ISO 10928:2024

Overview

ISO 10928:2024 - Plastics piping systems - Glass‑reinforced thermosetting plastics (GRP) pipes and fittings - Methods for regression analysis and their use - defines statistical procedures for analysing time‑dependent test data for GRP pipes and fittings. The standard describes how to work with data that, after logarithmic transformation, follow either a normal or a skewed distribution. It specifies two primary regression methods, guidance on extrapolation to long‑term values (commonly extending ~10 000 h test trends to a 50‑year prediction), and how analysis results are applied to product testing and design. Note: ISO 10928:2024 addresses only data analysis; test methods and sample collection are covered in referring standards.

Key topics and technical requirements

• Logarithmic transformation: Data are analysed as logarithms (lg of property and time) to stabilise variance and linearise trends for regression.
• Two analysis methods:
  • Method A - Covariance method: Designed to accommodate skewed distributions of log‑transformed data (common in destructive tests). Uses covariance calculations, error variances and a gamma parameter to determine best‑fit relationships.
  • Method B - Regression with time as independent variable: Simpler least‑squares approach suited to data whose logs are approximately normally distributed (often used for non‑destructive creep/relaxation tests).
• Statistical checks:
  • Calculation of linear correlation coefficient (r) and use of Student’s t thresholds to determine data suitability for regression and extrapolation.
  • Evaluation of slope variance and a T‑statistic to assess whether extrapolation is justified.
• Confidence and prediction limits: Procedures for lower confidence limits (LCL) and lower prediction limits (LPL) are provided (see Annex B).
• Alternative curve fitting: Annex A includes second‑order polynomial relationships for cases where non‑linear fitting better represents short‑term creep/relaxation behaviour.
• Application guidance: Clause 6 explains using analysis outcomes for product design, comparison to specified values and declaring long‑term properties.

Practical applications and users

ISO 10928:2024 is intended for:

• Test laboratories performing long‑term property tests on GRP pipes and fittings.
• Product designers and mechanical/material engineers using extrapolated long‑term property estimates for design and safety margins.
• Manufacturers declaring long‑term performance values and preparing conformity evidence.
• Standards developers and conformity assessment bodies integrating statistical analysis into product standards. Because the methods are generic, they can also be applied to other materials or test data sets where log‑transformed regression analysis is appropriate.

Related standards

• Referring product standards and test methods for GRP piping (these define specimen preparation, sampling and test durations).
• ISO/TC 138 (Plastics pipes, fittings and valves) publications - for related product requirements and test protocols.

Keywords: ISO 10928, GRP pipes, regression analysis, covariance method, method A, method B, extrapolation, long‑term prediction, statistical testing, confidence limits.