunderstorm outflows based on a database of 129 real-scale thunderstorm time-histories. Two equivalent methods of decomposition of the wind velocity are investigated, outlining the derivation of the EPSD. The analyses are based on the hypothesis that the residual turbulent fluctuations can be treated as a uniformly modulated process, whose reliability is verified studying the steadiness of its up-crossing rate. The assumption of considering a constant turbulence intensity is checked through the evaluation of the dynamic response of a set of SDOF systems in time-domain and through the response spectrum technique. Two models for the time-modulating function of the EPSD are proposed that include the parameters characteristic of both the thunderstorm event and the background wind. Finally, the steps for the derivation of the EPSD are outlined pointing out the limits of the representation of the PSD of the reduced turbulent fluctuations through spectral models commonly adopted in wind engineering. The final aim of this research is finalising a triad of complementary methods for evaluating the dynamic response of structures to thunderstorm outflows - EPSD, time-domain and response spectrum - consistent with real-scale time-history records. • The non-stationarity of turbulent fluctuations is analysed through their up-crossing rate. • The residual turbulent fluctuations can be dealt with a uniformly modulated process. • Assuming steady turbulence intensity causes overestimations in the dynamic response. • The models proposed for the time-modulating functions encase a physical meaning. • Two different ways of modelling the evolutionary power spectral density are outlined. [ABSTRACT FROM AUTHOR]