Mechanical properties of sustainable metakaolin/Rockwool based geopolymer mortar

Innovative Infrastructure Solutions, Год журнала: 2024, Номер 9(7)

Опубликована: Июнь 28, 2024

Язык: Английский

---

Interpreting the setting time of cement pastes for modelling mechanical properties

Case Studies in Construction Materials, Год журнала: 2023, Номер 19, С. e02364 - e02364

Опубликована: Авг. 3, 2023

The formulation of a quick, economical and trustworthy method for predicting the mechanical performance cement pastes in time is rather helpful when developing new materials. In this effort, two different types with increasing water/cement (w/c) ratios were evaluated terms consistency, setting times (ST) their compressive elastic moduli (E). A model has been proposed first to correlate behavior values at 28, 90 180 days. contact mechanics approach was used Vicat test and, thus predictive formulated. utilized ST data calibrated using strength E 28 After comparison theoretical (Epred) experimental (E), proved be accurate specific five w/c respectively. Finally,

Язык: Английский

---

Influence of the ANN Hyperparameters on the Forecast Accuracy of RAC’s Compressive Strength

Materials, Год журнала: 2023, Номер 16(24), С. 7683 - 7683

Опубликована: Дек. 17, 2023

The artificial neural networks (ANNs)-based model has been used to predict the compressive strength of concrete, assisting in creating recycled aggregate concrete mixtures and reducing environmental impact construction industry. Thus, present study examines effects training algorithm, topology, activation function on predictive accuracy ANN when determining concrete. An experimental database with 721 samples was defined considering literature. train, validate, test ANN-based models. Altogether, 240 ANNs were trained, by combining three algorithms, two functions, topologies a hidden layer containing 1-40 neurons. single including 28 neurons, trained Levenberg-Marquardt algorithm hyperbolic tangent function, achieved best level accuracy, coefficient determination equal 0.909 mean absolute percentage error 6.81%. Furthermore, results show that it is crucial avoid use overly complex Excessive neurons can lead exceptional performance during but poor ability testing.

Язык: Английский

---

Reconstructing missing InSAR data by the application of machine leaning-based prediction models: a case study of Rieti

Journal of Civil Structural Health Monitoring, Год журнала: 2023, Номер 14(1), С. 143 - 161

Опубликована: Окт. 30, 2023

Язык: Английский

---

Predicting Construction Accident Outcomes Using Graph Convolutional and Dual-Edge Safety Networks

Arabian Journal for Science and Engineering, Год журнала: 2023, Номер 49(10), С. 13315 - 13332

Опубликована: Дек. 28, 2023

Язык: Английский

---

A water quality prediction model based on signal decomposition and ensemble deep learning techniques

Water Science & Technology, Год журнала: 2023, Номер 88(10), С. 2611 - 2632

Опубликована: Ноя. 15, 2023

Accurate water quality predictions are critical for resource protection, and dissolved oxygen (DO) reflects overall river ecosystem health. This study proposes a hybrid model based on the fusion of signal decomposition deep learning predicting quality. Initially, complete ensemble empirical mode with adaptive noise (CEEMDAN) is employed to split internal series DO into numerous functions (IMFs). Subsequently, we multi-scale fuzzy entropy (MFE) compute values each IMF component. Time-varying filtered (TVFEMD) used further extract features in high-frequency subsequences after linearly aggregating sequences. Finally, support vector machine (SVM) long short-term memory (LSTM) neural networks predict low- subsequences. Moreover, by comparing it single models, models 'single layer decomposition-prediction-ensemble' combination using different methods, feasibility proposed data Xinlian section Fuhe River Chucha Ganjiang was verified. As result, combined prediction approach developed this work has improved generalizability accuracy, may be forecast complicated waters.

Язык: Английский

---

Predicting the impact strength and chloride permeability of concrete made with industrial waste and artificial sand using ANN

Innovative Infrastructure Solutions, Год журнала: 2024, Номер 9(8)

Опубликована: Июль 12, 2024

Язык: Английский

---

Evaluation of strength and modulus of elasticity (Ec) of concrete incorporated with recycled aggregate and rice straw ash (RSA)

Construction and Building Materials, Год журнала: 2024, Номер 448, С. 138016 - 138016

Опубликована: Сен. 6, 2024

Язык: Английский

---

Analysis of the Strength of Different Minerals-Modified MPC Based on Mathematical Models

International Journal of Concrete Structures and Materials, Год журнала: 2024, Номер 18(1)

Опубликована: Март 12, 2024

Abstract The study discussed the effects of different mineral incorporations and curing time on strength modified magnesium phosphate cement (MPC) mortars through mechanical tests, mathematical model analysis microstructure characterization. Fly ash (FA), silica fume (SF), metakaolin (MK), which exhibit excellent durability bonding properties, were used to modify MPC. A quantitative relationship was established between MPC incorporation time. First, each mineral-modified mortar cured in air with durations evaluated. strengths containing 10% fly ash, 15% fume, metakaolin—which perform best their incorporations—were compared analyze function three minerals. To establish time, models, linear model, general nonlinear data distribution shape (DDSNM), are commonly for material property based statistics. DDSNM describes trend change among models error is small Based DDSNM, influence various minerals quantitatively evaluated by calculating variable partial derivatives, verified scanning electron microscopy X-ray diffraction. MK performs improving flexural performance MPC, while SF compressive strength. FA-MPC has low sensitivity dosage fluctuations easy prepare.

Язык: Английский

---

Prediction of concrete compressive strength using a Deepforest-based model

Scientific Reports, Год журнала: 2024, Номер 14(1)

Опубликована: Авг. 14, 2024

Concrete compressive strength testing is crucial for construction quality control. The traditional methods are both time-consuming and labor-intensive, while machine learning has been proven effective in predicting the of concrete. However, current learning-based algorithms lack a thorough comparison among various models, researchers have yet to identify optimal predictor concrete strength. In this study, we developed 12 distinct regressors conduct model. To study correlation between factors, conducted comprehensive analysis selected blast furnace slag, superplasticizer, age, cement, water as optimized factor subset. Based on foundation, grid search fivefold cross-validation were employed establish hyperparameters each results indicate that Deepforest-based model demonstrates superior performance compared models. For more evaluation model's performance, its with state-of-the-art models using same independent dataset. demonstrate our achieving highest (R2 0.91), indicating accurate prediction capability

Язык: Английский

---