Overview
ISO 12107:2012 - "Metallic materials - Fatigue testing - Statistical planning and analysis of data" provides standardized methods for planning fatigue experiments and for statistically analysing fatigue test results so that material fatigue properties can be determined with high confidence using a practical number of specimens. The standard targets metallic materials and covers both fatigue life (N) at a given stress (S) and fatigue strength (S) at a given life. It emphasises reliable experimental design, appropriate statistical models (e.g., log‑normal life distributions), and reproducible reporting of S‑N curves and tolerance limits.
Key technical topics and requirements
- Statistical planning of fatigue tests
- Sampling strategies and allocation of specimens across stress levels to obtain representative data with efficient specimen counts.
- Distributions and modelling
- Common assumptions: logarithm of fatigue life often follows a normal (log‑normal) distribution; fatigue strength at a given life is often normally distributed. Alternative models (e.g., Weibull) are acknowledged for skewed cases.
- The standard includes methodology for the Bastenaire model and other regression approaches for S‑N behaviour.
- Data analysis methods
- Plotting on normal probability paper and estimating distribution parameters (mean, standard deviation).
- Quantitative tests for normality and estimation of lower tolerance limits for life or strength at chosen confidence levels.
- Procedures for estimating regression parameters of the S‑N curve, calculation of lower tolerance limits, and experimental planning for S‑N curve development.
- Censoring and runouts
- ISO 12107 addresses analysis of runout (censored) data in general, but for the S‑N curve estimation (Clause 8) it applies only to complete (uncensored) data-mixed or multi‑mechanism failure populations are outside scope.
- Reporting
- Prescribed content for the test report, including presentation of results, estimated fatigue strength at given lives, and S‑N curve parameters.
Practical applications and users
ISO 12107:2012 is intended for:
- Fatigue test laboratories establishing material S‑N data and tolerance limits.
- Materials and mechanical engineers evaluating fatigue life or strength for design validation.
- Quality assurance teams comparing batches or processes for consistent fatigue behaviour.
- Researchers developing fatigue-resistant alloys or validating statistical fatigue models.
Practical benefits include optimized specimen counts, statistically defensible design values (lower tolerance limits), and consistent reporting that supports design safety margins and material comparisons.
Related standards
- ISO 3534 (all parts) - Statistics: vocabulary and symbols (normative reference used in ISO 12107).
- ISO 12107 was prepared under ISO/TC 164 (Mechanical testing of metals), and should be used alongside other fatigue and material testing standards as appropriate.
Keywords: ISO 12107:2012, fatigue testing, statistical planning, statistical analysis, S‑N curve, fatigue life, fatigue strength, metallic materials, runout, censored data.