Dynamic Analysis of Pre-Cast RC Telecommunication Towers Using a Simplified Model

Abstract

The main goal of this paper is to propose a simplified procedure to accomplish the dynamic analysis of pre-cast RC (reinforced concrete) telecommunication towers subjected to wind loads. A new procedure, based on graphs and curves obtained using optimization techniques, uses the results of the static analysis to compute the dynamic response of this kind of structures. According to NBR-6123 code (ABNT, 1988), if the first natural frequency of vibration of a given structure is smaller than 1 Hz, it is necessary to perform dynamic analysis of the structure; otherwise a static model can be used. One peculiar characteristic of these pre-cast RC structures is that they often present the first natural frequency of vibration smaller than 1 Hz and so the dynamic analysis is needed. The main feature researched is the dynamic magnification factor, defined here as the ratio between the bending moment given by the dynamic and static models (ABNT, 1988). Surfaces are created to give the dynamic magnification factor as a function of the structure height and the first natural frequency of vibration. To create these surfaces, optimization problems (inverse problems) were formulated where the objective function is the error between the dynamic magnification factor, computed according to (ABNT, 1988), and other given by equations, defined in function of the structure height and first natural frequency of vibration. The design variables are the coefficients of these equations and constraints are imposed to avoid negative and also very large frequencies. With the methodology proposed here only the static results and the first natural frequency of vibration are needed to accomplish the dynamic analysis of a given structure. The method is easier and faster than the tradition dynamic analysis approach. In this work, results of the dynamic and static analysis of 90 real structures are used in the optimization process. The difference between the results given by the simplified method proposed here and the complete dynamic analysis are less than 2%.