Meaning
Mathematical algorithm used in surface metrology separates the long-wavelength waviness from the short-wavelength roughness of a measured profile. When evaluating the finish of a injection-moulded lens or textured plastic part, a gaussian l-filter establishes the mean line against which the surface roughness is calculated. This filter uses a symmetric probability density function to average the data points, which minimizes distortion at the edges of the measurement scan.
Utilizing this filter ensures that the moulder can verify that the textured surface meets the required gloss and friction specifications, which prevents cosmetic rejects.
Cutoff Frequency
Selected cutoff length of the filter determines which surface features are classified as roughness and which are classified as waviness. If the cutoff is set too wide, the high-frequency surface roughness is grouped with the macro-waviness of the mould insert, leading to incorrect roughness readings. Moulders must configure this parameter according to standard surface standards to ensure reproducible inspection results.
Data Distortion
Boundary effects can occur at the start and end of the measured profile during the filtering calculation. A gaussian l-filter minimizes these errors by applying a weighting function, though it still requires a run-up and run-down length that reduces the usable measurement area. This requires the technician to take a scan that is longer than the actual area of interest.
Metrology Standard
Quality control laboratories rely on this digital filtering technique to compare moulded part finishes directly with the CAD model or master insert. It provides a standardized method to assess whether tool wear has occurred over a long production run. This allows the moulder to plan maintenance before parts drift out of tolerance.