Abstract
In this report, the seismic response of the steel stack was treated from the statistical standpoint. The probability distribution of the maximum value of the response in the duration of the earthquake was calculated for the given value of them aximum base acceleration. The distribution spreaded over the considerably wide range, the upper value of the range (defined by the sufficiently small probablity (being 1.5〜2 times the value of mean value. Parallel to the theoretical study, the experiment was performed using the random shaking table. The models, steel canti-levers which had the same cylinderical cross-section and different hesghts, were tested. The results were in good agreement with the theroetical values. Scattering of the data of many ensembles was well explained by the probability distribution which was mentioned above. Thus it was concluded that the maximum response to a typical seismic acceleration pattern is only one point between probable wide range. Therefore we must consider this fact as well as the probable maximum seismic base acceleration.
