Estimation of a Minimum Allowable Structural Strength Based on Uncertainty in Material Test Data.

Area of Science:

• Engineering
• Materials Science
• Statistics

Background:

• Estimating the minimum fracture strength of full-scale components involves three key uncertainties: model selection, laboratory-scale strength, and full-scale strength.
• Accurate estimation is crucial for reliable engineering design and safety.

Purpose of the Study:

• To develop and present a novel three-step approach for estimating the minimum strength of full-scale components.
• To quantify uncertainties in statistical modeling and strength prediction for engineering applications.

Main Methods:

• A composite goodness-of-fit metric was used to select the best statistical distribution (from five candidates) for laboratory test data.
• Parameters of selected distributions were estimated with uncertainty, and laboratory-scale minimum strength was determined at a 95% confidence level.
• Full-scale allowable minimum strength was estimated using coverage concepts (99% A-basis and 90% B-basis) at a 95% confidence level.

Main Results:

• The developed uncertainty-based approach was applied to six sets of failure strength data from four engineering materials.
• The method successfully quantified uncertainties across model selection, parameter estimation, and full-scale strength prediction.
• Results demonstrated the applicability of the approach to diverse engineering materials.

Conclusions:

• The proposed three-step method provides a robust framework for estimating full-scale component strength with quantified uncertainties.
• This approach enhances the reliability of design allowables, particularly for critical aerospace components.
• The study highlights the significance of uncertainty quantification in materials strength prediction.