International Concrete Abstracts Portal

This study evaluated the effect of class F fly ash (5, 10, 15, and 20%) and silica fume (20%) as partial cement replacements on bacterial crack healing. Concrete cylinders were prepared, cracked into one-inch disks, and submerged in fresh water. Healing progress was monitored over 18 weeks using microscopy and quantified through a healing index. Results showed that bacterial activity substantially improved healing compared to natural hydration in control specimens. Fly ash replacement did not prevent healing, and several disks across all percentages achieved complete crack closure. However, higher fly ash levels shortened the duration of bacterial activity, indicating sensitivity to calcium availability. At 20% fly ash, healing progressed more slowly but remained active at 18 weeks. In contrast, specimens containing 20% silica exhibited significantly lower healing efficiency, with few disks achieving full closure and overall lower healing indices. These results confirm that bacteria-based self-healing concrete remains effective with fly ash but is constrained by high silica fume content due to very low to zero calcium content in silica fume. The findings indicated that lower calcium content in supplementary cementitious material (SCM) replacement, either due to higher fly ash content with lower calcium compared to OPC or with silica fume with almost zero calcium content, with bacteria, may have a significant effect on the healing progress.

The concept of multi-generational concrete recycling is increasingly relevant as many existing recycled concrete structures near the end of their service lives. This study examines the performance variation and recyclability of multi-generational concrete subjected to chloride salt dry-wet cycling. After 30 dry-wet cycles, natural aggregate concrete, designed with three different strength grades, was crushed to produce the first generation of recycled fine aggregate, which was then used to prepare the second generation of concrete. This second generation was subjected to the same dry-wet cycling and subsequently crushed to yield a second generation of recycled fine aggregate. The results demonstrate a significant decline in the performance of the second generation of concrete, with an average compressive strength reaching only 89.52% of the first generation. Notably, the performance deterioration was more pronounced in lower-strength mixes, which exhibited increased porosity, greater mass loss, and deeper chloride penetration. Both generations of recycled fine aggregate met the standards for Class III aggregate; however, some properties of the recycled fine aggregate derived from higher-strength concrete qualified for Class II aggregate status. Additionally, a regression analysis model was developed to predict the attenuation coefficients for the third generation of concrete with design strengths of 30, 45, and 60 MPa, yielding coefficients of 56.84, 67.75, and 71.72%, respectively. This study underscores the potential for multi-generational use of recycled fine aggregates and highlights the importance of selecting appropriate design strengths to enhance durability and recyclability in chloride-rich environments.

Advances in concrete material science have led to the development of a new class of cementitious materials, namely ultra-high-performance concrete (UHPC), which offers superior mechanical and durability properties. The control and characterization of the fresh properties of UHPC are crucial for successful mixture design. Among the methods for evaluating these properties, the mini-cone test has gained prominence due to its practicality. It requires smaller sample volumes than the standard slump cone test, making it especially suited for laboratory assessments of UHPC mixtures. In contrast, the slump flow test is the simplest and most widely used test for both laboratory and field testing of concrete. This study aims to establish a correlation between mini-cone flow and standard slump flow test results. A linear relationship is identified, which forms the basis for proposing consistency classes for UHPC using mini-cone flow values. These proposed classes align with the established consistency classifications for self-compacting concrete.

Design recommendations are presented for calculating the immediate deflection of cracked prestressed concrete members under service loads. Inconsistency and sometimes confusion regarding the calculation of immediate deflection for the different approaches presently available highlight the need for a rational approach toward computing deflection. The ACI 318-19 approach for reinforced (nonprestressed) concrete is broadened to include prestressed concrete. This involves the implementation of an effective moment of inertia, taken together with an effective eccentricity of the prestressing steel, used to define the effective curvature and/or camber from the prestressing force. Proposed revisions to ACI 318 are presented for prestressed Class T and Class C flexural members, and clear steps are provided for calculating immediate deflection. The effectiveness of the new approach is validated against an extensive database of test results, showing reasonable accuracy and reliability in predicting deflections. The paper concludes with practical recommendations for implementation and a worked-out example to illustrate the proposed methodology. These findings aim to enhance the accuracy and consistency of deflection predictions in prestressed concrete design, contributing to better serviceability and performance of concrete structures.

In this study, the challenge of automating concrete crack image classification by developing a lightweight machine learning model that balances accuracy with computational efficiency was addressed. Traditional deep learning models, while accurate, suffer from high computational demands, limiting their practicality in on-site applications. This study’s approach used the random forest (RF) classifier combined with histogram of oriented gradients (HOG) and local binary patterns (LBP) for feature extraction, offering a more feasible alternative for real-time structural health monitoring. Comparative analysis with the convolutional neural network (CNN) model highlights this model’s significantly reduced size and inference times, with only a marginal compromise in accuracy. The results demonstrated that the RF models, particularly RF with LBP, are well-suited for integration into resource-constrained environments, paving the way for their deployment in portable, on-site diagnostic systems in civil engineering. This study contributed a novel perspective to the field, emphasizing the importance of efficient machine learning solutions in practical applications of structural health monitoring.