VIBRATION CONTROL BY STRUCTURAL COUPLING IN ADJACENT STRUCTURES USING STOCHASTIC ANALYSIS

摘要

The structural coupling technique originated in the USA in the 70's, when Klein et al. [1] connected two adjacent structures through wire ropes to prevent pounding between them. This technique has been shown to be efficient not only to prevent the pounding phenomenon, but also to control the amplitude of vibration of coupled structures and to avoid possible severe structural damage caused by earthquakes or strong winds. The study of random vibrations arose from the need to assess the reliability structures that operate in a random environment. Stochastic dynamic analysis aims to evaluate the response of dynamic systems subject to stochastic inputs. One of the main premises of this analysis is to consider that the structure is known and deterministic, that is, its properties are not subject to random variations. The structure excitation is managed through probabilistic properties, generating the problem response as a function of variances. Thus, this paper aims to evaluate the difference between the deterministic and stochastic response in two adjacent structures, when using the structural coupling technique for vibration control. In addition, a parametric analysis is performed in order to verify the importance of the soil and filter parameters applied in the state matrix, in the stochastic analysis. Adjacent building models are used considering the shear frame structure behavior. The dampers that connect the structures are passive and their positions, quantities and mechanical properties are optimized by the particle swarm algorithm, in which the excitation in deterministic analysis is provided by the acceleration record of the 1940 El Centro earthquake. For stochastic excitation, a Gaussian stationary random process is used, in which it is applied through a filter, applied directly on the coupled system state matrix. The deterministic results of the uncoupled structures indicated the importance of choosing the base acceleration used as excitation in dynamic models. Reductions in vibration amplitude of up to 60% were observed. In the stochastic analysis, the reductions in floor vibration amplitude remained in the 50% range both structures. It has been seen that there are considerable differences when using deterministic analysis and stochastic analysis. However, the use of coupling structural control technique increases the system reliability.