Multi-Field Coupled Degradation and Life Prediction of Low-Temperature Corrosion-Resistant Steel: A Physics-Informed Neural Network (PINN) Approach

In cold regions, steel highway bridges face severe risks of low-temperature embrittlement and corrosion degradation under the coupled effects of extreme cold, freeze–thaw cycles, and corrosive media. Conventional periodic inspections and single-parameter monitoring fail to capture the nonlinear material degradation process throughout the entire service life. Taking the novel low-temperature corrosion-resistant steel (HRB400ERE) used for bridge decks as the research object, this paper proposes a full-life-cycle intelligent monitoring and service life prediction framework integrating multi-source sensing and physics-informed neural networks (PINNs). First, low-temperature Charpy impact and alternate immersion corrosion tests confirmed that the novel steel exhibits impact energies of 59.6 J (φ16 mm) and 57.5 J (φ20 mm) at −20 °C, both >27 J, with a DBTT below −25 °C and 72 h relative corrosion rates of 42.5% (<70%), meeting the intended service requirements. Using these measured physical parameters as boundary constraints, an “environment–corrosion–mechanics” coupled digital twin degradation model was established. To overcome the bottleneck of scarce long-term monitoring data in cold regions, the PINN model deeply integrates Faraday’s law of electrolysis, Arrhenius temperature kinetics, and sparse measured data to dynamically predict the trajectories of material cross-section thinning and low-temperature toughness degradation. Results demonstrate that the model achieves a root mean square error (RMSE) of less than 5% in long-term life prediction. This successfully realizes the transition from “passive inspection” to “active predictive maintenance” for low-temperature corrosion-resistant steel in cold regions, providing a theoretical basis for the intelligent operation and maintenance of transportation infrastructure in extreme environments.

Authors

Institutions

Publication Details

Journal
Metals
Published
2026-11-18
DOI
https://doi.org/10.3390/met16111265
Primary Topic
Structural Integrity and Reliability Analysis
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Multi-Field Coupled Degradation and Life Prediction of Low-Temperature Corrosion-Resistant Steel: A Physics-Informed Neural Network (PINN) Approach

Lubin Zhao, Xiaodong Hao, Dandan Xue, Yi Li et al.
6 links
Metals
Structural Integrity and Reliability Analysis
article

Multi-Field Coupled Degradation and Life Prediction of Low-Temperature Corrosion-Resistant Steel: A Physics-Informed Neural Network (PINN) Approach

Lubin Zhao, Xiaodong Hao, Dandan Xue, Yi Li, Guanghao Yang, Yating Wang
article en

Abstract

In cold regions, steel highway bridges face severe risks of low-temperature embrittlement and corrosion degradation under the coupled effects of extreme cold, freeze–thaw cycles, and corrosive media. Conventional periodic inspections and single-parameter monitoring fail to capture the nonlinear material degradation process throughout the entire service life. Taking the novel low-temperature corrosion-resistant steel (HRB400ERE) used for bridge decks as the research object, this paper proposes a full-life-cycle intelligent monitoring and service life prediction framework integrating multi-source sensing and physics-informed neural networks (PINNs). First, low-temperature Charpy impact and alternate immersion corrosion tests confirmed that the novel steel exhibits impact energies of 59.6 J (φ16 mm) and 57.5 J (φ20 mm) at −20 °C, both >27 J, with a DBTT below −25 °C and 72 h relative corrosion rates of 42.5% (<70%), meeting the intended service requirements. Using these measured physical parameters as boundary constraints, an “environment–corrosion–mechanics” coupled digital twin degradation model was established. To overcome the bottleneck of scarce long-term monitoring data in cold regions, the PINN model deeply integrates Faraday’s law of electrolysis, Arrhenius temperature kinetics, and sparse measured data to dynamically predict the trajectories of material cross-section thinning and low-temperature toughness degradation. Results demonstrate that the model achieves a root mean square error (RMSE) of less than 5% in long-term life prediction. This successfully realizes the transition from “passive inspection” to “active predictive maintenance” for low-temperature corrosion-resistant steel in cold regions, providing a theoretical basis for the intelligent operation and maintenance of transportation infrastructure in extreme environments.

MetalsVol. 16(11)
Automation Research and Design Institute of Metallurgical Industry (China) (CN)
Industry, innovation and infrastructure
Structural Integrity and Reliability Analysis
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon