抄録
A new approach to think about seismic input motions to a large-scale structure was proposed by a consistent idea of wave propagation from seismic source to structure, while focusing an attention on the formulation of semi-empirical equations on the significant upper bounds of period and the lower bounds of acceleration amplitude in relation to earthquake engineering. At first, a spectral acceleration amplitude at far field in an infinite elastic medium was theoretically introduced by means of an extention of a well-known earthquake fault model. The upper bounds of significant period was defined as the maximum period at which the spectral amplitude is kept constant. The actually estimated period T_c(sec) is as log T_c(sec) is as log T_c=0.5M-2.42. hence M is earthquake magnitude. Next, a semi-empirical equation of spectral acceleration amplitude at bedrock was derived as a function of M and epicentral distance γ(km), by substituting observed quantities into the theoretically developed spectrum. The formula is U=6.6×10^<0.5M-2>/γ kine. After taking into consideration of an amplification effect by deep soils this was led to the velocity response spectrum for a one-mass structural system and further to the expression of the Base shear coefficient as a function of M, γ, and period T, which is B_s=2.1×10^<0.5M-3>/(γ・T). This was compared with the recommended base shear coefficients by Architectural Institute of Japan. Finally, a discussion on the acceleration amplitude itself was given and suggested that even the acceleration as several gals is unnegligible for the aseismic construction of the large-scale structures.
