Enhancing the Design of Experiments on the Fatigue Life Characterisation of Fibre-Reinforced Plastics by Incorporating Artificial Neural Networks DOI Open Access
Christian Witzgall, Moh’d Sami Ashhab, Sandro Wartzack

et al.

Materials, Journal Year: 2024, Volume and Issue: 17(3), P. 729 - 729

Published: Feb. 3, 2024

Fatigue life testing is a complex and costly matter, especially in the case of fibre-reinforced thermoplastics, where other parameters addition to force alone must be taken into account. The number tests required therefore increases significantly, if influence different fibre orientations It important gain greatest possible amount knowledge from limited available tests. In order achieve this, this study aims utilise adaptive sampling, which used numerous areas computational engineering, for design experiments on fatigue testing. Artificial neural networks (ANNs) are trained data short-fibre-reinforced material PBT GF30, their model uncertainty queried. This was undertaken with ANNs various numbers hidden layers, were analysed performance. ideal turned out four squared error as small 1 × 10−3 recorded. Locally resolved, ANN identify region samples vertical orientation cycles. With information such additional can obtained uncertain regions improve prediction—almost halving recorded only 0.55 10−3. way, comparable value found less experimental effort, or better quality set up same effort.

Language: Английский

Prediction of the Compressive Strength of Vibrocentrifuged Concrete Using Machine Learning Methods DOI Creative Commons
Alexey N. Beskopylny, Sergey A. Stel’makh, Evgenii M. Shcherban’

et al.

Buildings, Journal Year: 2024, Volume and Issue: 14(2), P. 377 - 377

Published: Feb. 1, 2024

The determination of mechanical properties for different building materials is a highly relevant and practical field application machine learning (ML) techniques within the construction sector. When working with vibrocentrifuged concrete products structures, it crucial to consider factors related impact aggressive environments. Artificial intelligence methods can enhance prediction through use specialized algorithms materials’ strength determination. aim this article establish evaluate algorithms, specifically Linear Regression (LR), Support Vector (SVR), Random Forest (RF), CatBoost (CB), compressive in under diverse operational conditions. This achieved by utilizing comprehensive database experimental values obtained laboratory settings. following metrics were used analyze accuracy constructed regression models: Mean Absolute Error (MAE), Squared (MSE), Root-Mean-Square (RMSE), Percentage (MAPE) coefficient (R2). average MAPE range from 2% (RF, CB) 7% (LR, SVR) allowed us draw conclusions about possibility using “smart” development compositions quality control concrete, which ultimately entails improvement acceleration manufacture. best model, CatBoost, showed MAE = 0.89, MSE 4.37, RMSE 2.09, R2 0.94.

Language: Английский

Citations

10
The Influence of Materials on the Mechanical Properties of Ultra-High-Performance Concrete (UHPC): A Literature Review DOI Open Access

Mariana Lage da Silva,

Lisiane Pereira Prado, Emerson Felipe Félix

et al.

Materials, Journal Year: 2024, Volume and Issue: 17(8), P. 1801 - 1801

Published: April 14, 2024

Ultra-high-performance concrete (UHPC) is a cementitious composite combining high-strength matrix and fiber reinforcement. Standing out for its excellent mechanical properties durability, this material has been widely recognized as viable choice highly complex engineering projects. This paper proposes (i) the review of influence exerted by constituent materials on compressive strength, flexural tensile elastic modulus UHPC (ii) determination optimal quantities based simplified statistical analyses developed database. The data search was restricted to papers that produced with straight steel fibers at content 2% volume. mixture models were proposed graphical relationship versus properties, aiming optimize material’s performance each property. results proved be in accordance specifications present literature, characterized high cement consumption, significant presence fine materials, low water-to-binder ratio. divergences identified between mixtures reflect how uniquely impact property concrete. In general, shown play role increasing strength UHPC, while water superplasticizers stood their workability.

Language: Английский

Citations

9
Concrete Compressive Strength Prediction Using Combined Non-Destructive Methods: A Calibration Procedure Using Preexisting Conversion Models Based on Gaussian Process Regression DOI Open Access
Giovanni Angiulli, Salvatore Calcagno, Fabio La Foresta

et al.

Journal of Composites Science, Journal Year: 2024, Volume and Issue: 8(8), P. 300 - 300

Published: Aug. 1, 2024

Non-destructive testing (NDT) techniques are crucial in making informed decisions about reconstructing or repairing building structures. The SonReb method, a combination of the rebound hammer (RH) and ultrasonic pulse velocity (UPV) tests, is widely used for this purpose. To evaluate compressive strength, CS, concrete under investigation, Vp index R must be mapped to strength CS using suitable conversion model, identification which requires supplementing NDT measurements with destructive-type (DT) on relatively large number cores. An approach notably indicated all cases where minimization cores essential employ pre-existing i.e., model derived from previous studies conducted literature, appropriately calibrated. In paper, we investigate performance Gaussian process regression (GPR) calibrating models, exploiting their ability handle nonlinearity uncertainties. numerical results obtained experimental data collected literature show that GPR calibration very effective, outperforming, most cases, standard multiplicative additive calibrate models.

Language: Английский

Citations

3
Multi-objective optimization of the flow condition of binary constituent net-zero concretes towards carbon neutrality-built environment pathway DOI
César García, Kennedy C. Onyelowe, Paulina Elizabeth Valverde Aguirre

et al.

Journal of Building Pathology and Rehabilitation, Journal Year: 2024, Volume and Issue: 9(1)

Published: April 12, 2024

Language: Английский

Citations

1
Integrating machine learning and Monte Carlo Simulation for probabilistic assessment of durability in RC structures affected by carbonation-induced corrosion DOI
Emerson Felipe Félix,

Breno M. Lavinicki,

Tobias L. G. T. Bueno

et al.

Journal of Building Pathology and Rehabilitation, Journal Year: 2024, Volume and Issue: 9(2)

Published: Sept. 19, 2024

Language: Английский

Citations

1
Enhancing the predictive accuracy of recycled aggregate concrete’s strength using machine learning and statistical approaches: a review DOI
Jawad Tariq,

Kui Hu,

Syed Tafheem Abbas Gillani

et al.

Asian Journal of Civil Engineering, Journal Year: 2024, Volume and Issue: unknown

Published: Oct. 17, 2024

Language: Английский

Citations

1
Enhancing the Design of Experiments on the Fatigue Life Characterisation of Fibre-Reinforced Plastics by Incorporating Artificial Neural Networks DOI Open Access
Christian Witzgall, Moh’d Sami Ashhab, Sandro Wartzack

et al.

Materials, Journal Year: 2024, Volume and Issue: 17(3), P. 729 - 729

Published: Feb. 3, 2024

Fatigue life testing is a complex and costly matter, especially in the case of fibre-reinforced thermoplastics, where other parameters addition to force alone must be taken into account. The number tests required therefore increases significantly, if influence different fibre orientations It important gain greatest possible amount knowledge from limited available tests. In order achieve this, this study aims utilise adaptive sampling, which used numerous areas computational engineering, for design experiments on fatigue testing. Artificial neural networks (ANNs) are trained data short-fibre-reinforced material PBT GF30, their model uncertainty queried. This was undertaken with ANNs various numbers hidden layers, were analysed performance. ideal turned out four squared error as small 1 × 10−3 recorded. Locally resolved, ANN identify region samples vertical orientation cycles. With information such additional can obtained uncertain regions improve prediction—almost halving recorded only 0.55 10−3. way, comparable value found less experimental effort, or better quality set up same effort.

Language: Английский

Citations

1