Instytut Podstawowych Problemów Techniki

Streszczenie:

In this paper, an assessment of a topography measurement method for fracture surfaces of 10HNAP steel after bending-torsion fatigue tests was performed. Surface roughness was measured by using a non-contact Focus Variation Microscopy (FVM) technique in which the non-measured points (NMPs) and outliers (spikes) were removed by the application of general methods. The results revealed, that the optical measurement method introduced variations in the high-frequency errors, considered as noise within the selected bandwidth. Therefore, the minimization of the high-frequency noise (HFN) was proposed based on an extensive examination of ISO 25178 roughness parameters. Additionally, a general S-filter was applied, as recommended by international standards and commercial software. It was used to identify and remove noise from the measured data after pre-processing. Consequently, levelling and eliminating of NMPs and spikes was successfully performed. Subsequently, the results obtained by using various filters were compared to further assess the impact of different filtration bandwidths. Finally, the proposed procedure was validated by implementing different general functions, such as autocorrelation (ACF), power spectral densities (PSD), and texture direction (TD). It was concluded, that coupled characteristics, including profile and areal measurements, should be studied simultaneously since they are necessary to analyze the fracture surfaces comprehensively.

Słowa kluczowe:

bending-torsion fatigue ,surface topography ,roughness ,non-measured points ,outliers ,high-frequency noise,measurement errors

Afiliacje autorów: