A low-cost data acquisition system for dynamic structural identification

This work aims at the evaluation of an alternative data acquisition system to identify dynamics properties of structures, using low-cost devices with on-board micro-electro-mechanical systems (MEMS) accelerometers. The proposed acquisition system, using the Arduino platform, specifically an Arduino/Genuino 101 board, presented good results in the identification of natural frequencies of the structural models analyzed. It showed a maximum difference when compared to a professional accelerometer of 2.41%. In addition, for one of the models, by using two Arduino boards it was also possible to identify the mode shapes, with acceptable difference of 4.30% on natural frequencies, when compared to professional accelerometers. Models with frequency range between approximately 1.00 and 37.83 Hz were investigated and the results were considered promising. Thus, since the natural frequencies of conventional civil structures are close to or within this range, the proposed system demonstrates the ability to acquire the acceleration time series required for dynamic structural identification methods.