Ricardo Taborda

The dynamic response of a recently instrumented building in Acapulco City is studied. The objective is to know its dynamic properties and the characteristics of the soil-structure interaction to which it is subjected. Additionally, it is sought to establish a base which serves for later works, such as the elaboration of both linear and non-linear models of the building, or comparative studies with similar investigations carried out on instrumented buildings such as those in Mexico City.

For that purpose, the signals recorded during eight seismic events occurred between September, 2001 and April, 2002 are analyzed, along with the acceleration time histories of two ambient vibration tests made with the same accelerometric network.

The study of these signals is carried out through both parametric and non-parametric system identification techniques. Among the non-parametric methods, a conventional Fourier analysis is performed, by means of which the frequencies, modal damping values and modal shapes of the system are estimated.

This process is supplemented with a simplified method which allows to estimate the soil-structure interaction characteristics, understanding these as the frequencies and stiffnesses associated with the motions of the structure as well as the rocking, translational and torsional motions of the base.

The method of zero crossings is also applied for the analysis of the records in the time domain. With this method it is possible to identify the fundamental frequency of the system and additionally, by filtering the signals, the potential of the method to identify the properties of the torsional, rocking and translational movements of the base is examined.

In what concerns to parametric methods, a procedure that is based in the concepts of modal superposition is applied, by means of which the properties of the system, the structure and the rocking component of motion are determined through a least squares fit of the recorded and calculated signals, starting from models which idealize the system.

In addition to the simplified method, and looking forward to finding the most appropriate representation of the foundation system, its stiffnesses are estimated by using the computer program Dyna5 for a wide variety of models and by following the procedure proposed in the preliminary version of the Complementary Technical Standards for Earthquake Design of the Federal District Construction Regulation that, although not applicable in Acapulco City, constitute a guide for the design procedures used in practice.

The results obtained by the different methods are compared. Their analysis allows to highlight the strengths and limitations of each one, as well as to obtain interesting conclusions about the properties of the building and the soil-structure interaction effects identified.

Among the most interesting results, it can pointed out the possible appearance of non-linear behaviors in the system, which, given the identified characteristics, seem to be attributable to the motions due to soil-structure interaction, especially to the rocking component which reaches up to thirty percent of participation in the total response of the system. This motion was widely studied in both the Fourier analysis and the modal parametric technique, in which the use of filtered signals is explored.

Finally, it must be highlighted that in the estimation of the stiffnesses of the soil-structure system, it was found that the theoretical consideration of the pile-soil-pile group effects may lead to important differences with respect to the results obtained experimentally with the simplified method.