Artificial neural network-based forecasting of air temperature reduction in a tube bank under electric field
This study aims to investigate the potential of Artificial Neural Networks (ANN) in predicting the heat transfer behavior of air as it flows through a tube bank with low temperature and operates under an electric field. To date, no previous studies have applied ANN for predicting heat transfer chara...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-06-01
|
| Series: | Case Studies in Thermal Engineering |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2214157X25003466 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850195803659829248 |
|---|---|
| author | Ekuong Tang Attakorn Asanakham Thoranis Deethayat Tanongkiat Kiatsiriroat |
| author_facet | Ekuong Tang Attakorn Asanakham Thoranis Deethayat Tanongkiat Kiatsiriroat |
| author_sort | Ekuong Tang |
| collection | DOAJ |
| description | This study aims to investigate the potential of Artificial Neural Networks (ANN) in predicting the heat transfer behavior of air as it flows through a tube bank with low temperature and operates under an electric field. To date, no previous studies have applied ANN for predicting heat transfer characteristics under these specific conditions. ANN is employed to address the complexities in heat exchanger systems, as it effectively handles non-linear relationships and multiple variables, allowing for accurate predictions of air temperature and heat transfer rates across each column in the tube bank. The tube bank used in this study consisted of 10 rows of tubes with a diameter of 15 mm, a pitch ratio of 1–4, an inlet air temperature of 40–45 °C, a surface temperature of 30–40 °C, an air velocity of 0.1–0.25 m/s, and high voltage DC ranging from 7 to 15 kVDC. The heat transfer rate and outlet air temperature were the output results. The study evaluated the optimal ANN model structure, including the appropriate number of neurons in the hidden layer, the number of epochs, and the suitable activation function for predicting heat transfer performance under an electric field. The combination of the Sigmoid-Purelin activation function with 8 neurons and 30 epochs proved to be the most suitable for this study. The ANN model demonstrated exceptional accuracy in predicting performance across a variety of conditions for each row in the tube bank. Finally, correlations were developed between the temperature ratio and the maximum Reynolds number at different S/D ratios, both with and without the electric field. The results computed using ANN closely matched theoretical heat transfer predictions, with R-squared values greater than 0.99 and MSE less than 0.000025. Furthermore, when the ANN model was tested beyond the training data conditions, the predicted results still exhibited a high level of accuracy compared to the theoretical heat transfer calculations. |
| format | Article |
| id | doaj-art-b9be754a36fd4ceea8b6350becc0d2a9 |
| institution | OA Journals |
| issn | 2214-157X |
| language | English |
| publishDate | 2025-06-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Case Studies in Thermal Engineering |
| spelling | doaj-art-b9be754a36fd4ceea8b6350becc0d2a92025-08-20T02:13:40ZengElsevierCase Studies in Thermal Engineering2214-157X2025-06-017010608610.1016/j.csite.2025.106086Artificial neural network-based forecasting of air temperature reduction in a tube bank under electric fieldEkuong Tang0Attakorn Asanakham1Thoranis Deethayat2Tanongkiat Kiatsiriroat3Energy Engineering Program, Faculty of Engineering and Graduate School, Chiang Mai University, 239 Huay Kaew Road, Muang District, Chiang Mai, 50200, ThailandThermal System Research Laboratory, Department of Mechanical Engineering, Faculty of Engineering, Chiang Mai University, 239 Huay Kaew Road, Muang District, Chiang Mai, 50200, Thailand; Research Group for Renewable Energy, Faculty of Engineering, Chiang Mai University, 239 Huay Kaew Road, Muang District, Chiang Mai, 50200, ThailandThermal System Research Laboratory, Department of Mechanical Engineering, Faculty of Engineering, Chiang Mai University, 239 Huay Kaew Road, Muang District, Chiang Mai, 50200, Thailand; Research Group for Renewable Energy, Faculty of Engineering, Chiang Mai University, 239 Huay Kaew Road, Muang District, Chiang Mai, 50200, ThailandThermal System Research Laboratory, Department of Mechanical Engineering, Faculty of Engineering, Chiang Mai University, 239 Huay Kaew Road, Muang District, Chiang Mai, 50200, Thailand; Research Group for Renewable Energy, Faculty of Engineering, Chiang Mai University, 239 Huay Kaew Road, Muang District, Chiang Mai, 50200, Thailand; Corresponding author. Thermal System Research Laboratory, Department of Mechanical Engineering, Faculty of Engineering, Chiang Mai University, 239 Huay Kaew Road, Muang District, Chiang Mai, 50200, Thailand.This study aims to investigate the potential of Artificial Neural Networks (ANN) in predicting the heat transfer behavior of air as it flows through a tube bank with low temperature and operates under an electric field. To date, no previous studies have applied ANN for predicting heat transfer characteristics under these specific conditions. ANN is employed to address the complexities in heat exchanger systems, as it effectively handles non-linear relationships and multiple variables, allowing for accurate predictions of air temperature and heat transfer rates across each column in the tube bank. The tube bank used in this study consisted of 10 rows of tubes with a diameter of 15 mm, a pitch ratio of 1–4, an inlet air temperature of 40–45 °C, a surface temperature of 30–40 °C, an air velocity of 0.1–0.25 m/s, and high voltage DC ranging from 7 to 15 kVDC. The heat transfer rate and outlet air temperature were the output results. The study evaluated the optimal ANN model structure, including the appropriate number of neurons in the hidden layer, the number of epochs, and the suitable activation function for predicting heat transfer performance under an electric field. The combination of the Sigmoid-Purelin activation function with 8 neurons and 30 epochs proved to be the most suitable for this study. The ANN model demonstrated exceptional accuracy in predicting performance across a variety of conditions for each row in the tube bank. Finally, correlations were developed between the temperature ratio and the maximum Reynolds number at different S/D ratios, both with and without the electric field. The results computed using ANN closely matched theoretical heat transfer predictions, with R-squared values greater than 0.99 and MSE less than 0.000025. Furthermore, when the ANN model was tested beyond the training data conditions, the predicted results still exhibited a high level of accuracy compared to the theoretical heat transfer calculations.http://www.sciencedirect.com/science/article/pii/S2214157X25003466Artificial neural networkTube bank heat exchangerElectric field |
| spellingShingle | Ekuong Tang Attakorn Asanakham Thoranis Deethayat Tanongkiat Kiatsiriroat Artificial neural network-based forecasting of air temperature reduction in a tube bank under electric field Case Studies in Thermal Engineering Artificial neural network Tube bank heat exchanger Electric field |
| title | Artificial neural network-based forecasting of air temperature reduction in a tube bank under electric field |
| title_full | Artificial neural network-based forecasting of air temperature reduction in a tube bank under electric field |
| title_fullStr | Artificial neural network-based forecasting of air temperature reduction in a tube bank under electric field |
| title_full_unstemmed | Artificial neural network-based forecasting of air temperature reduction in a tube bank under electric field |
| title_short | Artificial neural network-based forecasting of air temperature reduction in a tube bank under electric field |
| title_sort | artificial neural network based forecasting of air temperature reduction in a tube bank under electric field |
| topic | Artificial neural network Tube bank heat exchanger Electric field |
| url | http://www.sciencedirect.com/science/article/pii/S2214157X25003466 |
| work_keys_str_mv | AT ekuongtang artificialneuralnetworkbasedforecastingofairtemperaturereductioninatubebankunderelectricfield AT attakornasanakham artificialneuralnetworkbasedforecastingofairtemperaturereductioninatubebankunderelectricfield AT thoranisdeethayat artificialneuralnetworkbasedforecastingofairtemperaturereductioninatubebankunderelectricfield AT tanongkiatkiatsiriroat artificialneuralnetworkbasedforecastingofairtemperaturereductioninatubebankunderelectricfield |