Page 60 - IJAMD-2-1

International Journal of AI for
Materials and Design

ORIGINAL RESEARCH ARTICLE
Improvement of multiaxial fatigue life prediction

performance based on contrastive learning
feature extraction

Ziyu Cui , Xingyue Sun * , and Xu Chen 1,3
1
2
1 Department of Process Equipment & Control Engineering, School of Chemical Engineering and
Technology, Tianjin University, Tianjin, China
2 Department of Aeronautical Structure Engineering, School of Aeronautics, Northwestern
Polytechnical University, Xi’an, Shaanxi, China
3 Zhejiang Institute of Tianjin University, Ningbo, Zhejiang, China
(This article belongs to the Special Issue: AI for Multiscale Analysis and Defect Identification in
Packaging Structures and Semiconductor Chips)

Abstract
Accurate prediction of multiaxial fatigue life was crucial for structural integrity
assessment, yet the variability in material responses under complex loading paths
made it challenging for both classical and data-driven models to achieve high
accuracy. To address this issue, a contrastive learning-based framework was proposed
in this study, enabling the construction of more generalized low-dimensional feature
representations across different loading paths. This framework enhanced the robustness
*Corresponding author: of fatigue life prediction without relying on mechanical assumptions. Experimental
Xingyue Sun validation demonstrated that, compared to existing methods, the contrastive learning
(xysun@nwpu.edu.cn) model learned more suitable feature encodings, significantly improving prediction
Citation: Cui Z, Sun X, Chen X. performance. This framework provided a reference solution for engineering applications
Improvement of multiaxial fatigue requiring reliability assessment under multiaxial stress conditions.
life prediction performance based
on contrastive learning feature
extraction. Int J AI Mater Design.
2025;2(1):54-72. Keywords: Contrastive learning; Deep learning; Feature engineering; Life prediction;
doi: 10.36922/IJAMD025040004 Multiaxial fatigue
Received: January 22, 2025
Revised: March 6, 2025
Accepted: March 14, 2025 1. Introduction
Published online: March 28, 2025 In modern high-tech industries such as electronic packaging, aerospace, and nuclear
power generation, the role of multiaxial fatigue analysis has become increasingly
Copyright: © 2025 Author(s).
1-5
This is an Open-Access article critical. Electronic packaging materials are subjected to complex thermal and
distributed under the terms of the mechanical loads, which can precipitate premature material and structural failures,
Creative Commons Attribution thereby severely damaging structural integrity. Consequently, comprehensive research
6,7
License, permitting distribution,
and reproduction in any medium, into multiaxial fatigue is essential for enhancing the reliability and service life of
provided the original work is electronic devices at the design and serving stages. In addition, precise assessments of
properly cited. fatigue behavior under complex stress environments are crucial for ensuring the safe
Publisher’s Note: AccScience operation of aerospace vehicles, nuclear power stations, and other fields. 8-10 Such studies
Publishing remains neutral with not only facilitate a better understanding of material responses under multiaxial stresses
regard to jurisdictional claims in
published maps and institutional but also advance the design of materials and the evaluation of structural integrity, pivotal
affiliations. for the development of safer and more efficient technological solutions.

Volume 2 Issue 1 (2025) 54 doi: 10.36922/IJAMD025040004