This paper presents two applications of a Probabilistic Seismic Hazard Analysis (PSHA) methodology which is both site- and soil-specific. The ground motion hazard is evaluated at the surface either by oscillator-frequency-dependent hazard curves for spectral acceleration, S-a(s) (f), or by acceleration uniform hazard spectra associated with a given mean return period. The soil response is represented statistically via multiple nonlinear dynamic analyses of the soil column with uncertain properties.The ground motion estimates presented here are more precise than those which could be found by means of conventional PSHA with standard attenuation laws for generic soil conditions. The use of generic predictive equations may in fact lead to inaccurate results especially for soft soil sites, where significant amplification is expected at long periods, and for saturated sandy sites, where liquefaction or cyclic mobility may be expected for severe levels of ground shaking. Both such cases are considered in this article.

A tunnel excavated by a tunnel boring machine was monitored extensively by means of extensometers installed near the tunnel face. Consequently, the three-dimensional state existing at the time of installation must be considered for the interpretation of the monitoring data. Results from three-dimensional finite element simulations are used to back-calculate rock mass strength and deformation properties. The purpose of this study was to establish and test various approaches of back-analysis. Results are compared with field and laboratory measurements. On the basis of these analyses, the paper provides guidance on how field data can be used for back-analysis purposes even when the ground behaves in a nonelastic manner. Key words: tunnelling, monitoring, tunnel boring, back-analysis, nonlinearity.

Convergence, radial displacements, and stress changes are often recorded during the advance of a tunnel for the observational tunnel design approach. In deep tunnels, instruments must be installed from underground and can seldom be placed in undisturbed ground. Consequently, observations are only partial records of the total change induced by an excavation and the influence of the three-dimensional state near the face must be considered. This paper presents results from numerical simulations to assess face effects on monitoring data. The influence of such aspects as in situ state of stress, anisotropy, nonlinearity, and plasticity (yielding ground) are evaluated. Guidelines for underground monitoring of deformations are given. Key words: tunnelling, monitoring, back-analysis, convergence, extensometers, numerical modelling.

Probabilistic seismic hazard assessments (PSHA) keeping into account amplification effects in non linear soil deposits require a relatively complex, multidisciplinary approach. The purpose of this presentation is to describe the preliminary version of an on-line system (www.eriskzone.net) suitable to define, with a limited effort, the probabilistic seismic hazard at the ground surface accounting for site amplification effects. The software relies on advanced probabilistic procedures and analytical tools for soil amplification assessment, and provides an example on how state of the art procedures could be applied more extensively, even to smaller projects and for multiple analyses throughout large areas where limited information on local soil characteristics is available. At present the system, which is freely available on the Internet and is intended for demonstration purposes only, covers the whole Italian territory and, including some features which will be made available on the web in the near future (see below for details), it presents the following main characteristics: 1) The seismic hazard at the bedrock is contained in a database and cannot be changed by the User. 2) Site amplification is computed based on advanced amplification functions using local parameters as defined by the User. The system as described above has been available on the Internet since November 2005, with the exception of the advanced amplification functions (item 2) which are currently being tested and will be made available shortly replacing the basic amplification criteria adopted in the original version. The seismic hazard at the bedrock is the result of a conventional PSHA carried out throughout the Italian territory over a grid of nodes. It includes pre-defined hazard curves obtained for several oscillation frequencies and magnitude+distance values estimated by a disaggregation procedure. As far as site amplification is concerned, the non-linear effect of the soil layers on the intensity of the ground motion at the surface is captured by a site-specific, frequency-dependent amplification function, AF(f), where f is a generic oscillator frequency. The method exploits the results of thousands of nonlinear dynamic analyses of different types of "pilot" soil columns subject to real rock ground motions. Median AF(f) and related uncertainty can be coupled via convolution with the site-specific bedrock hazard to obtain the surface ground motion hazard for the site (e.g., Bazzurro and Cornell, 2004a and b). Although the system focuses on seismic hazard, it includes also a simplified procedure to predict structural damage based on building characteristics as defined by the User. The purpose is to show that besides hazard other aspects of seismic risk can also be included in the same framework. At present the system does not perform a rigorous convolution procedure to establish structural damage, but simply provides an estimate of the probability of having a pre-defined damage level based on fragility curves obtained from the literature. In the future the structural damage capability could be enhanced by implementing a proper convolution procedure and adopting user-defined fragility curves.