| Abstract: | One of the most common hypotheses implicit in the seismic analyses of structures is that the earthquake input motion is identical at all points beneath the structure. Very little experimental evidence presently is available to supplant this viewpoint. However, one may infer a spatially distributed surface motion of the soil if the earthquake is simply assumed to consist of a complex of surface waves traversing the plan of the structural site. Under these conditions, as shown in the paper, the effects of passing waves must be integrated over the structural area to obtain their net effects as exciting functions to the structure. When this is done for individual Fourier components of the quake, one important result is the diminution or ‘self-cancelling’ effect of some inputs, particularly for those waves the wavelengths of which are comparable to the dimensions of the structure, or shorter. Another important effect is the torsional excitation of the structure. The present paper is not necessarily aimed at replacing present analysis methods but at discussing some of the effects which will inevitably be entrained by the introduction of any information or hypotheses regarding the spatial distribution of earthquake motions. This analysis tends to suggest why higher frequencies are of lesser importance for a structure having a large rigid foundation. |
