Application of filtering strategies for 3D striation spacing measurements

Quantifying the spacing of fatigue striations on a fracture surface offers a direct, post-mortem method for estimating fatigue crack growth (FCG) rates. Here, a 3-dimensional (3D) surface metrology approach is utilised on reconstructed scanning electron microscope (SEM) images of striated fracture surfaces to assess true striation spacings. 3D surface metrology provides the raw topographic data, although isolating these specific features from the multi-scale profile remains a significant metrological challenge. This paper investigates the limitations of conventional filtering strategies for this application. It is demonstrated that analyses performed on unfiltered primary profiles produce skewed mean spacing values due to the inclusion of long-wavelength form and waviness. Furthermore, the use of standard, fixed cut-off filters is shown to be unreliable, leading to inconsistent measurements from either over-filtering (loss of feature data) or under-filtering (inclusion of unrelated waviness). To overcome these limitations, an accurate and robust methodology is reported. The R Sm parameter (mean spacing of roughness profile elements) was first identified as a suitable proxy for striation spacing. A novel, dynamic filtering strategy was then developed to isolate the striation-dominant roughness profile, setting the filter’s cut-off wavelength (λ c ) based on the evaluation length (L) and the number of striations (N s ) present. Application of this adaptive method yields R Sm values that are robust and consistent, independent of the chosen evaluation length. These measurements reveal a clear, positive correlation between R Sm (striation spacing) and increasing ΔK in strong agreement with Paris’ Law. This work establishes a validated, adaptive filtering technique that provides a reliable pathway for linking 3D surface metrology to FCG rate estimation.