Avaliação da influência da temperatura no método de ISHM aplicado em eixos de material compósito de máquinas rotativas

Abstract

This master’s thesis proposes a methodology capable of increasing the robustness of the ISHM method in order to make it less sensitive to variations in environmental and operational conditions. In this case, two shafts of composite material are investigated considering temperature and operational condition variations. For this aim, the first shaft was subjected to experiments carried out in a climatic chamber for temperature control purposes, thus simulating environmental variations. On the other hand, the second shaft was used in a rotating machine operating at different rotation speeds. In these two experimental tests, a nut was attached to the surface of the shafts in order to simulate the damage conditions. Impedance signatures were measured using an impedance analyzer and an impedance meter. The obtained results demonstrated that the ISHM method was able to detect the considered damages. Additionally, it was found experimentally that variations in temperature and operating conditions caused unwanted changes to the impedance signatures. The optimization process using a hybrid strategy, Differential Evolution (ED) and Nelder-Mead (NM), was not efficient to minimize the influence of these variations throughout the investigated frequency range. This difficulty is associated with the nonlinear behavior of the impedance signatures measured in the composite shafts according to the considered temperature and operational condition variations. Thus, a new compensation methodology was proposed aiming at seeking optimal frequency regions in the measured impedance signatures. The damage condition is evaluated within the optimal frequency range. Thus, the results obtained by the proposed methodology, when compared with the previous one, were more interesting, since the influence of temperature and operational condition variations in the impedance signatures were minimized, avoiding the detection of false positives.