Search

This work presents the design and application of a low-cycle reciprocating loading test on 23 recycled aggregate concrete-filled steel tube columns and 3 ordinary concrete-filled steel tube columns. Additionally, a systematic study on the influence of various parameters (e.g., slenderness ratio, axial compression ratio, etc.) was conducted on the seismic performance of the specimens. The results show that all the specimens have good hysteresis performance and a similar development trend of skeleton curve. The influence of slenderness ratio on the seismic index of the specimens is more significant than that of the axial compression ratio and the steel pipe wall thickness. Furthermore, artificial intelligence was applied to estimate the influence of parameter variation on the seismic performance of concrete columns. Specifically, Random Forest with hyperparameters tuned by Firefly Algorithm was chosen. The high correlation coefficients (R) and low root mean square error values from the prediction results showed acceptable accuracy. In addition, sensitivity analysis was applied to rank the influence of the aforementioned input variables on the seismic performance of the specimens. The research results can provide experimental reference for the application of steel tube recycled concrete in earthquake areas.

Waste glass is promising to be recycled and reused in construction for sustainability. Silicon dioxide is the main component of glass, however, its pozzolanic activity is latent mainly due to its stable silica tetrahedron structure. To excite the activation of waste glass, chemical activation and mechanical grinding of waste glass powder (WGP) were investigated. As the supplementary, hydrothermal and combined (mechanical-chemical-hydrothermal) treatments were conducted on part of the WGP samples. The unconfined compression strength (UCS), expansion caused by alkali–silica reaction (ASR), and the microstructural morphology of WGP were investigated. The results showed the dosage threshold (around 2%) of the chemical activators (alkali and sodium sulfate) and the combined activation were optimal. Besides, a firefly algorithm (FA) based multi-objective optimisation model (MOFA) was applied to seek the Pareto fronts based on three objectives: UCS, ASR expansion, and Cost of mixture proportion. The objective functions of UCS and expansion were established through training the machine learning (ML) models where FA was used to tune the hyperparameters. The cost was calculated by a polynomial function. The ultimate values of root mean square error (RMSE) and correlation coefficient (R) showed the robustness of the ML models. Moreover, the Pareto fronts for mortars containing 300 μm and 75 μm WGPs were successfully obtained, which contributed to the practical application of waste glass in mortar production. In addition, the sensitivity analysis was conducted to rank the importance of input variables. The results showed that curing time, activator's content, and WGP particle size were three essential parameters.

For low-carbon sustainability, recycled rubber particles (RPs) and recycled aggregate (RA) could be used to make rubber-modified recycled aggregate concrete (RRAC). The characteristics (compressive strength and peak strain) of RRAC with various amounts of RA and RPs after heating at various temperatures were studied in this work. The results show that high temperatures significantly decreased the uniaxial compressive strength (UCS), whereas the addition of RA (e.g., 50%) and RPs (e.g., 5%) can mitigate the negative effect caused by high temperatures. The peak strain can also be improved by increasing the replacement ratios of RA and RP. Support vector regression (SVR) models were trained using a total of 120 groups of UCS and peak strain experimental datasets, and an SVR-based multi-objective optimization model was proposed. The excellent correlation coefficients (0.9772 for UCS and 0.9412 for peak strain) found to illustrate the remarkable accuracy of the SVR models. The Pareto fronts of a tri-objective mixture optimization design (UCS, strain, and cost) were successfully generated as the decision reference at varying temperature conditions. A sensitivity analysis was performed to rank the importance of the input variables where temperature was found as the most important one. In addition, the replacement ratio of RA is more important compared with that of the RP for both the UCS and strain datasets. Among the mechanical properties of concrete, compressive strength and peak strain are two key properties. This study provides guidance for the study of RRAC constitutive models under high temperatures.

In dam engineering, the presence of cracks and crack width are important indicators for diagnosing the health of dams. The accurate measurement of cracks facilitates the safe use of dams. The manual detection of such defects is unsatisfactory in terms of cost, safety, accuracy, and the reliability of evaluation. The introduction of deep learning for crack detection can overcome these issues. However, the current deep learning algorithms possess a large volume of model parameters, high hardware requirements, and difficulty toward embedding in mobile devices such as drones. Therefore, we propose a lightweight MobileNetV2_DeepLabV3 image segmentation network. Furthermore, to prevent interference by noise, light, shadow, and other factors for long-length targets when segmenting, the atrous spatial pyramid pooling (ASPP) module parameters in the DeepLabV3+ network structure were modified, and a multifeature fusion structure was used instead of the parallel structure in ASPP, allowing the network to obtain richer crack features. We collected the images of dam cracks from different environments, established segmentation datasets, and obtained segmentation models through network training. Experiments show that the improved MobileNetV2_DeepLabV3 algorithm exhibited a higher crack segmentation accuracy than the original MobileNetV2_DeepLabV3 algorithm; the average intersection rate attained 83.23%; and the crack detail segmentation was highly accurate. Compared with other semantic segmentation networks, its training time was at least doubled, and the total parameters were reduced by more than 2 to 7 times. After extracting cracks through the semantic segmentation, we proposed to use the method of inscribed circle of crack outline to calculate the maximum width of the detected crack image and to convert it into the actual width of the crack. The maximum relative error rate was 11.22%. The results demonstrated the potential of innovative deep learning methods for dam crack detection.

The dynamic increase factor (DIF) of the concrete material strength, obtained using a split Hopkinson pressure bar (SHPB), includes structural effects that do not precisely reflect the real strain-rate effect of concrete. To further clarify the real strain-rate effects of rubberised concrete (RC), an experimental investigation regarding the dynamic compressive response of ordinary concrete (NC) and RC with three rubber contents (10%, 20%, and 30%) was performed in this study. Additionally, based on a dynamic constitutive model, i.e., the Karagozian and Case (K&C) concrete model, numerical SHPB tests were conducted using the LS-DYNA software. According to the experimental results, all parameters of the K&C model were discussed, and the damage factors were modified to satisfy the mechanical properties of RC. After validating the numerical model, it was observed that the experimental DIF included the inertial enhancement and the real DIF. Moreover, because rubber particles effectively reduce the density and improve the deformation capacity of concrete, the real strain-rate effect of RC was found to be more rate-sensitive than that of NC by analysing the radial stress distribution. In addition to lateral inertia, another external source, namely, the interface friction between the specimen and bars, which can produce lateral confinement, was further studied. It was found that interface friction significantly contributes to lateral confinement; however, as the strain rate increased, the impact generally decreased. Finally, the mechanism of the strain-rate effect of RC was clarified.

The single use of nano-silica (NS) or fly ash (FA) can improve the strength of recycled aggregate concrete (RAC). However, the studies on NS-FA-modified RAC (NFRAC) are limited, and further research is required to understand the mechanical behaviors of NFRAC over different curing ages. In this study, to further improve the properties of RAC, these two supplementary cementitious materials (SCMs) were utilized in RAC together: NFRAC. The workability as well as the compressive and split tensile properties were determined by an orthogonal test along with the effects of various factors, including curing age, recycled coarse aggregate (RCA) replacement rate, and SCM content on the performance of NFRAC. NFRAC was found to have adequate mechanical properties, both in the early and later curing ages. Based on NFRAC with 100% RCA replacement rate as the main research object, the strain ductility, energy consumption, and constitutive model were further explored. Scanning electron microscopy was performed to explore the modified mechanism owing to the incorporation of NS and FA. Both the SCMs can successfully accelerate the pozzolanic reaction, thereby improving the mechanical properties. In addition, NS can also fill the micropores in RCAs and react with the CH of mortar to form C-S-H gels, thereby improving the properties of ITZ. However, NS weakens the workability that can be compensated by incorporating spherical FA. Finally, 10% FA and 2% or 3% NS are recommended as SCMs for RAC.