Issue |
E3S Web Conf.
Volume 497, 2024
5th International Conference on Energetics, Civil and Agricultural Engineering (ICECAE 2024)
|
|
---|---|---|
Article Number | 02011 | |
Number of page(s) | 12 | |
Section | Civil Engineering | |
DOI | https://doi.org/10.1051/e3sconf/202449702011 | |
Published online | 07 March 2024 |
A Non-destructive Method for the determination of Carbonation Time for Nominal Concrete Cover Depth Using Non-Linear En-semble Prediction
1 Institute of Sustainable Built Environment, Heriot-Watt University , United Kingdom
2 Kano University of Science and Technology, Kano, Nigeria
3 Bahçeşehir Cyprus University, Nicosia, N. Cyprus, via Mersin 0, Turkey
* Corresponding author: sim2000@hw.ac.uk
Carbonation, a process involving the reaction of carbon dioxide and moisture, results in the for-mation of powdery calcium carbonate, a critical durability issue causing reinforcement corrosion. The study analyzed carbonation data from coastal and inland buildings in the Turkish Republic of Northern Cyprus, re-vealing higher carbonation rates than anticipated within their lifespan. An artificial intelligence model named Support Vector Machine (SVM) was applied to predict carbonation time (T) to penetrate concrete cover of 25mm in the TRNC. Subsequently used two ensemble techniques, namely Neural Network Ensembles (NNE) and Support Vector Machine Ensembles (SVME) to enhance the performance of the prediction of T. Four performance criteria namely Correlation Coefficient (CC), Root Mean Square Error (RMSE), Correlation Co-efficient (R2), Mean Absolute Error (MAE) was applied to verify the modelling accuracy. The Values of R2 of Ensemble techniques indicated significant increase in the performance, greater than the SVM model. This shows that using ensemble techniques is promising in getting better predictions of carbonation time (T) to penetrate concrete cover. The results obtained showed that NNE and SVME combination demonstrated the best performance under the evaluation criteria of R2 = 0.8721 and R2 = 0.8644 in testing phases respectively in comparison SVM-M1 to SVM-M3.
© The Authors, published by EDP Sciences, 2024
This is an Open Access article distributed under the terms of the Creative Commons Attribution License 4.0, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.