DEVELOPMENT OF A LOW-COST PROTOTYPE FOR PREDICTIVE MAINTENANCE APPLICATIONS BY VIBRATION ANALYSIS

Resumo

Rotary machines typically have vibrations which, due to deterioration and lack of proper
maintenance, cause equipment failure. Predictive maintenance by vibration analysis has the task of
evaluating the operating conditions from global root mean square (RMS) velocity measurements,
comparing them with standards established by technical norms. It is possible to define the causes of
vibrations from the frequencies. Characteristics observed in the frequency spectrum. Nowadays the
conventional equipment used for these analyzes has a high cost and often have a very complex user
interface. In this context, this work proposes the development of a low-cost instrument for vibration
analysis, as this is the main barrier to the application of this technique. The prototype should be able to
acquire system acceleration from this data, calculate instantaneous speeds and RMS speed within a fixed
time frame, the latter being sent to an online server allowing the evolution of vibration severity to be
monitored remotely. By the user. The data obtained will be recorded on a micro SD card so that you can
perform more detailed analyzes with a computer, such as frequency domain spectrum analyzes. To assist
in the analysis, a software with an interactive interface capable of importing the data from the generated
text files, performing the fast Fourier transform (FFT) and plotting the graph for analysis was developed.
For validation of vibration analysis, an Agilent® benchtop analyzer was used. In the calibration tests
with the commercial model, the low-cost prototype was able to identify the defects, unbalances and
misalignments, provided in a rotary system, compared to the Agilent® analyzer. RMS acceleration and
25% for speed. Even with the errors obtained in the measurements, the developed prototype proved to
be effective for the application of low-cost predictive maintenance, since besides being able to identify
the defects in the rotating systems, it can remotely follow the evolution of the vibration severity.