Abstract

ZUNIGA-OSORIO, H.J. et al. Vibrational Modal Model for a Compressor Blade. RIIIT. Rev. int. investig. innov. tecnol. [online]. 2019, vol.7, n.41, pp.23-39.  Epub Feb 05, 2021. ISSN 2007-9753.

The analysis of thermodynamic machinery (cooling systems and gas turbine technologies) exposed at certain set of critical ambient conditions, is always a continuously issue for the mechanical engineering; where the state of the art statement is concerned not just with the dynamics magnitude measurement and the regulation of the process variables into the critical stages from the strategical machine systems, nowadays it also needs to observed and control the spatial distribution of the energetic variables into the volumes of interest. As a first step to address this engineering area of opportunity, and based on the hypothesis that the flow incidence in solid bodies induces energy on them as mechanical vibrations, in this work, a methodology is proposed to get a mathematical model based on the vibrations behavior from a set of blades of a gas turbine compressor; which is based on the lumped mass hypothesis. In the future, with the validation of the gotten model, authors will intend to characterize in-vivo the spatial behavior of the air flow into the ventilation systems grids and/or the flow conditioning stages (expansion or compression) from the rotating machinery, by the measuring and analyzing the vibration in these components; since this is not exposed instrumental system, neither the performance of the concerned machinery, nor the integrity of the sensor device is affected by the energy magnitude and distribution in which the flow insides. The proposed methodology consists of two stages: the first stage is a linear approximation of the model parameters for mass and stiffness from the experimental results of an experimental modal testing based on the impact test, and the second stage is a recursive adjustment of the parameters for damping to reduce the error in the estimation of the resonance frequencies and their corresponding modal shapes. The result of the computational approximation agrees with the experimental results and also by applying the concerned recursive adjustment, the maximum average error has been reduced from 28.3% to 2.5%.

Keywords : Vibrations; Experimental modal testing; Parameters identification algorithms.