On the characterisation of steel corrosion and the resulting concrete damage using tomography
Doctoral thesis, 2025
Steel corrosion is a major cause of deterioration in reinforced concrete structures. As the process initiates and progresses from within the structure, it remains largely hidden behind the concrete cover, making the assessment of internal damage challenging without invasive or destructive methods. An improved understanding and predictive capability of the internal processes over time is required to improve inspection methods based on observations at the surface. In this research, the internal characteristics of steel corrosion and the mechanisms that lead to damage in reinforced concrete were investigated at the material scale using experimental methods that allow internal processes to be monitored within the specimen.
The overall aim was to characterise steel corrosion and the resulting damage in reinforced concrete at the material scale using X-ray and neutron computed tomography, within the context of accelerated corrosion experiments. The focus was on extracting quantitative information from the tomography data. A subset of the image-derived characteristics were then integrated into a finite element model to gain additional insight into the mechanical effects of corrosion. In addition, other complementary techniques were employed, including distributed optical fibre sensing, electrical resistance measurements and chemical analysis, for monitoring processes that tomography alone could not capture.
The work demonstrated that tomography could be used to quantify a broad range of corrosion and damage-related characteristics. These include the size and spatial distribution of interfacial voids, corrosion penetration depth, corrosion morphology, the volumetric expansion coefficient of corrosion products and the volumetric strain and normal stress in the corrosion layer. Moreover, 3D deformations in the cementitious matrix were measured through local digital volume correlation.
A key finding was the identification of a spatial correlation between larger interfacial voids and pitting corrosion, highlighting void size as a critical factor influencing localised attack. In addition, the volumetric expansion coefficient of corrosion products was found to be close to four, consistent with previous image-based studies. Furthermore, the mechanical response in the corrosion layer indicated a non-linear behaviour of corrosion products.
Together, the findings demonstrate how experimental imaging and numerical modelling can be combined to advance the understanding of corrosion-induced damage in reinforced concrete, providing insights that neither approach could provide on its own.
The overall aim was to characterise steel corrosion and the resulting damage in reinforced concrete at the material scale using X-ray and neutron computed tomography, within the context of accelerated corrosion experiments. The focus was on extracting quantitative information from the tomography data. A subset of the image-derived characteristics were then integrated into a finite element model to gain additional insight into the mechanical effects of corrosion. In addition, other complementary techniques were employed, including distributed optical fibre sensing, electrical resistance measurements and chemical analysis, for monitoring processes that tomography alone could not capture.
The work demonstrated that tomography could be used to quantify a broad range of corrosion and damage-related characteristics. These include the size and spatial distribution of interfacial voids, corrosion penetration depth, corrosion morphology, the volumetric expansion coefficient of corrosion products and the volumetric strain and normal stress in the corrosion layer. Moreover, 3D deformations in the cementitious matrix were measured through local digital volume correlation.
A key finding was the identification of a spatial correlation between larger interfacial voids and pitting corrosion, highlighting void size as a critical factor influencing localised attack. In addition, the volumetric expansion coefficient of corrosion products was found to be close to four, consistent with previous image-based studies. Furthermore, the mechanical response in the corrosion layer indicated a non-linear behaviour of corrosion products.
Together, the findings demonstrate how experimental imaging and numerical modelling can be combined to advance the understanding of corrosion-induced damage in reinforced concrete, providing insights that neither approach could provide on its own.
Tomography
Corrosion Characteristics
Steel Corrosion
Corrosion-Induced Cracking
Finite Element Modelling
Reinforced Concrete
SB-H6, Sven hultins gata 6, Chalmers
Opponent: Associate Professor Els Verstrynge, Department of Civil Engineering, KU Leuven, Belgium
Author
Andreas Alhede
Chalmers, Architecture and Civil Engineering, Structural Engineering
A two-stage study of steel corrosion and internal cracking revealed by multimodal tomography
Construction and Building Materials,;Vol. 394(2023)
Journal article
Monitoring corrosion-induced concrete cracking adjacent to the steel-concrete interface
Materials and Structures/Materiaux et Constructions,;Vol. 56(2023)
Journal article
Characterisation of steel corrosion and matrix damage in reinforced mortar combining analytical, electrical and image-based techniques
Cement and Concrete Research,;Vol. 190(2025)
Journal article
Alhede, A., Dijkstra, J., Lundgren, K. Linking image-based corrosion characterisation to the mechanical response in reinforced mortar
How can we inspect what we cannot see? Steel corrosion is one of the main reasons why reinforced concrete structures deteriorate over time. The process starts invisibly, inside the concrete, making it difficult to assess without removing the protective concrete cover. This poses a major challenge when assessing damage and planning targeted repair strategies, especially when the goal is to avoid invasive and destructive methods.
In this research, advanced 3D imaging techniques; X-ray and neutron computed tomography, were used to look inside concrete specimens and study how corrosion develops and affects the material from within. These techniques are primarily suited for small-scale samples and are typically performed in laboratory environments. Still, they made it possible to observe internal processes in great detail, without damaging the samples.
The goal was to better understand what happens inside the concrete when steel corrodes. By studying corrosion in detail under controlled conditions, this knowledge is intended to contribute to future efforts to assess corrosion in full-scale structures. Using the 3D image data, it was possible to measure the loss of steel cross-sectional area, track how this loss was distributed along the steel surface, and observe how the corrosion products expanded and exerted pressure on the surrounding concrete. Other tools, like distributed fibre-optics and chemical analysis, were also used to monitor processes that imaging alone could not capture.
One important finding was that large air voids at the steel–concrete interface significantly increase the risk of corrosion initiation and propagation in those areas. By combining experimental observations with computer modelling, the study also revealed how corrosion products behave under pressure: how they expand, compress and interact with the surrounding concrete. This made it possible to estimate mechanical effects that had not previously been measured directly. Together, these methods offer a new way to study corrosion -- one that in future could help engineers detect hidden corrosion damage earlier and design concrete structures that last longer.
In this research, advanced 3D imaging techniques; X-ray and neutron computed tomography, were used to look inside concrete specimens and study how corrosion develops and affects the material from within. These techniques are primarily suited for small-scale samples and are typically performed in laboratory environments. Still, they made it possible to observe internal processes in great detail, without damaging the samples.
The goal was to better understand what happens inside the concrete when steel corrodes. By studying corrosion in detail under controlled conditions, this knowledge is intended to contribute to future efforts to assess corrosion in full-scale structures. Using the 3D image data, it was possible to measure the loss of steel cross-sectional area, track how this loss was distributed along the steel surface, and observe how the corrosion products expanded and exerted pressure on the surrounding concrete. Other tools, like distributed fibre-optics and chemical analysis, were also used to monitor processes that imaging alone could not capture.
One important finding was that large air voids at the steel–concrete interface significantly increase the risk of corrosion initiation and propagation in those areas. By combining experimental observations with computer modelling, the study also revealed how corrosion products behave under pressure: how they expand, compress and interact with the surrounding concrete. This made it possible to estimate mechanical effects that had not previously been measured directly. Together, these methods offer a new way to study corrosion -- one that in future could help engineers detect hidden corrosion damage earlier and design concrete structures that last longer.
Corrosion cracking: Image-based methods combined with modelling
Formas (2019-00497), 2020-01-01 -- 2022-12-31.
Splitting cracks as performance indicator for existing structures
Swedish Transport Administration (TRV2021/27819), 2021-01-01 -- 2023-12-31.
Splitting cracks as performance indicator for existing structures, part 2
Swedish Transport Administration (TRV2024/62846), 2024-08-01 -- 2026-12-15.
Structural safety of corroded reinforced concrete structures from visual inspection
Formas (2022-01175), 2023-01-01 -- 2025-12-31.
Subject Categories (SSIF 2025)
Structural Engineering
ISBN
978-91-8103-247-5
Doktorsavhandlingar vid Chalmers tekniska högskola. Ny serie: 5705
Publisher
Chalmers
SB-H6, Sven hultins gata 6, Chalmers
Opponent: Associate Professor Els Verstrynge, Department of Civil Engineering, KU Leuven, Belgium