Improved maximum likelihood method for P-S-N curve fitting method with small number specimens and application in...

Area of Science:

• Materials Science
• Mechanical Engineering
• Fatigue Analysis

Background:

• Accurate fatigue P-S-N (probability-S-N) curve fitting is crucial for structural integrity.
• Limited specimen availability poses significant challenges in fatigue data analysis, particularly for stress-standard deviation relationships.

Purpose of the Study:

• To propose a sample information reconstruction method to address P-S-N curve fitting issues with small datasets.
• To introduce an improved maximum likelihood method for enhanced P-S-N curve fitting.

Main Methods:

• Development of a sample information reconstruction technique.
• Application of the life equivalent principle.
• Implementation and comparison of the improved maximum likelihood method against least square, standard maximum likelihood, BS7608, and IIW standards.

Main Results:

• The improved maximum likelihood method demonstrated effective P-S-N curve fitting with limited fatigue test data.
• Fitted P-S-N curve parameters for 99.9% survival probability using the improved method closely matched established standards (BS7608, IIW).

Conclusions:

• The proposed sample information reconstruction and improved maximum likelihood method offer a superior approach for P-S-N curve fitting when dealing with a small number of fatigue test specimens.
• This method enhances the reliability of fatigue life prediction in materials like Q450NQR1 steel under limited testing conditions.