Predicting concrete corrosion of sewers using artificial neural network

Predictive analysis of concrete slump using a stochastic search-consolidated neural network

Attaining a dependable measurement of concrete slump is crucial as it is a valuable indication of concrete workability. On the other hand, complexities associated with costly traditional approaches have driven engineers to use indirect efficient models such as metaheuristic-based machine learning for approximating the slump. While the literature shows promising application of some metaheuristic techniques for this purpose, the large variety of these algorithms calls for evaluating the most capable ones to keep the solution updated. Stochastic fractal search (SFS) is one of the most powerful optimization algorithms in the literature that has not received appropriate attention in analyzing concrete mechanical parameters. In the present research, a multi-layer perceptron neural network (NN-MLP), is enhanced using the SFS. The proposed SFS-NN-MLP model aims to predict the slump based on the amount of ingredients in the mixture, as well as the curing age of specimens. Accuracy assessment revealed that the proposed model can deal with the assigned task with excellent accuracy. It indicates that the SFS could properly tune the parameters required for training the NN-MLP, and consequently, the trained network could reliably calculate the slump of specimens that were not analyzed before. For comparative validation, the SFS was replaced with two similar optimizers, namely elephant herding optimization algorithm (EHO) and slime mould algorithm (SMA). Based on the calculated mean square errors of 5.6526, 6.1129, and 7.3561 along with mean absolute errors of 4.6657, 5.0078, and 6.3066, as well as the percentage-Pearson correlation coefficients of 78.06 %, 73.95 %, and 58.11 %, respectively for the SFS-NN-MLP, EHO-NN-MLP, and SMA-NN-MLP, it was shown that the SFS-NN-MLP is the most accurate predictor. Hereupon, the SFS-NN-MLP model is recommended to be effectively used for obtaining a cost-efficient approximation of concrete slump in real-world projects.

A multi-tier deterioration assessment models for sewer and stormwater pipelines in Hong Kong

Sewerage and stormwater networks are subjected to several deterioration factors, including aging, environmental conditions, and traffic. Maintaining these critical assets in good condition is essential to avoid harmful consequences, such as environmental contamination and negative implications on other infrastructure systems (e.g., water and road networks). Deterioration assessment models are effective and cost-efficient means for proactive management systems that can reduce such consequences. In this connection, this study aims to develop deterioration assessment models for sewer and stormwater pipelines in Hong Kong. First, critical factors that impact the deterioration process of these pipelines were identified. Data for these factors were then collected from the Drainage Services Department (DSD) and open-source data provided by the Hong Kong government. To improve prediction accuracy, a multi-tier concept was utilized in building the models. The first tier categorized pipelines into two groups: fail and not fail, whereas the second tier assigned a grade range from 1 to 3 to the "not fail" pipelines. Several artificial intelligence approaches, such as random forest, neural network, and SVM, were tested. Random forest achieved the highest accuracy in predicting pipelines condition, followed by neural networks. A sensitivity analysis was carried out to investigate the combined impact of two factors, with age being one of them, on the pipeline's performance. The findings of this study provide a robust decision-making tool that DSD authorities and consultants can use to optimize inspection and maintenance activities.

A critical review of chemical uses in urban sewer systems

Chemical dosing is the most used strategy for sulfide and methane abatement in urban sewer systems. Although conventional physicochemical methods, such as sulfide oxidation (e.g., oxygen/nitrate), precipitation (e.g., iron salts), and pH elevation (e.g., magnesium hydroxide/sodium hydroxide) have been used since the last century, the high chemical cost, large environmental footprint, and side-effects on downstream treatment processes demand a sustainable and cost-effective alternative to these approaches. In this paper, we aimed to review the currently used chemicals and significant progress made in sustainable sulfide and methane abatement technology, including 1) the use of bio-inhibitors, 2) in situ chemical production, and 3) an effective dosing strategy. To enhance the cost-effectiveness of chemical applications in urban sewer systems, two research directions have emerged: 1) online control and optimization of chemical dosing strategies and 2) integrated use of chemicals in urban sewer and wastewater treatment systems. The integration of these approaches offers considerable system-wide benefits; however, further development and comprehensive studies are required.

Hydrogen sulfide control in sewer systems: A critical review of recent progress

In sewer systems where anaerobic conditions are present, sulfate-reducing bacteria reduce sulfate to hydrogen sulfide (H₂S), leading to sewer corrosion and odor emission. Various sulfide/corrosion control strategies have been proposed, demonstrated, and optimized in the past decades. These included (1) chemical addition to sewage to reduce sulfide formation, to remove dissolved sulfide after its formation, or to reduce H₂S emission from sewage to sewer air, (2) ventilation to reduce the H₂S and humidity levels in sewer air, and (3) amendments of pipe materials/surfaces to retard corrosion. This work aims to comprehensively review both the commonly used sulfide control measures and the emerging technologies, and to shed light on their underlying mechanisms. The optimal use of the above-stated strategies is also analyzed and discussed in depth. The key knowledge gaps and major challenges associated with these control strategies are identified and strategies dealing with these gaps and challenges are recommended. Finally, we emphasize a holistic approach to sulfide control by managing sewer networks as an integral part of an urban water system.

A Review of Trends in Corrosion-Resistant Structural Steels Research-From Theoretical Simulation to Data-Driven Directions

This paper provides a review of models commonly used over the years in the study of microscopic models of material corrosion mechanisms, data mining methods and the corrosion-resistant performance control of structural steels. The virtual process of material corrosion is combined with experimental data to reflect the microscopic mechanism of material corrosion from a nano-scale to macro-scale, respectively. Data mining methods focus on predicting and modeling the corrosion rate and corrosion life of materials. Data-driven control of the corrosion resistance of structural steels is achieved through micro-alloying and organization structure control technology. Corrosion modeling has been used to assess the effects of alloying elements, grain size and organization purity on corrosion resistance, and to determine the contents of alloying elements.

Predicting the concentrations of enteric viruses in urban rivers running through the city center via an artificial neural network

Viral waterborne diseases are widespread in cities due largely to the occurrence of enteric viruses in urban rivers, which pose a significant concern to human health. Yet, the application of rapid detection technology for enteric viruses in environmental water remains undeveloped globally. Here, multiple linear regression (MLR) modeling and artificial neural network (ANN) modeling, which used frequently measured physicochemical parameters in river water, were constructed to predict the concentration of enteric viruses including human enteroviruses (EnVs), rotaviruses (HRVs), astroviruses (AstVs), noroviruses GⅡ (HuNoVs GⅡ), and adenoviruses (HAdVs) in rivers. After training, testing, and validating, ANN models showed better performance than any MLR model for predicting the viral concentration in Jinhe River. All determined R-values for ANN models exceeded 0.89, suggesting a strong correlation between the predicted and measured outputs for target enteric viruses. Furthermore, ANN models provided a better congruence between the observed and predicted concentrations of each virus than MLR models did. Together, these findings strongly suggest that ANN modeling can provide more accurate and timely predictions of viral concentrations based on frequent (or routine) measurements of physicochemical parameters in river water, which would improve assessments of waterborne disease prevalence in cities.

Towards a decentralized solution for sewer leakage detection - a review

Sewer pipelines often leak due to physical, operational, and environmental deterioration factors. Due to the hidden infrastructure of the sewer systems, leakage detection is often costly, challenging, and crucial at the city scale. Various sewer inspection methods (SIMs) have been developed and implemented at this time. This study evaluates the existing SIMs and categorizes them based on their area of impact (AoI) into three classes. Tier-one (T-I) methods, such as deterioration models and hotspot mapping, tend to grasp a broader and reliable understanding of the sewer systems' structural health and pinpoint the network sections that are more prone to leakage. As an intermediate solution, Tier-two (T-II) non-destructive methods, such as aerial thermal imagery (ATI) and electrical resistivity tomography (ERT), inspect the potential pipe clusters regardless of their material and visualize the leaked plume generated from defects and cracks. Tier-three (T-III) methods include in-pipe SIMs, such as visual and multi-sensory inspections, that can provide an in-depth understanding of the pipe and its deterioration stage. In this study, we suggest that a sustainable sewer inspection plan should include at least two SIMs belonging to different tiers to provide a dual investigation of precision and AoI, a balance between cost and time as well as an equilibrium between self-sufficiency and decentralization.

Predicting Compressive Strength of Blast Furnace Slag and Fly Ash Based Sustainable Concrete Using Machine Learning Techniques: An Application of Advanced Decision-Making Approaches

The utilization of waste industrial materials such as Blast Furnace Slag (BFS) and Fly Ash (F. Ash) will provide an effective alternative strategy for producing eco-friendly and sustainable concrete production. However, testing is a time-consuming process, and the use of soft machine learning (ML) techniques to predict concrete strength can help speed up the procedure. In this study, artificial neural networks (ANNs) and decision trees (DTs) were used for predicting the compressive strength of the concrete. A total of 1030 datasets with eight factors (OPC, F. Ash, BFS, water, days, SP, FA, and CA) were used as input variables for the prediction of concrete compressive strength (response) with the help of training and testing individual models. The reliability and accuracy of the developed models are evaluated in terms of statistical analysis such as R2, RMSE, MAD and SSE. Both models showed a strong correlation and high accuracy between predicted and actual Compressive Strength (CS) along with the eight factors. The DT model gave a significant relation to the CS with R2 values of 0.943 and 0.836, respectively. Hence, the ANNs and DT models can be utilized to predict and train the compressive strength of high-performance concrete and to achieve long-term sustainability. This study will help in the development of prediction models for composite materials for buildings.

Potential of supervised machine learning algorithms for estimating the impact of water efficient scenarios on solids accumulation in sewers

Understanding the negative effects of widespread implementation of optimal water efficient solutions may have on existing centralised sewer systems is still limited - one of these effects is the accumulation of solids in sewer pipes. Predicting these effects requires setting up and simulating complex detailed hydraulic sewer network models. Often, precise details of the sewer network layout and diurnal patterns of the wastewater flows are not available, limiting the applicability of using model predictions for such phenomena. In this study, the applicability of supervised machine learning (ML) algorithms for the development of a simplified surrogate model to predict solid accumulation in sewer pipes was investigated. A large number of highly variable sewer networks were synthetically generated and used to produce results that can be generalizable within the limitations of the current study. A hydrodynamic sewer model was set up and simulated for each synthetic sewer network and various scenarios in which different water-efficient solutions were considered. Simulation results indicated that the most impacts are expected to occur in the upstream part of the sewer networks, and that with 50% reduction in (waste-)water flows, 3-20% more pipes are expected to accumulate solids. It was further found that ML algorithms can be used to successfully predict locations of solids accumulation in sewer pipes without using hydrodynamic models. A simple tool based on the findings of this study, sparing the need to conduct complex hydraulic simulations, was developed. It allows the user to enter a set of pipe characteristics and the proportion of flow that is reduced due to the implementation of water efficient solutions, and it predicts whether the pipe will accumulate solids or not. The study results and the proposed ML algorithms can support the implementation of optimal water-efficient solutions that will promote designing and managing the water sensitive cities of the future.

Congo red dye removal from aqueous environment by cationic surfactant modified-biomass derived carbon: Equilibrium, kinetic, and thermodynamic modeling, and forecasting via artificial neural network approach

Herein, it was aimed to optimize, model, and forecast the biosorption of Congo Red onto biomass-derived biosorbent. Therefore, the waste-orange-peels were processed to fabricate biomass-derived carbon, which was activated by ZnCl₂ and modified with cetyltrimethylammonium bromide. The physicochemical properties of the biosorbents were explored by scanning electron microscopy and N₂ adsorption/desorption isotherms. The effects of pH, initial dye concentration, temperature, and contact duration on the biosorption capacity were investigated and optimized by batch experimental process, followed by the kinetics, equilibrium, and thermodynamics of biosorption were modeled. Furthermore, various artificial neural network (ANN) architectures were applied to experimental data to optimize the ANN model. The kinetic modeling of the biosorption offered that biosorption was in accordance both with the pseudo-second-order and saturation-type kinetic model, and the monolayer biosorption capacity was calculated as 666.67 mg g^-1 at 25 °C according to Langmuir isotherm model. According to equilibrium modeling, the Freundlich isotherm model was better fitted to the experimental data than the Langmuir isotherm model. Moreover, the thermodynamic modeling revealed biosorption took place spontaneously as an exothermic process. The findings revealed that the best ANN architecture trained with trainlm as the backpropagation algorithm, with tansig-purelin transfer functions, and 14 neurons in the single hidden layer with the highest coefficient of determination (R² = 0.9996) and the lowest mean-squared-error (MSE = 0.0002). The well-agreement between the experimental and ANN-forecasted data demonstrated that the optimized ANN model can predict the behavior of the anionic dye biosorption onto biomass-derived modified carbon materials under various operation conditions.

Investigation of ANN architecture for predicting the compressive strength of concrete containing GGBFS

An extensive simulation program is used in this study to discover the best ANN model for predicting the compressive strength of concrete containing Ground Granulated Blast Furnace Slag (GGBFS). To accomplish this purpose, an experimental database of 595 samples is compiled from the literature and utilized to find the best ANN architecture. The cement content, water content, coarse aggregate content, fine aggregate content, GGBFS content, carboxylic type hyper plasticizing content, superplasticizer content, and testing age are the eight inputs in this database. As a result, the optimal selection of the ANN design is carried out and evaluated using conventional statistical metrics. The results demonstrate that utilizing the best architecture [8-14-4-1] among the 240 investigated architectures, and the best ANN model, is a very efficient predictor of the compressive strength of concrete using GGBFS, with a maximum R2 value of 0.968 on the training part and 0.965 on the testing part. Furthermore, a sensitivity analysis is performed over 500 Monte Carlo simulations using the best ANN model to determine the reliability of ANN model in predicting the compressive strength of concrete. The findings of this research may make it easier and more efficient to apply the ANN model to many civil engineering challenges.

A Two-Stage Data-Driven Spatiotemporal Analysis to Predict Failure Risk of Urban Sewer Systems Leveraging Machine Learning Algorithms

Risk-informed asset management is key to maintaining optimal performance and efficiency of urban sewer systems. Although sewer system failures are spatiotemporal in nature, previous studies analyzed failure risk from a unidimensional aspect (either spatial or temporal), not accounting for bidimensional spatiotemporal complexities. This is owing to the insufficiency of good-quality data, which ultimately leads to under-/overestimation of failure risk. Here, we propose a generalized methodology/framework to facilitate a robust spatiotemporal analysis of urban sewer system failure risk, overcoming the intrinsic challenges of data imperfections-e.g., missing data, outliers, and imbalanced information. The framework includes a two-stage data-driven modeling technique that efficiently models the highly right-skewed sewer system failure data to predict the failure risk, leveraging a bidimensional space-time approach. We implemented our analysis for Bogotá, the capital city of Colombia. We train, test, and validate a battery of machine learning algorithms-logistic regression, decision trees, random forests, and XGBoost-and select the best model in terms of goodness-of-fit and predictive accuracy. Finally, we illustrate the applicability of the framework in planning/scheduling sewer system maintenance operations using state-of-the-art optimization techniques. Our proposed framework can help stakeholders to analyze the failure-risk models' performance under different discrimination thresholds, and provide managerial insights on the model's adequate spatial resolution and appropriateness of decentralized management for sewer system maintenance.

Data-driven estimation of COVID-19 community prevalence through wastewater-based epidemiology

Wastewater-based epidemiology (WBE) has been regarded as a potential tool for the prevalence estimation of coronavirus disease 2019 (COVID-19) in the community. However, the application of the conventional back-estimation approach is currently limited due to the methodological challenges and various uncertainties. This study systematically performed meta-analysis for WBE datasets and investigated the use of data-driven models for the COVID-19 community prevalence in lieu of the conventional WBE back-estimation approach. Three different data-driven models, i.e. multiple linear regression (MLR), artificial neural network (ANN), and adaptive neuro fuzzy inference system (ANFIS) were applied to the multi-national WBE dataset. To evaluate the robustness of these models, predictions for sixteen scenarios with partial inputs were compared against the actual prevalence reports from clinical testing. The performance of models was further validated using unseen data (data sets not included for establishing the model) from different stages of the COVID-19 outbreak. Generally, ANN and ANFIS models showed better accuracy and robustness over MLR models. Air and wastewater temperature played a critical role in the prevalence estimation by data-driven models, especially MLR models. With unseen datasets, ANN model reasonably estimated the prevalence of COVID-19 (cumulative cases) at the initial phase and forecasted the upcoming new cases in 2-4 days at the post-peak phase of the COVID-19 outbreak. This study provided essential information about the feasibility and accuracy of data-driven estimation of COVID-19 prevalence through the WBE approach.

Water quality modeling in sewer networks: Review and future research directions

Urban sewer networks (SNs) are increasingly facing water quality issues as a result of many challenges, such as population growth, urbanization and climate change. A promising way to addressing these issues is by developing and using water quality models. Many of these models have been developed in recent years to facilitate the management of SNs. Given the proliferation of different water quality models and the promise they have shown, it is timely to assess the state-of-the-art in this field, to identify potential challenges and suggest future research directions. In this review, model types, modeled quality parameters, modeling purpose, data availability, type of case studies and model performance evaluation are critically analyzed and discussed based on a review of 110 papers published between 2010 and 2019. The review identified that applications of empirical and kinetic models dominate those of data-driven models for addressing water quality issues. The majority of models are developed for prediction and process understanding using experimental or field sampled data. While many models have been applied to real problems, the corresponding prediction accuracies are overall moderate or, in some cases, low, especially when dealing with larger SNs. The review also identified the most common issues associated with water quality modeling of SNs and based on these proposed several future research directions. These include the identification of appropriate data resolutions for the development of different SN models, the need and opportunity to develop hybrid SN models and the improvement of SN model transferability.

Estimation of in-situ biogas upgrading in microbial electrolysis cells via direct electron transfer: Two-stage machine learning modeling based on a NARX-BP hybrid neural network

With the increasing of data in wastewater treatment, data-driven machine learning models are useful for modeling biological processes and complex reactions. However, few data-driven models have been developed for simulating the microbial electrolysis cells (MECs) and traditional models are too ambiguous to comprehend the mechanisms. In this study, a new general data-driven two-stage model was firstly developed to predict CH₄ production from in-situ biogas upgrading in the biocathode MECs via direct electron transfer (DET), named NARX-BP hybrid neural networks. Compared with traditional one-stage model, the model could well predict methane production via DET with excellent performance (all R² and MES of 0.918 and 6.52 × 10^-2, respectively) and reveal the mechanisms of biogas upgrading, for the new systematical modeling approach could improve the versatility and applicability by inputting significant intermediate variables. In addition, the model is generally available to support long-term prediction and optimal operation for anaerobic digestion or complex MEC systems.

Review of Fiber Optical Sensors and Its Importance in Sewer Corrosion Factor Analysis

Adverse effects of wastewater on the hygiene of human and circumstances is a major issue in society. Appropriate refining systems with high efficiency is required to treat the wastewater. Sewage treatment plant plays a major and important role in conserving incredible nature of the environment. Microbiologically Induced Corrosion (MIC) is an important phenomenon in sewage structures which causes the deterioration of infrastructures. Huge capital has been spent and efforts have been made on wastewater treatment infrastructure to increase operating efficiency and reliability of compliance. The investments in reimbursement and maintenance of sewer structures upsurge with an increase in the rate of MIC. The focus of this review is to describe MIC in sewer structure and the factors influencing the corrosion such as the generation of Sulfuric acid (H2SO4), Relative Humidity (RH), pH of the concrete structure and temperature. Modern developments in the design of Fiber Optical Sensors (FOSs) for observing the parameters including pH, Hydrogen Sulfide (H2S), RH and temperature will be discussed.

Acoustic Emission Study on the Damage Evolution of a Corroded Reinforced Concrete Column under Axial Loads

In this paper, the damage of a reinforced concrete (RC) column with various levels of reinforcement corrosion under axial loads is characterized using the acoustic emission (AE) technique. Based on the AE rate process theory, a modified damage evolution equation of RC associated with the axial load and different corrosion rates is proposed. The experimental results show that the measured AE signal parameters during the loading process are closely related to the damage evolution of the RC column as well as the reinforcement corrosion level. The proposed modified damage evolution equation enables dynamic analysis for the damage of corrosion on a RC column under axial loading for a further real-time quantitative evaluation of corrosion damage on reinforced concrete.

The use of artificial neural network (ANN) for modeling adsorption of sunset yellow onto neodymium modified ordered mesoporous carbon

Discharging coloring products in water bodies has degraded water quality irreversibly over the past several decades. Order mesoporous carbon (OMC) was modified by embedding neodymium(III) chloride on the surface of OMC to enhance the adsorptive removal towards these contaminants. This paper represents an artificial neural network (ANN) based approach for modeling the adsorption process of sunset yellow onto neodymium modified OMC (OMC-Nd) in batch adsorption experiments. Neodymium modified OMC was characterized using N₂ adsorption-desorption isotherm, TEM micrographs, FT-IR and XPS spectra analysis techniques. 2.5 wt% Nd loaded OMC was selected as the final adsorbent for further experiments because OMC-2.5Nd showed highest removal efficiency of 93%. The ANN model was trained and validated with the adsorption experiments data where initial concentration, reaction time, and adsorbent dosage were selected as the variables for the batch study, whereas the removal efficiency was considered as the output. The ANN model was first developed using a three-layer back propagation network with the optimum structure of 3-6-1. The model employed tangent sigmoid transfer function as input in the hidden layer whereas a linear transfer function was used in the output layer. The comparison between modeled data and experimental data provided high degree of correlation (R² = 0.9832) which indicated the applicability of ANN model for describing the adsorption process with reasonable accuracy.

Rebar corrosion and its interaction with concrete degradation in reinforced concrete sewers

Concrete corrosion, as a major issue in sewer management, has attracted considerable research. In comparison, the corrosion of reinforcing steel bar (rebar) is not well understood. Particularly, fundamental knowledge of rebar corrosion and its interactions with concrete corrosion/cracking is largely lacking. This study investigated rebar corrosion and concrete degradation using reinforced concrete coupons exposed in a pilot sewer system. The physical-chemical corrosion characteristics were investigated in local regions; the nature of rebar rusts was analyzed using the advanced mineral analytical techniques, including Scanning Electron Microscope (SEM), Energy Dispersive X-ray Spectroscopy (EDS) and X-ray Diffraction (XRD); further, the interactions between rebar corrosion and concrete corrosion/cracking were elucidated by characterizing the microstructure and element distribution in interfacial areas using Mineral Liberation Analysis (MLA). The rebar corrosion products were found to be iron oxides, oxyhydroxides, chlorides, sulfides and sulfates. The predominant rebar corrosion reactions varied with exposure time and the development of concrete corrosion. When concrete corrosion reached rebar surface, the cracking of the concrete cover was influenced by multiple effects, including the macro-cracking induced by the corrosion products expansion, and the micro-cracking accelerated by the dissolution, diffusion and deposition of Fe derived from rebar rusts at the concrete corrosion front. A conceptual model elucidating rebar corrosion and the complex interactions between rebar corrosion and concrete degradation is proposed to support the development of corrosion prevention and refurbishment strategies for reinforced concrete sewers.

The rapid chemically induced corrosion of concrete sewers at high H2S concentration

Concrete corrosion in sewers is primarily caused by H₂S in sewer atmosphere. H₂S concentration can vary from several ppm to hundreds of ppm in real sewers. Our understanding of sewer corrosion has increased dramatically in recent years, however, there is limited knowledge of the concrete corrosion at high H₂S levels. This study examined the corrosion development in sewers with high H₂S concentrations. Fresh concrete coupons, manufactured according to sewer pipe standards, were exposed to corrosive conditions in a pilot-scale gravity sewer system with gaseous H₂S at 1100 ± 100 ppm. The corrosion process was continuously monitored by measuring the surface pH, corrosion product composition, corrosion loss and the microbial community. The surface pH of concrete was reduced from 10.5 ± 0.3 to 3.1 ± 0.5 within 20 days and this coincided with a rapid corrosion rate of 3.5 ± 0.3 mm year ^-1. Microbial community analysis based on 16S rRNA gene sequencing indicated the absence of sulfide-oxidizing microorganisms in the corrosion layer. The chemical analysis of corrosion products supported the reaction of cement with sulfuric acid formed by the chemical oxidation of H₂S. The rapid corrosion of concrete in the gravity pipe was confirmed to be caused by the chemical oxidation of hydrogen sulfide at high concentrations. This is in contrast to the conventional knowledge that is focused on microbially induced corrosion. This first-ever systematic investigation shows that chemically induced oxidation of H₂S leads to the rapid corrosion of new concrete sewers within a few weeks. These findings contribute novel understanding of in-sewer corrosion processes and hold profound implications for sewer operation and corrosion management.

Prediction and Knowledge Mining of Outdoor Atmospheric Corrosion Rates of Low Alloy Steels Based on the Random Forests Approach

The objective of this paper is to develop an approach to forecast the outdoor atmospheric corrosion rate of low alloy steels and do corrosion-knowledge mining by using a Random Forests algorithm as a mining tool. We collected the corrosion data of 17 low alloy steels under 6 atmospheric corrosion test stations in China over 16 years as the experimental datasets. Based on the datasets, a Random Forests model is established to implement the purpose of the corrosion rate prediction and data-mining. The results showed that the random forests can achieve the best generalization results compared to the commonly used machine learning methods such as the artificial neural network, support vector regression, and logistic regression. In addition, the results also showed that regarding the effect to the corrosion rate, environmental factors contributed more than chemical compositions in the low alloy steels, but as exposure time increases, the effect of the environmental factors will gradually become less. Furthermore, we give the effect changes of six environmental factors (Cl− concentration, SO2 concentration, relative humidity, temperature, rainfall, and pH) on corrosion with exposure time increasing, and the results illustrated that pH had a significant contribution to the corrosion of the entire process. The paper also dealt with the problem of the corrosion rate forecast, especially for changing environmental factors situations, and obtained the qualitative and quantitative results of influences of each environmental factor on corrosion.

Artificial Intelligence Approaches for Prediction of Compressive Strength of Geopolymer Concrete

Geopolymer concrete (GPC) has been used as a partial replacement of Portland cement concrete (PCC) in various construction applications. In this paper, two artificial intelligence approaches, namely adaptive neuro fuzzy inference (ANFIS) and artificial neural network (ANN), were used to predict the compressive strength of GPC, where coarse and fine waste steel slag were used as aggregates. The prepared mixtures contained fly ash, sodium hydroxide in solid state, sodium silicate solution, coarse and fine steel slag aggregates as well as water, in which four variables (fly ash, sodium hydroxide, sodium silicate solution, and water) were used as input parameters for modeling. A total number of 210 samples were prepared with target-specified compressive strength at standard age of 28 days of 25, 35, and 45 MPa. Such values were obtained and used as targets for the two AI prediction tools. Evaluation of the model's performance was achieved via criteria such as mean absolute error (MAE), root mean square error (RMSE), and coefficient of determination (R²). The results showed that both ANN and ANFIS models have strong potential for predicting the compressive strength of GPC but ANFIS (MAE = 1.655 MPa, RMSE = 2.265 MPa, and R² = 0.879) is better than ANN (MAE = 1.989 MPa, RMSE = 2.423 MPa, and R² = 0.851). Sensitivity analysis was then carried out, and it was found that reducing one input parameter could only make a small change to the prediction performance.

Evaluation of data-driven models for predicting the service life of concrete sewer pipes subjected to corrosion

Concrete corrosion is one of the most significant failure mechanisms of sewer pipes, and can reduce the sewer service life significantly. To facilitate the management and maintenance of sewers, it is essential to obtain reliable prediction of the expected service life of sewers, especially if that is based on limited environmental conditions. Recently, a long-term study was performed to identify the controlling factors of concrete sewer corrosion using well-controlled laboratory-scale corrosion chambers to vary levels of H₂S concentration, relative humidity, temperature and in-sewer location. Using the results of the long-term study, three different data-driven models, i.e. multiple linear regression (MLR), artificial neural network (ANN), and adaptive neuro fuzzy inference system (ANFIS), as well as the interaction between environmental parameters, were assessed for predicting the corrosion initiation time (t_i) and corrosion rate (r). This was performed using the sewer environmental factors as the input under 12 different scenarios after allowing for an initiation corrosion period. ANN and ANFIS models showed better performance than MLR models, with or without considering the interactions between environmental factors. With the limited input data available, it was observed that t_i prediction by these models is quite sensitive, however, they are more robust for predicting r as long as the H₂S concentration is available. Using the H₂S concentration as a single input, all three data driven models can reasonably predict the sewer service life.

Thermochemical Humidity Detection in Harsh or Non-Steady Environments

We present a new method of chemical quantification utilizing thermal analysis for the detection of relative humidity. By measuring the temperature change of a hydrophilically-modified temperature sensing element vs. a hydrophobically-modified reference element, the total heat from chemical interactions in the sensing element can be measured and used to calculate a change in relative humidity. We have probed the concept by assuming constant temperature streams, and having constant reference humidity (~0% in this case). The concept has been probed with the two methods presented here: (1) a thermistor-based method and (2) a thermographic method. For the first method, a hydrophilically-modified thermistor was used, and a detection range of 0–75% relative humidity was demonstrated. For the second method, a hydrophilically-modified disposable surface (sensing element) and thermal camera were used, and thermal signatures for different relative humidity were demonstrated. These new methods offer opportunities in either chemically harsh environments or in rapidly changing environments. For sensing humidity in a chemically harsh environment, a hydrophilically-modified thermistor can provide a sensing method, eliminating the exposure of metallic contacts, which can be easily corroded by the environment. On the other hand, the thermographic method can be applied with a disposable non-contact sensing element, which is a low-cost upkeep option in environments where damage or fouling is inevitable. In addition, for environments that are rapidly changing, the thermographic method could potentially provide a very rapid humidity measurement as the chemical interactions are rapid and their changes are easily quantified.

Modeling Self-Healing of Concrete Using Hybrid Genetic Algorithm-Artificial Neural Network

This paper presents an approach to predicting the intrinsic self-healing in concrete using a hybrid genetic algorithm–artificial neural network (GA–ANN). A genetic algorithm was implemented in the network as a stochastic optimizing tool for the initial optimal weights and biases. This approach can assist the network in achieving a global optimum and avoid the possibility of the network getting trapped at local optima. The proposed model was trained and validated using an especially built database using various experimental studies retrieved from the open literature. The model inputs include the cement content, water-to-cement ratio (w/c), type and dosage of supplementary cementitious materials, bio-healing materials, and both expansive and crystalline additives. Self-healing indicated by means of crack width is the model output. The results showed that the proposed GA–ANN model is capable of capturing the complex effects of various self-healing agents (e.g., biochemical material, silica-based additive, expansive and crystalline components) on the self-healing performance in cement-based materials.

Prediction of concrete corrosion in sewers with hybrid Gaussian processes regression model

A hybrid Gaussian Processes Regression (GPR) model is to approach the evolution of the corrosion rate and corrosion initiation time, thereby supporting the calculation of service life of sewers.