摘要：Mechanical wear has its origin in the friction of surfaces under contact and it is characterized for low normal forces by the mean coefficient of friction (COF) test on tribometers. The roughness of the worn surfaces is well established that is independent of the wear. Nevertheless a main hypothesis of this work is the research for evidences of the friction test variability influence on the surface roughness of worn materials in the test. The experimental setup includes the COF test ball-on-flat with steel on titanium and the surface roughness measured with an optical profilometer (confocal microscope) is analyzed. The space domain study of roughness parameters is developed on profiles (1D) and the surface (2D). The correlation study of the variability of the COF test with roughness metrics seeks the most significant roughness parameters. Its coherence and significance to respect the friction process are discussed. The similarities in the surface topography and its link with the COF signal are analyzed in the frequency domain based on power spectral analysis (PSA) from the fast Fourier transformation (FFT). Next, the whole profiles are also decomposed in frequency ranges by wavelet filtering that reveal the traces of the COF test on the surfaces and their similarities in the frequency domain.
