During the contact between metallic surfaces in mechanical systems, some dynamic characteristics present significant changes as the vibration signal patterns. The acquisition and characterization of these signals, by the use of Fourier Transform (FT), are non-intrusive tools in the evaluation of lubricity. This study aimed to evaluate dynamically the fuels lubricity (diesel S50, biodiesel and its blends) by the reciprocating sliding of a ball against a flat disc (AISI 52100 steel). After that, the vibration signals were evaluated by statistical test and FFT (Fast Fourier Transform) and STFT (Short-Time Fourier Transform) analyses and verified their associations with parameters supplied by ASTM D6079. The results present a correlation between lubricity, frequency spectrum and the parameters of profiles from the worn disc surface. In addition, the evolution of wear scars presented an influence of the biodiesel content in these fuels.

In this work, the electrical signals obtained by application of microwaves in chemical and bone tissues are analyzed and classified using techniques of signal processing and pattern recognition. For this, Wavelet Transform is applied as a method to extract relevant features of signal and KNN is used as a classification technique. The results showed that microwave signals can be analyzed using Wavelet Transform, which can be used to reconstruct the signals with minimal error rate and KNN showed satisfactory results.