Stress, Strain, or Energy - Which One Is the Superior Parameter to Estimate Fatigue Life of Notched Components? - An Answer by a Novel Machine Learning-Based Framework

Aug 5, 2026

This paper presents a framework to predict the fatigue lifetime of notched components by applying multiple machine learning algorithms across diverse materials, loading conditions, notch geometries, and fatigue lives. It compares traditional empirical measures based on stress, strain, or energy by examining how the gradients of these mechanical properties differentiate notch geometries. The approach is validated on isotropic materials, then extended to additively manufactured samples and carbon fiber laminated composites. Gradient Boosting and Random Forest perform best. Accuracy improves further by adding new data generated via the Basquin equation.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages36

Reading Time1 h 31 min

AccessFree to download

Abstract

Introduction

Machine learning for fatigue life estimation

Applications and recent advances

Proposed framework (learning-based)

Property gradients for notch geometry identification

Validation across material sets

Stress, Strain, or Energy: Which One Is the Superior Parameter to Estimate Fatigue Life of Notched Components? An Answer by a Novel Machine Learning-Based Framework A. M. Mirzaei 1 Department of Structural, Geotechnical and Building Engineering, Politecnico di Torino, Corso Duca degli Abruzzi 24, 10129 Torino, Italy Abstract This paper introduces a simple framework for accurately predicting the fatigue lifetime of notched components by employing various machine learning algorithms applied to a wide range of materials, loading conditions, notch geometries, and fatigue lives. Traditional approaches for this task have relied on empirical relationships involving one of the mechanical properties, such as stress, strain, or energy. This study goes further by exploring which mechanical property serves asa better measure. The key idea of the framework is to use the gradient of the mechanical properties (stress, strain, and energy) to distinguish between different notch geometries. To demonstrate the accuracy and broad applicability of the framework, it is initially validated using isotropic materials, subsequently applied to samples produced through additive manufacturing techniques, and ultimately tested on carbon fiber laminated composites. The research demonstrates that the gradient of all three measures can be effectively employed to estimate fatigue lifetime, with stressbased predictions exhibiting the highest accuracy. Among the machine learning algorithms investigated, Gradient Boosting and Random Forest yield the most successful results. A noteworthy finding is the significant improvement in prediction accuracy achieved by incorporating new data generated based on the Basquin equation. Keywords: Fatigue life prediction; Notch; Machine learning; Additive manufacturing; Carbon fiber laminated composites 1Corresponding author. Tel.: +39 011 090 4910; Fax: +39 011 090 4899. E-mail address: amir.mirzaei@polito.it 1. Introduction Estimating fatigue life is a vital task in the design and maintenance of various mechanical systems and structures. This process involves predicting how many cycles (life) a material can endure before failure due to cyclic loading, a phenomenon commonly encountered in real-world structures. Traditionally, fatigue life estimation has relied on empirical relationships derived from extensive experimental testing, focusing on mechanical properties like stress, strain, or energy. Therefore, these methods can be time-consuming, expensive, and may lack precision for diverse materials or loading conditions. Recently, machine learning (ML) has emerged as a leading solution for fatigue life estimation. It offers the potential to enhance prediction accuracy by identifying patterns within existing datasets, revolutionizing the field. This section provides areview of the current state of the art in fatigue analysis through machine learning, initially exploring its broad applications across various problems and materials, then narrowing down to its use in additively manufactured samples, and ultimately presenting advanced models to set the context for the problem statement. ML techniques have significantly improved the predictive accuracy of fatigue life estimation of various materials, as evidenced by several compelling applications. For instance, hybrid physicsinformed and data-driven models (HPDM) have proven to be effective, laying a methodological foundation further substantiated by the successful application of the XGBoost model in predicting the fatigue life of high-strength bolts [1] . This approach has also been extended to the prediction of low-cycle fatigue life for lead-free solders, demonstrating ML's reliability across various fatigue conditions [2] . Another illustrative example is the development of a knowledge-based ML framework that integrates empirical formulas with data-driven models, resulting in improved accuracy and resource efficiency [3] . The versatility of ML is further

FAQs

What is the main goal of the proposed framework?

To accurately predict the fatigue lifetime of notched components using machine learning models trained on data spanning materials, loading conditions, notch geometries, and fatigue lives.

How does the framework decide which mechanical property is the better parameter?

It uses the gradients of stress, strain, and energy to distinguish among different notch geometries and then evaluates which property-based predictions are most accurate.

Which machine learning algorithms and data augmentation method show the best results?

Gradient Boosting and Random Forest yield the most successful predictions, and prediction accuracy improves significantly when new data generated using the Basquin equation is incorporated.

Aug 5, 2026

Upload by Lucas Martin