The present work deals with the modelling of Equivalent Static Wind loads ESWL. The ESWL are expressed as a combination of mean, resonant and background parts.
Thereby, the aim is to give a simple static load case based on experimental data. The resonant part is formulated as modal inertial load whilst the background part
is determined with the Load-Response-Correlation LRC approach as well as a modification of it. The LRC approach considers correlation effects in combination with
influence functions and gains a better generality through the modification. As both methods result in the same static response, the latter might have more potential
to be integrated into standards. In general, the LRC reproduce responses in the alongwind direction that are in a comparable magnitude to the results from time history
analyses. For the crosswind direction, the LRC tend to overestimate the responses compared to time history analyses. As in this case the resonant part is dominating the
total ESWL, an adaption of the resonant peak factor could be necessary.
A 3D beam model is available for the required structural analysis, including eigenvalue, static and dynamic analyses. Optimization procedures are used to adjust the
material properties to obtain a beam with the dynamic behaviour of a particular tall building. By introducing parametric, off-diagonal terms in the stiffness matrix,
this model is extended to a bending-torsion coupled beam. In order to reach the given dynamic properties, other stiffness terms have to be adapted subsequently. By comparing
results of time history analyses of the uncoupled to the coupled model, it can be observed that the peak base moments in the three primary directions as well as the total
top point acceleration are reduced. Only the torsional acceleration at the top increases. However, these results cannot be attributed exclusively to the coupling due to
necessary stiffness adjustments in the optimization procedures.
For the determination of peak loads from results of time history analyses, methods of extreme value analysis are evaluated in detail. Methods based on block maxima and
peak over threshold extreme value selection are examined. For the evaluation of extreme values, several values for the input parameters block size and threshold, respectively
should be evaluated. Individual parameter values result in extreme values that do not represent a meaningful extreme. In this regard, the translational process approach proposed
by the National Institute of Standards and Technology NIST proposes a reasonable alternative. The entire sample is used to determine an extreme value. Consequently, no critical
parameter selection is necessary and reasonable values are obtained.
«
The present work deals with the modelling of Equivalent Static Wind loads ESWL. The ESWL are expressed as a combination of mean, resonant and background parts.
Thereby, the aim is to give a simple static load case based on experimental data. The resonant part is formulated as modal inertial load whilst the background part
is determined with the Load-Response-Correlation LRC approach as well as a modification of it. The LRC approach considers correlation effects in combination with
influence...
»