Recieved:

07/07/2026

Accepted:

22/09/2026

Page: 

–

doi:

http://dx.doi.org/10.17515/resm2026-1834st0707rs

Views:

5

Research on a multi-factor time-varying prediction model for the corrosion rate of reinforcing bars in existing concrete structures

Liying Zhu 1, Weizhong Wang 1, Haiyong Cai 1, Fang Jin 1

1College of Engineering, Huanghe Science and Technology College, ZhengZhou, 450063, China

Abstract

The corrosion of reinforcing bars in existing concrete structures exhibits strong time-varying characteristics under the combined action of multiple factors. Most existing corrosion rate prediction models rely on single-factor or static analyses. They are insufficient for accurately depicting the coupling of multiple factors and the law of time-series evolution, and they also exhibit deficiencies in prediction accuracy and adaptability to actual projects. Starting from the electrochemical nature of rebar corrosion, this paper systematically analyses the mechanisms that induce concrete deterioration during corrosion initiation and development, identifies key influencing variables, including environmental factors, material properties, structural characteristics, and crack evolution, and constructs a computer model of the coupling relationships among multiple factors. In combination with the time-series evolution law of the corrosion rate, a multi-factor features input module and a time-series feature fusion mechanism are established to achieve the in-depth integration of static and dynamic features, thereby forming a time-varying prediction model that considers both mechanism constraints and has data-driven capabilities. The results of multi-factor coupling prediction show that the rate deviation is 8.9 μA/cm² under the weak-coupling condition, the model feature matching degree is 67.5%; the average rate is 72.3 μA/cm² under the medium-coupling condition, the fitting degree of time-varying features is 49.1%; the fitting accuracy under strong coupling is 88.7%, the proportion of load – coupling influence is 23.6%, the comprehensive prediction relative error is 59.8‰, and the global prediction accuracy of the model reaches 97.4%, indicating a stable time-varying prediction effect.

Keywords

Reinforcing bar corrosion; Multi-factor coupling; Time-varying prediction model; Concrete deterioration; Durability assessment

Cite this article as: 

Share This Article
LinkedIn
X
Facebook
journal cover
News & Upcoming Events