Overview
ISO/TR 16610-32:2023 is part of the ISO 16610 series focusing on geometrical product specifications (GPS), specifically related to filtration methods applied to surface and profile analysis. This technical report addresses robust profile filters using spline filters. It provides detailed guidance on both linear spline filters for uniform and non-uniform sampling, and robust spline filters for processing surface profiles, expanding on concepts introduced in ISO 16610-22, ISO 16610-30, and ISO 16610-31.
Spline filters are essential in modern metrology for analyzing and managing surface texture data, especially for handling measurement data with outliers or non-Gaussian distributions. Their robust variants offer enhanced resistance to noise and surface anomalies, improving the reliability of extracted waviness and roughness characteristics.
Key Topics
- Spline Filters: Filters based on piecewise polynomials (splines), effective for both uniform and non-uniform sampled data. Linear spline filters are especially useful for profiles where measurement points are not evenly distributed.
- Robust Estimation: Application of robust statistical estimators to spline filters, making them less sensitive to outliers in the input data. This is key when analyzing surfaces with irregularities or noise.
- Tukey’s Biweight Influence Function: The robust spline filter utilizes the Tukey biweight function as the weight mechanism in its M-estimator, suppressing the impact of extreme values and supporting the extraction of meaningful profile and waviness data.
- Filter Parameters: Detailed explanation of regularization and tension parameters, enabling control over the smoothing and shape retention capabilities of the filter.
- Objective Functions: The document discusses the formulation of optimization problems for spline filters, both with L2-norm (standard least-squares) and L1-norm (for enhanced outlier resistance).
- Profile and Sampling: The standard covers both profiles with equidistant (uniform) and non-equidistant (non-uniform) sampling points, a practical consideration for real-world measurement scenarios.
- Filter Designation: Defines symbols for linear spline filters (FPLS) and robust spline filters (FPRS), ensuring standardized communication within the GPS framework.
Applications
ISO/TR 16610-32:2023 is highly relevant in industrial and scientific metrology, particularly in surface texture analysis and geometrical specification of products. Key application areas include:
- Surface Texture Measurement: Useful in characterizing surface finish on machined, abraded, or otherwise processed components, where robust filtering ensures accurate separation of roughness and waviness.
- Quality Control: Spline filters support reliable surface profile assessments for conformity with specified product standards, vital in high-precision manufacturing sectors such as aerospace, automotive, and medical device industries.
- Research and Development: Ideal for developers of surface metrology software and advanced measurement instruments, enabling the integration of robust and flexible filtering techniques.
- Data Interpretation: When analyzing profiles with outliers or irregular distributions, the robust spline filter offers improved extraction of surface parameters, supporting better material ratio and performance evaluations.
Related Standards
Understanding and implementing ISO/TR 16610-32:2023 is complemented by familiarity with other relevant GPS filtration standards, including:
- ISO 16610-1: Overview and basic concepts for filtration methods.
- ISO 16610-22: Linear profile filters using spline filters.
- ISO 16610-30: Robust profile filters - basic concepts.
- ISO 16610-31: Robust profile filters based on Gaussian regression.
- ISO 14638: GPS matrix model, providing the organizational structure for the ISO GPS system.
- ISO 8015: General GPS principles.
- ISO 14253-1: Decision rules for verifying conformity.
These standards together form a comprehensive framework for surface texture filtration and geometrical product specification in modern manufacturing and metrology.
Keywords: ISO/TR 16610-32, robust spline filter, geometrical product specifications, GPS, surface texture, profile filter, spline filter, Tukey biweight, metrology, surface roughness, non-uniform sampling, industrial filtration standard.