Variabilities in the Earthquake Ground-motion Modelling-Source,Path and Site Effects, for Engineering Designs

摘要

Engineers face many types of variability in their design of structures under earthquake loads, including estimates of the design ground motions, probable material strength, and reliability and performance of numerical models for their seismic design analyses. Performance-based engineering means that engineers need to design, evaluate and construct engineered facilities that meet, as economically as possible, the "uncertain" future demands that both owner-users and nature will put upon them. Modern structural engineers can design and build structures to achieve a performance that is reasonably similar to that intended. Critical dynamic parameters, such natural periods, failure mechanisms and displacement capacity can be predicted well within a factor of two. However, the estimates of ground motions used for engineering designs are very "uncertain", and in many cases it is even not possible to know "precisely" how large is the uncertainty of any particular type. In this manuscript, the author will present the uncertainty associated with the estimates of strong ground-motion for engineering designs. The uncertainty can be divided into two types: that associated with strong-motion attenuation models for strong-motion parameters such as response spectra, and that associated with site effects using numerical modelling. The uncertainty from site effect (or site response spectral amplification ratios) can be further divided into two types: record-to-record variability caused by different frequency contents from one earthquake record to another, and soil material properties such as soil shear-wave velocity in numerical modelling for a soil site. Over-simplification in establishing numerical models, such as a 2-dimensional site with buried soft soil layers being modelled as a 1-D site, will also be discussed to demonstrate the extent of this type of error in estimating ground motions for engineering designs. This manuscript does not present a state-of-the-art review of the uncertainties for engineering designs, but gives some results from the author's past and current research projects, with some insights on the amplitude of the uncertainties and how to minimize the effect of these uncertainties. With appropriate estimate of the uncertainties, performance-based engineering design and the use of smart but simple design measures/products would be the key in mitigating the uncertainties in estimating future strong-ground motions for engineering designs.