Heat transfer analysis using ANNs with experimental data for air flowing in corrugated channels

doi:10.1016/j.ijheatmasstransfer.2003.07.031

International Journal of Heat and Mass Transfer

Volume 47, Issues 6–7, March 2004, Pages 1361-1365

https://doi.org/10.1016/j.ijheatmasstransfer.2003.07.031 Get rights and content

Abstract

The objective of this work is to use artificial neural networks (ANNs) for heat transfer analysis in corrugated channels. A data set evaluated experimentally is prepared for processing with the use of neural networks. Back propagation algorithm, the most common learning method for ANNs, was used in training and testing the network. To solve this algorithm a computer program using C++ has been developed. The accuracy between experimental and ANNs approach results was achieved with a mean absolute relative error less than 4%.

Introduction

Artificial neural networks (ANNs) have been used in many engineering applications because of providing better and more reasonable solutions [1], [2]. Some examples are: analysis of thermosyphon solar water heaters, heat transfer data analysis, HVAC computations and prediction of critical heat flux [3]. Sreekanth et al. [4] for evaluation of surface heat transfer coefficient at the liquid–solid interface, Diaz et al. [5] for simulation of heat exchanger performance, Kalogirou [6] used ANNs for performance prediction of forced circulation type solar domestic water heating. Singh et al. modeled the entire flow field around an automobile using ANNs and Schreck et al. used ANNs models to predict the unsteady separated flow field on a wing [4]. Farshad et al. [7] for predicting temperature profiles in producing oil wells used an artificial neural network algorithm. Recently, Parcheco-Vega et al. [3], [8] modeled the heat transfer phenomena in heat exchanger systems using neural network. In addition, same authors performed a simulation of the time-dependent behavior of a heat exchanger [9]. Boccaletti et al. [10], for simulation of gas turbine with a waste heat recovery section, Bechtler et al. [11], to model the steady-state performance of a vapor-compression liquid heat pump, and Sablani [12], for non-iterative calculation of heat transfer coefficient in fluid-particle systems, used ANNs. It should be understood from the literature review mentioned above that ANNs better serve to thermal analysis in engineering applications. However, the ANNs methods have not been used or tested for heat transfer analysis in corrugated channels yet. For this reason, the study was focused on the applicability of ANNs method for heat transfer analysis in corrugated channels, employed in the design of plate heat exchangers because of achieving enhanced heat transfer, and the best candidate for high heat flux applications.

Section snippets

Experimental details

A schematic diagram of the experimental apparatus used for the heat transfer analysis in this study for data gathering are presented in Fig. 1. A detailed presentation of the design, fabrication of the experimental apparatus and evaluation method of the data are available in [13], [14], [15], [16].

Air from the laboratory room, the working fluid, was drawn through the systems by a downstream fan that was supplied the electrical power by the autotransformer to use it as variable speed. The mass

Artificial neural networks approach

ANNs consisting of very simple and highly interconnected processors called neuron are a computational structure inspired by biological neural systems. The processors are analogous to biological neurons in human brain. The neurons are connected to each other by weighted links over which signals can pass. Each neuron receives multiple inputs from other neurons in proportion to their connection weights and generates a single output which may be propagated to several other neurons [4].

Among the

Conclusion

In this study, ANNs model was developed for the analysis of heat transfer. Results indicate that the ANNs model can be trained to provide satisfactory estimations of Nusselt numbers for air flowing in corrugated channels. It should be advised that in preliminary engineering studies, the networks can be used an easy-to-use tool for engineers.

References (22)

T.Y. Lin et al.
Optimum design for artificial neural networks: an example in a bicycle derailleur system
Eng. Appl. Artif. Intell.
(2000)
K. Genel et al.
Boriding response of AISI W1 steel and use of artificial neural network for prediction of borided layer properties
Surf. Coat. Technol.
(2002)
S.A. Kalogirou
Long-term performance prediction of forced circulation solar domestic water heating systems using artificial neural networks
Appl. Energy
(2000)
G. Diaz et al.
Dynamic prediction and control of heat exchangers using artificial neural networks
Int. J. Heat Mass Transfer
(2001)
S.S. Sablani
A neural network approach for non-iterative calculation of heat transfer coefficient in fluid-particle systems
Chem. Eng. Process.
(2001)
E.M. Sparrow et al.
Effect of rounding of protruding edges on heat transfer and pressure drop in a duct
Int. J. Heat Mass Transfer
(1984)
W. Huang et al.
Neural network modeling of salinity variation in Apalachicola River
Water Res.
(2002)
G.E. Nasr et al.
Back propagation neural networks for modeling gasoline consumption
Energy Convers. Manage.
(2003)
A. Parcheco-Vega et al.
Neural network analysis of fin-tube refrigerating heat exchanger with limited experimental data
Int. J. Heat Mass Transfer
(2001)
S. Sreekanth et al.
A neural network approach for evaluation of surface heat transfer coefficient
J. Food Process. Preserv.
(1999)

G. Diaz et al.

Simulation of heat exchanger performance by artificial neural networks

HVAC Res.

(1999)

Cited by (114)

Convection heat transfer augmentation inside a hybrid corrugated channel by varying the positions of corrugated configurations
2023, Case Studies in Thermal Engineering
A numerical investigation was carried out to evaluate the thermo-hydraulic characteristics of a corrugated channel utilizing both single and hybrid corrugated configurations. The single corrugation configurations consisted of trapezoidal and saw-tooth corrugations, denoted as Single Corrugation-1 (SC1) and Single Corrugation-2 (SC2). In addition, the hybrid corrugated configurations included four distinct arrangements, each combining trapezoidal, saw-tooth, and triangular cavities in different ways, represented as Hybrid Corrugation-1 (HC1), Hybrid Corrugation-2 (HC2), Hybrid Corrugation-3 (HC3), and Hybrid Corrugation-4 (HC4). The flow was considered steady, two-dimensional, laminar with Reynolds numbers (Re = 500–900), while a heat flux of 2000 W/m² was applied along the corrugated wall. The findings demonstrated that the flow and temperature contours were significantly influenced by both single and hybrid corrugated configurations. All hybrid corrugations showed Nusselt number (Nu) enhancements of 20%–36% compared to SC1 and 40%–60% compared to SC2 at Re = 900. Moreover, hybrid corrugated configurations showed better Thermal Hydraulic Performance (THP) compared to single corrugations. At Re = 900, the THP values for HC1, HC2, HC3, and HC4 were 0.95, 1.12, 1.05, and 1.02 compared to SC1, and 1.26, 1.49, 1.39, and 1.35 compared to SC2. Notably, among all the cases, HC2 demonstrated the most significant increase in heat transfer.
Comparative study on different cooling techniques for photovoltaic thermal management: Hollow fins, wavy channel and insertion of porous object with hybrid nanofluids
2023, Applied Thermal Engineering
The present work considers performance of different cooling methods applicable to photovoltaic (PV) panels. Hollow fins and two different channel cooling systems are considered. Flat and hollow fins types are used while cooling channels with wavy wall and with porous layer insert are compared. Effects of different fin numbers (between 2 and 16), hollow fin radius (between 0.8h and 3.2h, h-panel height), Reynolds number (between 100 and 500), wavy channel amplitude (between 0.01H and 0.5H, H-channel height), porous insert permeability (Darcy number between $1 0^{- 6}$ and $1 0^{- 1}$ ), porous layer size (between 0.1L and 0.7L, L-channel length) and nanoparticle volume fraction (between 0 and 0.02) on the cooling performance are numerically assessed. When nanofluid is used in the cooling channels, additional improvements in the cooling performance are obtained while temperature drop of 2.5 °C is achieved at the highest loading. When different cooling systems are compared, channel cooling with porous layer and wavy walls provide the highest cooling performance followed by hollow fin and flat fin systems while 37.6 °C , 37 °C, 33.6 °C and 29 °C temperature drops are obtained as compared to reference panel case without cooling system. When compared to flat fins, hollow fins perform better and additional temperature drop of 8.5 °C can be obtained by using 16 number of fins. It also provides a lightweight design as less material is used. Artificial neural networks are successfully used for performance estimation of wavy cooling channel system with 15 neurons in the hidden layer. The outcomes of this work is useful in further optimization and system development of PV cooling technologies as the need for thermal management in renewable energy sources is increasing due to the energy cost and environmental side concerns.
A universal correlation for flow condensation heat transfer in horizontal tubes based on machine learning
2023, International Journal of Thermal Sciences
Accurate and universal prediction of in-tube condensation heat transfer coefficients (HTCs) is vital for designing compact condensers. This study presents machine learning (ML) methods for predicting flow condensing HTCs inside horizontal tubes based on an assembled database. The database contains 6064 experimental data of 28 pure fluids and covers broad operating conditions. Furthermore, the database is serviced to train and evaluate five ML models based on K-nearest neighbors (KNN), artificial neural network (ANN), convolutional neural network (CNN), random forest (RF), and Extreme gradient boosting (XGBoost) algorithms. The designed ML models show excellent predictive performances for 1213 test data points with the best mean absolute relative deviation (MARD) of 5.82% achieved by CNN and the coefficient of determination (R²) of 0.98 or higher for both XGBoost models. Based on the parametric importance analysis conducted by the trained XGBoost models, a new universal correlation is developed by incorporating several key parameters to characterize condensation heat transfer. This correlation shows reliable predictions for all data points in the consolidated database with the mean relative deviation (MRD) of −1.77% and MARD of 19.21%. An additional excluded database is collected to validate the universality of the new correlation and these ML models. It illustrates that ML models are effective predictive tools for HTCs, and can also help develop the heat transfer correlation with high accuracy and universality.
Air side performance characterization of wavy Fin-and-tube heat exchangers having elliptic tubes with large waffle heights
2022, Applied Thermal Engineering
Demand of less energy consumption and higher energy efficiency suggests increasing heat transfer performance of a fin-and-tube heat exchanger. This paper experimentally and numerically examines the air side performance of wavy fin-and-tube heat exchangers (FTHXs) with elliptic tubes. In a properly regulated and evenly distributed airflow supply loop, the inlet/outlet temperatures and pressure drops across the test section are obtained using two measuring meshes and a differential pressure transducer to resolve thermal and frictional outcomes in terms of the heat transfer coefficient and pressure drop. The theoretical formulation considers the thermo-fluid analysis of crossing airflows over the wavy finned tube heat exchanger. In this research, the test heat exchanger contains elliptic tubes incorporated with 4-row wavy fins, while the proposed design adopts a large setting of 3.24-mm waffle height and 3.02-mm fin pitch, respectively. The predicted heat transfer coefficients and pressure drops are in good agreement with the measured data at varied inlet air velocities to validate the CFD model. Numerical simulations are then extended to investigate the influences of waffle height, fin pitch, wave length and inlet air velocity on thermofluid characteristics of FTHXs. Six common correlations of Colburn factors (j) and friction factor (f) factors are compared with predictions to evaluate their applicabilities for reasonable estimates of thermal and frictional performance of FTHXs. It is noticed that the correlation by Wang et al. can achieve the most accurate results in calculating the j and f factors.
Machine-learning-aided design optimization of internal flow channel cross-sections
2022, International Journal of Heat and Mass Transfer
The design optimization of various thermal management components such as cold plates, heat sinks, and heat exchangers relies on accurate prediction of flow heat transfer and pressure drop. During the iterative design process, the heat transfer and pressure drop is typically either computed numerically or obtained using geometry-specific correlations for Nusselt number and friction factor. Numerical approaches are accurate for evaluation of a single design but become computationally expensive if many design iterations are required (such as during formal optimization processes). Correlation-based approaches restrict the design space to a specific set of geometries for which correlations are available. Surrogate models for the Nusselt number and friction factor, which are more universally applicable to all geometries than traditional correlations, would enable flexible and computationally inexpensive design optimization. The current work develops machine-learning-based surrogate models for predicting the Nusselt number and friction factor under fully developed internal flow in channels of arbitrary cross section and demonstrates use of these models for optimization of the cross-sectional channel shape. The predictive performance and generality of the machine learning surrogate models is first verified on various shapes outside the training dataset, and then the models are used in the design optimization of flow cross sections based on performance metrics that weigh both heat transfer and pressure drop. The optimization process leads to novel shapes outside the training data, and so numerical simulations are carried out on these optimized shapes to compare with the surrogate model predictions and show their performance is at least as good as that of shapes with known correlations available. A three-lobed shape was found to reduce friction factor, whereas a pentagon with rounded corners and an ice cream cone-shaped duct, both found using different performance metrics. Although the ML model predictions lose accuracy outside the training set for these novel shapes, the predictions follow the correct trends with parametric variations of the shape and therefore successfully direct the search toward optimized shapes.
Prediction of Trypanosoma evansi infection in dromedaries using artificial neural network (ANN)
2022, Veterinary Parasitology
Surra is caused by Trypanosoma evansi, a flagellated parasite that affects domestic and wild animals. Surra is a neglected tropical disease causing serious problems to camels breed in Algeria. The aim of our study consists to extract the major risk factors that predict T.evansi infection in dromedaries using artificial neural networks. This investigation was conducted on 115 dromedaries from Ghardaïa district, Southern Algeria. The immune trypanolysis test was used to detect antibodies against T. evansi. Firstly, the gamma test has been used to choose optimal input parameters. The obtained results indicate that the age, gender, breed, clinical manifestations history, herd size, as well as the animal activities were the most predictors of T. evansi infection. Afterward, an artificial neural network method has been performed for modelling the proposed optimal inputs and their accuracy was assessed through seven statistical indicators. The comparative study indicates the effectiveness of the (6−9−1) model trained by the Tansig transfer function. The proposed model has demonstrated a good performance: 0.925 for training data and 0.962 for validation data. Furthermore it could be very useful for the rapid intervention of veterinarians as close as possible to the point-of-care (POC).

View all citing articles on Scopus

View full text

Heat transfer analysis using ANNs with experimental data for air flowing in corrugated channels

Abstract

Introduction

Section snippets

Experimental details

Artificial neural networks approach

Conclusion

Eng. Appl. Artif. Intell.

Surf. Coat. Technol.

Appl. Energy

Int. J. Heat Mass Transfer

Chem. Eng. Process.

Int. J. Heat Mass Transfer

Water Res.

Energy Convers. Manage.

Neural network analysis of fin-tube refrigerating heat exchanger with limited experimental data

Int. J. Heat Mass Transfer

A neural network approach for evaluation of surface heat transfer coefficient

J. Food Process. Preserv.

Simulation of heat exchanger performance by artificial neural networks

HVAC Res.