Dissertação
Ligações semirrígidas na análise dinâmica de estruturas metálicas treliçadas
Autor
Grotto, Luiz Guilherme
Institución
Resumen
In Brazil, lattice metalic towers are dimensioned to support their own weight, weight of cables or equipments that they have and mainly Wind load. But, even low-intensity dynamic actions can bring such a structure to collapse when excitation frequency is very close to one of tower's natural vibration frequencies. When this occurs, structure goes into resonance and amplitudes of displacements increase significantly. One of ways of controlling amplitudes and prevent a structure from collapsing is to increase its damping, so that energy induced by dynamic load can be dissipated more quickly. In order to evaluate response of structure to these loads, a mechanical model that represents reality is necessary. This work presents a numerical and experimental study about subject, evaluating dynamic properties of a lattice structure of aluminum. Natural frequencies and damping are evaluated experimentally by applying impacts to model. Flexibility of frame connections is changed by changing torque applied to screws and by inserting rubber rings between profiles that make up model. Accelerometers and computer programs are used to measure, receive, and process dynamic response of structure. In this way, first natural frequencies of vibration and damping behavior of structure are determined, latter calculated by logarithmic decrement. Same properties are evaluated in a computational numerical model, where same conditions of experimental stage are applied with aid of a finite element program and a direct integration program using central finite differences. It was observed that introduction of flexible elements increased damping rate of structure to the point of reducing natural frequency of vibration. In addition, it was identified that change in torque did not cause a significant change in dynamic properties of structure without rubber rings in connections, since no slip was observed between profiles.