No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Application of an Artificial Neural Network (ANN) for predicting low-GWP refrigerant boiling heat transfer inside Brazed Plate Heat Exchangers (BPHE)
Abstract
This paper presents an Artificial Neural Network (ANN) model for predicting refrigerant boiling heat transfer coefficients inside Brazed Plate Heat Exchangers (BPHE). The model accounts for the effect of plate geometry, operating conditions and refrigerant properties. The model shows a fair agreement with a database of 1760 data points comprising 15 plate geometries and 16 refrigerants (including 4 natural refrigerants and 6 other low-GWP refrigerants). The Mean Absolute Percentage Error (MAPE) of the model predictions is 4.8%. The ANN model exhibits a better predictive capability than most of the state-of-the-art analytical-computational procedures for boiling inside BPHE available in the open literature. The characteristic parameters of the ANN model are fully reported in the paper.
... Some application subjects of predictive quality systems are proposed in the literature. Such as, the deep drawing manufacturing process of car body parts [8], tool flank wears at a turning operation [9], refrigerant brazed plate heat exchangers (BPHE) [10], battery cells production [11]. Moreover, we can add topics to consider like crankshaft production line [12], rare quality event detection, ultrasonic metal welding of battery tabs, sensorless drive diagnosis [14]. ...
... Digitalization needs specific subjects for using machine learning algorithms aspect of predictive quality applications. Also, these special concepts need a wide range and different type of variables such as; cutting speed (rpm), feed rate (mm/rev), depth of cut (mm), lubrication variables [9], plate geometry, operating conditions [10], x-ray inspection with height, % shape 2D, % shape 3D, % surface, % volume, % offset X μm, offset Y μm [13], flange retraction laser data, strain gauge sensory data, signal data, the occurrence of process failures [8], etc. ...
... Predictive Quality models and methods in Machine Learning (ML) algorithms are used in the literature can be listed as follows. Adaptive Neuro-Fuzzy Approach (ANFIS) [9], Artificial Neural Network (ANN) [8,10,11,15], Lasso-Lars Regression ...
Industry 4.0 is generally defined as a development system that compels the digitalization of processes to create integrated and autonomous systems. The process tracking of parts is very important in terms of detecting missed faulty products. Some defects that escape from quality control directly affect the end-user. Machine learning algorithms have been used to predict changes in the quality control processes and defective products, toward real-time and effective data processing. Thus, the highest quality of the final product will be delivered to the customer and to reduce the defective production coming out of the manufacturing chain. In this article, the study aims to establish a predictive quality model that can detect defect-free approved but faulty products overlooked during the quality inspection operations. Machine learning methods are used to analyze the relationship between quality control data and customer complaints. For this purpose, we use the last quality stage data of an automobile manufacturer’s brake system from 2018 to 2020. Machine learning models are constructed using logistic regression, ridge regression, support vector machine, random forest classification tree, gradient boost, XGBoost, LightGBM, and CatBoost algorithms. The results of specificity and negative prediction value show that the Gradient Boost and CatBoost algorithms have the best classification benefit for detecting the rare events.
... In addition, various studies were carried out on the optimization of the plate heat exchanger design and operation parameters. Wang et al. [14] presented a method for optimizing the heat exchanger network (HEN) for plate heat exchangers, and Longo et al. [15] presented the artificial neural network (ANN) method for determining the total heat transfer. Fernández-Seara et al. [16] investigated numerous parametric studies on the use of the Wilson method and the modified Wilson method to determine the thermal performance of the plate heat exchanger. ...
... The RANSbased 2-equation k-ε turbulence model was chosen. The 2-equation turbulence model gives more accurate results than other models in defining of both boundary layer and wall functions in heat exchangers [7,15]. In addition, the RNG k-ε model used in this study is a relatively inexpensive two-equation model that describes the eddy viscosity as a function of turbulent kinetic energy and turbulent propagation velocity and is often used in similar applications [6]. ...
Determination of the geometrical parameters of the heat exchanger has important effects on the thermohydraulic performance of the heat exchanger. In this study, the effects of geometric parameters of a plate heat exchanger on thermohydraulic performance have been extensively investigated using computational fluid dynamics (CFD). Parametric studies were performed on 8 different corrugated channel geometries with various chevron angles (β) and aspect ratios (2b/λ) for Reynolds numbers ranging from 500 to 3000. An entire fluid channel was numeri- cally studied using the same mass flow rate and the Reynolds number. As results of the study, temperature distribution, pressure gradient, velocity, turbulent kinetic energy distribution, Nusselt number, friction factor, and flow properties were evaluated comparatively for each case. It was determined that the sinusoidal corrugations promote the turbulence intensity and the swirling flow which leads to thermal boundary layer mitigation and enhanced convection heat transfer in response to the increasing of aspect ratio. The results of the study show that the (CFD) model is a reasonable and effective technique for displaying 3D contour plots, stream- lines, and determining performance parameters.
... Because of the new requirements and applications for plate heat exchangers in terms of enhancing heat exchange and performance, in addition to restrictions on energy consumption and impact on the environment, new geometrical designs for the PHE have emerged that are superior to the chevron plate in terms of thermohydraulic properties of flow under specific operating conditions [14][15][16][17][18]. Longo et al employed the ANN model to forecast heat transfer coefficients during refrigerant condensation within plates of herringbone configuration [19]. Optimal operating conditions for the flows of the working fluids were predicted using the gray-box model by Khan et al [20]. ...
Reducing fuel consumption and toxic gas emissions is a major concern in modern energy research. This paper investigates the performance and heat transfer enhancement of an innovative plate heat exchanger (IPHE) using machine learning techniques. By optimizing the geometric parameters of the plate, we predict thermohydraulic characteristics—represented by the Nusselt number (Nu), coefficient of friction (f), and performance (P) within the Reynolds number range of 500–5000 based on numerical modeling data. This study addresses the need for improved efficiency in plate heat exchangers (PHEs) amid rising energy demands and environmental concerns. Traditional methods like numerical simulations or costly experiments have limitations, prompting interest in artificial intelligence (AI) and machine learning (ML) for thermal analysis and property prediction in PHEs. Various ML models, including Decision Trees, XGBoost, Gradient Boosting, and ensemble methods, are evaluated in predicting f, Nu, and overall performance (P). Our comprehensive experimentation and analysis identify top-performing models with robust predictive capabilities. For f, the highest R² score was 0.98, indicating excellent prediction accuracy, with mean squared error (MSE) values consistently below 0.0016. Similarly, for Nu and P, top models achieved R² scores of 0.979 and 0.9628, respectively, with MSE values below 0.0347 and 0.05. These results highlight the effectiveness of machine learning techniques in accurately predicting thermohydraulic properties and optimizing PHE performance.
... The authors trained their algorithms based on visualisation data to infer the heat flux from nucleate boiling experiments. Longo et al. [66,67,68] demonstrated the interesting capabilities of the use of ML algorithms to predict both the condensation and boiling heat transfer coefficients and the two-phase pressure drops inside brazed plate heat exchangers. Several other applications also exist, which can be found in the review article by Hughes et al. [23]. ...
... Longo et al. [14] used an ANN model to predict refrigerant boiling heat transfer coefficients inside Brazed Plate Heat Exchangers (BPHE). The input parameters were corrugation enlargement ratio, reduced inclination angle, liquid Prandtl number, equivalent Reynolds number, boiling number, and reduced pressure. ...
Artificial neural network (ANN) methods have been gradually used in the field of nuclear reactor thermal-hydraulics as new methods to improve accuracy or fast prediction. This study establishes a back propagation (BP) neural network model based on the ANN methods to predict the steam condensation heat transfer coefficient outside a heat tube in the presence of air. The main factors affecting condensing heat transfer, such as pressure, air mass fraction, subcooling, and tube diameter, are used as input quantities, and the condensation heat transfer coefficient is used as output quantity. A complete set of neural networks for predicting the heat transfer coefficient for steam condensation in the presence of air is established based on the relevant experimental data collected over the world. The results predicted by the ANN model are compared with the experimental data and those of traditional correlation methods. The data from 2276 experiments are distributed within a ±10% error band at the 95% confidence level. This means that the prediction accuracy of the ANN model is higher than that of the traditional experimental correlation. Therefore, the neural network model developed in this study can be used for the prediction of the heat transfer coefficient for steam condensation in the presence of air.
... A large database of LGWP refrigerants boiling inside brazed PHEs were collected by Longo et al. [108] who also developed an ANN model to estimate the refrigerant boiling HTCs inside a brazed PHE with a low MAPE. Their research aimed to precisely predict the behavior of other LGWP refrigerants, whose thermophysical and thermodynamic properties have not yet been well characterized. ...
This review presents the nucleate/convective boiling performance for a variety of important low global warming potential (LGWP) alternatives to commonly used high-global warming potential (GWP) refrigerants (such as R-134a, R404A, and R-410A, etc.). Efforts are stressed on the assessment of their evaporation pressure drop and heat transfer coefficient (HTC) characteristics. These alternatives include R-1234ze(Z), R-1234ze(E), R-1233zd (E), R-1234ze(E), R-410A, R-1234yf, and R-513A. The authors investigated the thermo-fluid properties within and outside a tube, mini-channel, micro-fin tube, and plate heat exchanger. The investigation of the numerical, experimental, and simulated results revealed that the evaporation pressure drop and HTC characteristics were dependent on a variety of variables. These factors include the working fluid’s thermodynamics and transport properties, the refrigerant’s mass flux, heat flux, saturation temperature, the vapor quality, the conditions and flow patterns, the orientation of the heating surface, and the geometry (shape, size, and surface area smooth/enhanced) of the heating surface. An expanded LGWP refrigerants, surfaces, and conditions database is needed. Mechanistic models may assist. These models can optimize boiling, anticipate heat transfer, and develop high-performance geometries.
... Pendekatan ANN sangat cocok untuk aplikasi pengenalan pola spektrum getaran. Arsitektur propagasi mundur dipilih karena secara simultan dapat menampilkan berbagai jenis kerusakan motor sebagai keluaran sehingga dapat dilihat identifikasi alternatif berupa faktor korelasi [29], [30]. Konsep berbeda seperti Hebb, Heteroassociative, dan Autoassociative hanya mendefinisikan keluaran yang menyatakan apakah pola yang diuji adalah pola kelas yang disimpan jaringan atau tidak. ...
Expert system design is an effective and sophisticated way of diagnosing a fault in a 12 kW DC Motor. This study aims to design an ANN system to determine damage to the motor. The research method uses spectrum data from the vibration analyzer which is collected based on different types of damage. The training data patterns from the spectrum characteristics to be used in the system, the goal is that the systems can recognize the patterns that have been made. The training data patterns that have been successfully recognized by the system are then tested. The results of training and ANN testing are quite good, with the greatest Cross-Entropy value of 9.94, having 0% error value, the largest Mean Square Error value 8.33e-6 and the smallest regression 0.998. A testing of 8 new spectrums resulted in accurate predictions.
... In such conditions, obtaining thermal and hydraulic correlations is quite difficult or even not feasible due to the requirement of huge computational resources. Hence, machine learning comes into the picture that is reported to predict complex heat transfer and fluid flow characteristics with minimal computational cost [24], [25], [26], [27]. In this context, Behnam et al. [28] and k-nearest neighbors (kNN), were implemented to predict the effect of air injection on the performance of shell and tube heat exchanger. ...
In the current study, a multi-layered approach, combining 3D Reynolds Averaged Navier–Stokes (RANS) model, Artificial Neural Network (ANN), and in-house precooler design and analysis code (PDAC), is developed to investigate the influence of various parameters on the design of the precooler used in sCO2 cycle. The validated 3D-RANS model is utilized to compute thermal and hydraulic characteristics of a zigzag channeled PCHEs operating under the precooler's conditions. The generated CFD data is used to develop empirical correlations and train a machine learning (ML) model based on Artificial Neural Network (ANN). Finally, both trained ML model and empirical correlation are utilized in the PDAC to design and analyze the precooler under different design conditions. Regression data for the precooler, operating in the pseudocritical zone, show that the prediction accuracy of the developed Nusselt number and pressure drop correlation is relatively poor. At the same time, the trained Artificial Neural Network (ANN) can estimate 99% of the data with 90% confidence in the same operating regime. Further, precooler design and analysis code (PDAC) based on the trained ML model can accurately mimic the CFD prediction at a significantly reduced computational time. It is found that pressure losses on the sCO2-side and water pumping power are exceptionally sensitive to desired exit temperature and channel flow rate. Exit temperature lower than the pseudo critical temperature and higher channel flow rate result in pinch point location well inside the precooler, resulting in a massive rise in the heat exchanger’s size and pressure losses. The pinch point can be avoided by choosing the channel mass flow rate ratio of the hot and cold-sides fluid comparable with their ratio of specific heat capacities corresponding to pseudocritical temperature.
... Except for the MLR method, the artificial neural network (ANN) method has been widely used for engineering experimental data processing in recent years (Huang et al. 2021). ANN is inspired by biological nervous systems and has been used in many areas, such as engineering applications, medicine, and business for dealing with nonlinearities and complexities (Longo et al. 2020;Ramalingam et al. 2020;Xue et al. 2020). ANN has also become very popular for simulating biodiesel properties (Meng et al. 2014;Miraboutalebi et al. 2016;Piloto-Rodríguez et al. 2013;Zheng et al. 2020). ...
Iodine value (IV) is an indicator to evaluate the degree of unsaturation (DU) of biodiesel. It reflects the biodiesel degradation and oxidation stability (OS) and also has an effect on viscosity, low-temperature flow properties (LTFP), and the combustion performance. To construct a theoretical system for the simultaneous optimization of LTFP and OS of biodiesel using IV, 52 measured experimental data are used to investigate the qualitative and quantitative relationship between IV and biodiesel composition. The relationships between biodiesel physicochemical properties and IV are investigated in this work. The qualitative analysis shows that the poly-unsaturated fatty acid methyl esters (FAMEs) contribute to an increase in IV, whereas saturated and mono-unsaturated FAMEs decrease IV. Multiple linear regression (MLR) and artificial neural network (ANN) are used to estimate IV from FAMEs. The correlation coefficient, root mean squared error (RMSE), and mean absolute percentage error (MAPE) are respectively 0.976, 2.45, and 1.76% for the MLR model and 0.983, 2.14, and 1.57% for the back propagation neural network (BPNN) model; these values indicate the high accuracy of these methods. The performances of the proposed models were compared with three existing IV prediction models and validated by another databank. The results indicate that the application of the developed BPNN model is better and more comprehensive. Additionally, a preliminary conclusion is that biodiesel with a low percentage of both long-chain saturated and poly-unsaturated FAMEs can have solidifying point (SP) and OS in the proper range. Biodiesel with a low IV is generally more combustible and efficient.
... Considering two-phase heat transfer problems, a few papers have recently appeared on the application of ANNs. In particular, Longo et al. [32] developed an ANN model to predict the refrigerant boiling heat transfer coefficients inside brazed plate heat exchangers based on a database of 1760 data points, which included 16 refrigerants and 15 plate geometries. The ANN predicted the data with a Mean Absolute Percentage Error (MAPE) equal to 4.8%. ...
This paper shows the promising capabilities that the aluminum metal foams offer in enhancing the water pool boiling on flat surfaces. Different metal foam samples, with pore density ranging between 5 and 40 Pore Per Inch (PPI), with similar porosity values, around 0.92, and two different thickness values, 5 and 10 mm, were tested to investigate their pertinent pool boiling performance. The saturated boiling curves at atmospheric pressure were obtained. Furthermore, using a high speed camera, heat and fluid flow inside and above the porous layer were investigated. The results showed that boiling heat transfer can be greatly enhanced by the use of metal foams partly because of an earlier onset of the nucleate boiling. Moreover, over a wide range of operating conditions foams lead to remarkably higher heat transfer coefficients , compared with a smooth aluminum reference surface. For a more comprehensive understanding of the problem, we tried to correlate the existing data in the literature for different foam geometries and base materials. However, the existing correlations and independent test data in the literature do not agree well; not even within the same order of magnitude. Hence, in order to offer a generic solution, predictions based on the results of an artificial neural network (ANN) are sought. A large database, comprising 758 experimental data points available in the open literature, was collected and then used to develop, train and validate a model based on ANN to estimate the heat transfer performance of metal foams during water pool boiling. Our data show that the predicted Nusselt number is within 10% of the measured experimental data. Moreover, we demonstrate that the developed ANN tool can be successfully implemented to predict the intrinsic complexity of water pool boiling inside metal foams that in most cases cannot be articulated by merely relying on conventional semi-empirical methods.
... After an experimental study, two novel correlations were calculated by the curve-fitting method [66][67][68][69][70][71][72][73][74] for HN and DHN, individually. In the end, to predict the other domain (more or less than the mentioned temperature and Vf.), an artificial neural network [75][76][77][78][79][80][81], or MD & Nanofluid simulations [82-85] have been modeled. A comparative study between achieved data and those trained by novel correlation reveals a fine certainty. ...
Nanofluid refers to the mixture of fluid and solid nanoparticles. If this mixture contains more than one NP or fluid, it is called "hybrid nanofluid"; further, if HN contains more than one NP and also more than one fluid, it is called "dihybrid nanofluid." In this research, first, titanium dioxide NP was dispersed in the water-ethylene glycol basefluid and formed an HN. Then, thermal conductivity of HN was measured. After that, MWCNT NP was added to the HN and formed a DHN. Further, TC of DHN measured. Both HN and DHN TCs were compared, and the results revealed that by adding MWCNT, thermal conductivity enhanced about 30.83% (from 25.65% of HN to 56.48% of DHN). On the other hand, to analyze the phase structure, and to observe the microstructure, X-ray diffraction analysis, energy-dispersive X-ray analysis, and field emission scanning electron microscopy were examined. The measured TC for both samples was at volume fractions up to 1.0% and temperatures up to 50 °C. After an experimental study, two novel correlations were calculated by the curve-fitting method for HN and DHN, individually. In the end, to predict the other Vf and temperature, an artificial neural network has been modeled for both HN and DHN.
... After an experimental study, two novel correlations were calculated by the curve-fitting method [66][67][68][69][70][71][72][73][74] for HN and DHN, individually. In the end, to predict the other domain (more or less than the mentioned temperature and Vf.), an artificial neural network [75][76][77][78][79][80][81], or MD & Nanofluid simulations [82-85] have been modeled. A comparative study between achieved data and those trained by novel correlation reveals a fine certainty. ...
Nanofluid refers to the mixture of fluid and solid nanoparticles. If this mixture contains more than one NP or fluid, it is called "hybrid nanofluid"; further, if HN contains more than one NP and also more than one fluid, it is called "dihybrid nanofluid." In this research, first, titanium dioxide NP was dispersed in the water-ethylene glycol basefluid and formed an HN. Then, thermal conductivity of HN was measured. After that, MWCNT NP was added to the HN and formed a DHN. Further, TC of DHN measured. Both HN and DHN TCs were compared, and the results revealed that by adding MWCNT, thermal conductivity enhanced about 30.83% (from 25.65% of HN to 56.48% of DHN). On the other hand, to analyze the phase structure, and to observe the microstructure, X-ray diffraction analysis, energy-dispersive X-ray analysis, and field emission scanning electron microscopy were examined. The measured TC for both samples was at volume fractions up to 1.0% and temperatures up to 50 °C. After an experimental study, two novel correlations were calculated by the curve-fitting method for HN and DHN, individually. In the end, to predict the other Vf and temperature, an artificial neural network has been modeled for both HN and DHN.
In recent times, employing machine learning techniques (MLTs) to forecast the evaporation/condensation performance of refrigerants, namely the heat transfer coefficient (HTC) and frictional pressure drop (FPD), has become increasingly significant. In the current investigation, an experimental comparison of the evaporation HTC and FPD of R1234yf, R290, and R13I1/R290 was explored in an offset strip fin-plate heat exchanger. Following this, six MLTs—support vector regressor (SVR), multilayer perceptron regressor (MLPR), gradient boosting regressor (GBR), AdaBoost regressor (ABR), ridge regressor (RR), and K-nearest neighbors regressor (KNNR)—were proposed to forecast the HTC and FPD of these refrigerants. The comparative experimental analysis revealed that the evaporation-HTC of R290 was 27.8–76.2 % and 46.2–73.8 % higher than that of R1234yf and R13I1/R290, respectively, while FPD declined by up to 74.8 % and 57.2 %, respectively. Under the same working conditions, the first transition from nucleation to convective boiling occurred in the order of R1234yf, R13I1/290, and R290. In addition, nucleate boiling dominated at lower vapor quality, while convective boiling prevailed at higher vapor quality. A total of 336 experimental data points corresponding to various test conditions from existing studies and current experiments were considered for the MLT prediction analysis. According to the results, GBR (without any enhancement approaches) was the best approach for forecasting the FPD and HTC, with mean absolute errors (MAE) of 0.125 % and 0.131 %, respectively. Additionally, to enhance the forecasting efficiency of the MLTs, feature selection, analysis of principal components, and hyperparameter tuning were employed. According to feature importance, mass flux, reduced pressure, saturation temperature, heat flux, and entry vapor quality were identified as the most influential parameters on the FPD and HTC. Ultimately, the most accurate predictions of HTC and FPD, with the lowest MAE of 0.111 %, were achieved using MLPR and SVR, respectively, with feature selection.
This research investigates the complex phenomenon of nanofluid flow boiling and its associated heat transfer characteristics. Employing advanced numerical simulations and machine learning techniques, we explore the behavior of dimensionless heat transfer correlations in subcooled flow boiling scenarios using water and SiO2/water nanofluids. The study begins with an adaptive design of experiment, yielding a comprehensive dataset of nearly 250 simulations. In the numerical simulation phase, the extracted data are examined, and the grid independence of the modeling process is confirmed, ensuring robustness and reliability. The data, validated against experimental results, provides critical insights into the intricate interplay of dimensionless numbers influencing Nusselt number behavior. Subsequently, the extracted dataset from the numerical simulations underwent a two-stage feature selection process, incorporating Pearson correlation and iterative techniques, to identify the most influential dimensionless parameters for the calculation of the Nusselt number. Later, the data are randomly split into training and testing sets (70–30%), and predictive models are developed using Python, leveraging libraries such as Pandas, NumPy, scikit-learn, and Keras. A tenfold cross-validation approach is employed to ensure model stability and accuracy. Through response surface methodology (RSM), we establish regression equations for average and local Nusselt numbers, achieving minimal mean absolute error (MAE) and high R-squared () values, demonstrating the effectiveness of our approach. Further enhancing predictive capabilities, we explore random forest, support vector machine, and artificial neural network (ANN) models. The ANN model emerges as the top performer, offering exceptional accuracy with MAE below 2.21% and above 0.95 for both average and local Nusselt numbers. Notably, the machine learning algorithms, from data preprocessing to the final model evaluation, required only about 10% of the time invested in the numerical simulations.
This paper proposes a universal approach that employs artificial neural networks (ANN) to predict the off-design performance of printed circuit heat exchangers (PCHEs) using supercritical carbon dioxide (sCO2) fluid. PCHEs are widely studied precoolers and recuperators of sCO2 power cycles, for which predicting their off-design heat transfer performance is challenging due to the fluid’s dramatic non-linear property variations. Due to their narrow valid range and constrained expressing ability, the existing empirical correlations might introduce significant out-of-scope errors that result in diverged simulation results, slow convergence and abnormal quitting. This work addresses the issue by proposing an ANN-based Nusselt number correlation, ANNu, to predict the off-design performance of PCHEs quickly and universally over the vast operating range of sCO2 cycles. Multiple techniques were adopted to promote ANNu’s prediction accuracy and generalization, such as consolidating 494 training data, data-driven input selection, and global optimization for the networks’ architecture and hyperparameters. This paper comprehensively evaluates ANNu’s performance by comparing its predictions with ten empirical correlations over four typical real-world scenarios, where ANNu features an overall MSE of 83.4222 and overall MAPE of 16.285 % that rank the second best and third best among eleven candidates (even overpassing two references), respectively. Moreover, ANNu exhibits good generalization and reliability, featuring MAPEs ranging from 9.89 % to 27.73 % when processing unmet experiment data from sCO2 PCHE applications.
This study is part of the Flow Boiling and Condensation Experiment (FBCE) and utilizes flow boiling data collected in both microgravity onboard the International Space Station (ISS) and Earth gravity at different channel orientations. The goal is to develop a prediction technique for heat transfer and critical heat flux (CHF) for flow boiling in both microgravity and Earth gravity using artificial neural networks (ANNs). The working fluid, n-perfluorohexane or FC-72, flows through a rectangular channel of 114.6 mm heated length, 2.5 mm heated width, and 5.0 mm unheated height with either one or two walls heated. The consolidated FBCE database for heat transfer coefficient comprises 29,226 datapoints spanning a mass velocity of 173-3200 kg/m 2 s, pressure of 102-238 kPa, subcooling of 0-44 • C, and thermodynamic equilibrium quality of-0.60-0.95 (spanning from highly subcooled to high-quality saturated boiling). Following a statistical analysis of various input parameters relevant to flow boiling and optimization of key model parameters, a fully connected feed-forward ANN is developed to predict Nu tp. It predicts the entire test database with an overall mean absolute error (MAE) of just 7.99% with consistent and accurate predictions in each subset. Similarly, 641 CHF datapoints from FBCE were consolidated into a database spanning a mass velocity of 99-3212 kg/m 2 s, inlet pressure of 97-239 kPa, inlet subcooling of 0-46 • C, inlet thermodynamic equilibrium quality of-0.61-0.86, and CHF values of 4-54 W/cm 2. A separate ANN is developed by following the same methodology as heat transfer, and it predicts dimensionless CHF, Bo CHF , with an overall MAE of just 12.05%. Existing seminal correlations are assessed for subsets of the two consolidated FBCE databases, and the ANNs are shown to have better accuracies in each subset of the database. The ANNs' high prediction accuracy, in conjunction with their ability to predict physical parametric trends in previously unseen data, shows their potential as prediction tools for both heat transfer and CHF for flow boiling in microgravity and Earth gravity.
Heat exchanger modeling has been widely employed in recent years for performance calculation, design optimizations, real-time simulations for control analysis, as well as transient performance predictions. Among these applications, the model’s computational speed and robustness are of great interest, particularly for the purpose of optimization studies. Machine learning models built upon experimental or numerical data can contribute to improving the state-of-the-art simulation approaches, provided careful consideration is given to algorithm selection and implementation, to the quality of the database, and to the input parameters and variables. This comprehensive review covers machine learning methods applied to heat exchanger applications in the last 8 years. The reviews are generally categorized based on the types of heat exchangers and also consider common factors of concern, such as fouling, thermodynamic properties, and flow regimes. In addition, the limitations of machine learning methods for heat exchanger modeling and potential solutions are discussed, along with an analysis of emerging trends. As a regression classification tool, machine learning is an attractive data-driven method to estimate heat exchanger parameters, showing a promising prediction capability. Based on this review article, researchers can choose appropriate models for analyzing and improving heat exchanger modeling.
Although HFC refrigerants have high global warming potential (GWP) values, they are preferred due to their satisfactory cooling performance and A1 fire protection classification. If possible, alternatives of HFC-type refrigerants should be used; if not, they should be used with the least charge value. In this study, the effect of metal foam heat exchanger was investigated to reduce the amount of refrigerant in the refrigeration system. The performance of the metal foam incorporated internal heat exchanger (IHX) was estimated by trained artificial neural networks (ANNs) using the correlations given in the literature, and the results were compared with the experimental data presented in the literature. For the same cooling capacity, a higher performance is achieved by using IHX with metal foam additives. Although the developed correlation has been extracted for IHX, it could be applied for all HE with gas flow.
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.
During contemporary industrial design and production, traditional numerical simulation is incapable of predicting the temperature field of undetectable surfaces with high accuracy, rapidity and board applicability. In this paper, we present a temperature reconstruction generative adversarial network (TRe-GAN), which can rapidly predict the temperature field of all equipment surfaces through one readily observable two-dimensional temperature image rather than specific thermodynamic parameters. Compared with previous surface temperature prediction networks, TRe-GAN reconstructs surface temperature field via vertex temperatures instead of surface texture, which innovatively incorporates both computer vision semantics and heat transfer semantics. Conditions inputs and an optimization output module were designed to strengthen vertex association and mitigate over-fitting. An orthogonal experimental design scheme considering heat transfer theory is employed to construct reliable data sets representative enough for heat transfer phenomenons. Model details, such as loss function and normalized scheme, are discussed to improve the overall temperature accuracy for thousands of nodes (MAPE = 0.93K). TRe-GAN trained in numerical-simulation data performs well in predicting the global temperature field of the measured experimental data (MAPE=0.89%), which proves that our network effectively avoids over-fitting and can predict measured data from easily available simulated data. Therefore, TRe-GAN has great application potential in producing or designing industrial equipment and target characteristics prediction on the battlefield.
In this paper, experimental data of Nusselt number and pressure drop utilizing TiO2/water, ZnO/water, and Ag/water nanofluids in helical coils under isothermal conditions were used to design an artificial neural network to predict Nusselt number and pressure drop for experimental and predict data. Feed Forward Neural Network (FFNN) and Generalized Regression Neural Network (GRNN) were designed. All predicted results were compared with the original experimental data and the results showed that both networks could be used to predict Nusselt number and pressure drop. Root Mean Square Error (RMSE) was used to measure the accuracy of each neural network. Predicted Nusselt number average RMSE for FFNN and GRNN were 4.84 and 0.004, while expected Nusselt numbers were 4.009 and 0.63. Predicted pressure drop average RMSE for FFNN and GRNN were 6.78 and 0.005 respectively while the average RMSE for expected pressure drop were 6.173 and 0.105 respectively. GRNN showed more accuracy than FFNN. The maximum deviation between trained data and experimental data for FFNN and GRNN were ±15.5% and ±0.02%, respectively. The maximum deviation between predicted data and experimental data for FFNN and GRNN were ±13.1% and ±1.6% respectively. The generated networks to predict non-experimental data showed good and logical behavior and proved that they are sufficient in Nusselt number and pressure drop prediction.
The goal of this study is to improve the accuracy and the validity of the prediction of the heat transfer coefficient (HTC) throughout flow boiling of different water-based nanofluids in a horizontal tube by developing an artificial neural network model using Ag/water, Cu/water, CuO/water, Al 2 O 3 /water, and TiO 2 /water nanofluids. The multiple layer perceptron (MLP) neural network was designed and trained by 354 experimental data points that were collected from the literature. Thermal conductivity of nanoparticle, mass flux, volumetric concentration, and heat flux were used to serve as input variables of the model. The heat transfer coefficient (HTC) was used as the output variable. Via the method of the trial-and error, MLP with 8 neurons in the hidden layer was attained as the optimal artificial neural network structure. This developed smart model is more accordant with the experimental data than the correlations of the literature. The accuracy of the developed smart model was validated by the value of mean squared error (MSE=0.042) and the value of determination coefficient (R ² = 0.9992 ) for all data.
Most chemical processes, such as distillation, absorption, extraction, and catalytic reactions, are extremely complex processes affected by multiple factors. As a result, the relationships between their input and output variables are non-linear, and it is not easy to optimize or control them using traditional methods. Artificial neural network is a systematic structure composed of multiple neuron models. By simulating many basic functions of the nervous system of living organisms, nonlinear control can be realized without relying on mathematical models, and it is especially suitable for more complex control objects. This article will introduce artificial neural networks' basic principles and development history, and review its application research progress in chemical process control, fault diagnosis, and process optimization.
Most chemical processes, such as distillation, absorption, extraction, and catalytic reactions, are extremely complex processes that are affected by multiple factors. The relationships between their input variables and output variables are non-linear, and it is difficult to optimize or control them using traditional methods. Artificial neural network (ANN) is a systematic structure composed of multiple neuron models. Its main function is to simulate multiple basic functions of the nervous system of living organisms. ANN can achieve nonlinear control without relying on mathematical models, and is especially suitable for more complex control objects. This article will introduce the basic principles and development history of artificial neural networks, and review its application research progress in chemical process control, fault diagnosis, and process optimization.
With the rapid development of machine learning techniques, data-mining for processes in chemistry, materials, and engineering has been widely reported in recent years. In this discussion, we summarize some typical applications for process optimization, design, and evaluation of chemistry, materials, and engineering. Although the research and application targets are various, many important common points still exist in their data-mining. We then propose a generalized strategy based on the philosophy of data-mining, which should be applicable for the design and optimization targets for processes in various fields with both scientific and industrial purposes.
This paper presents the heat transfer coefficients and the frictional pressure drops of R1234ze(Z) and R1233zd(E) boiling inside a commercial Brazed Plate Heat Exchanger (BPHE): the effects of heat flux / mass flux, saturation temperature / pressure, outlet conditions and fluid properties are investigated. The boiling heat transfer coefficients are controlled mainly by heat flux / mass flux and evaporator outlet conditions. The degrees of superheating at the outlet of the evaporator produces a degradation of the average boiling heat transfer coefficient in the whole evaporator, particularly for R1234ze(Z). The frictional pressure drops exhibit a quadratic dependence on refrigerant mass flux. Saturation temperature / pressure has a remarkable influence only on the frictional pressure drops. R1234ze(Z) exhibits boiling heat transfer coefficients higher and frictional pressure drops lower than those of R1233zd(E). There is a reasonable agreement between the saturated boiling heat transfer coefficients and the calculated values by two recent models for boiling inside BPHE.
Measurements of viscosity of trans-1‑chloro‑3,3,3-trifluoropropene (R-1233zd(E)) in liquid and vapor phases are the key attention of the present study and we developed correlations to predict viscosities of saturated liquid and vapor by extrapolating the data to saturation condition, which are useful in industrial design and simulation. R-1233zd(E) is being introduced as a potential candidate to be an alternative working fluid for high temperature heat pumps and organic Rankine cycles (ORCs). In this work, the viscosity of R-1233zd(E) was measured by a tandem capillary tubes method up to 4.07 MPa pressure over a temperature ranges from 314 K (40.85 °C) to 434 K (160.85 °C) and 394 K (120.85 °C) to 474 K (200.85 °C) for liquid and vapor phases, respectively. Total standard combined uncertainties in liquid and gas viscosity measurements are lower than ± 3.0% and ± 3.1%, respectively. On the other hand, there were large deviations between experimental data and REFPROP version 9.1 for the liquid viscosity in a range of -25 to -37% and those of vapor phase are −7 to −16%. Experimental data of condensation heat transfer which cannot be correlated well with Nusselt's theory by REFPROP version 9.1 viscosity give a good agreement by using the present viscosity data.
In this study, the flow boiling heat transfer characteristics of R-1234ze(E) and R-134a in plate heat exchangers with different Chevron angles are measured and analyzed as a function of the mass flux, saturation temperature, vapor quality, and heat flux. The effect of the mass flux on the heat transfer and pressure drop of R-1234ze(E) is substantial. The heat transfer coefficient of R-1234ze(E) for a Chevron angle of 60° is approximately 3.7 times higher than that for a Chevron angle of 30° at high vapor qualities owing to the intensified turbulent flow. Moreover, for a Chevron angle of 60°, the average heat transfer coefficient of R-1234ze(E) is on average 4.7% higher than that of R-134a due to its higher equivalent Reynolds number. However, the average pressure drop of R-1234ze(E) is higher than that of R-134a owing to the lower vapor density of R-1234ze(E). Finally, the correlations for the heat transfer and pressure drop of R-1234ze(E) are developed in the plate heat exchangers with different Chevron angles.
Although R-1233zd(E) has been considered as an alternative to R-245fa used in the organic Rankine cycle (ORC), experimental studies on the heat transfer characteristics of R-1233zd(E) in plate heat exchangers are limited. In this study, the evaporation heat transfer coefficient and pressure drop of R-1233zd(E) in a brazed plate heat exchanger are measured with respect to the mass flux, heat flux, saturation temperature, and vapor quality. As a result of the experiment in this study, the heat transfer coefficient of R-1233zd(E) is strongly dependent on the mass flux and vapor quality, and not on the heat flux and saturation temperature because the flow is in the convective boiling regime. The frictional pressure drop of R-1233zd(E) shows a strong dependence on the mass flux, vapor quality, and saturation temperature. Moreover, the heat transfer coefficient and pressure drop of R-1233zd(E) are compared with those of R-245fa. Finally, empirical correlations for the heat transfer coefficient and friction factor of R-1233zd(E) are developed based on the measured data.
Organic Rankine cycle power systems for low quality waste heat recovery applications can play a major role in achieving targets of increasing industrial processes efficiency and thus reducing the emissions of greenhouse gases. Low capacity organic Rankine cycle systems are equipped with brazed plate heat exchangers which allows for efficient heat transfer with a compact design. Accurate heat transfer correlations characterizing these devices are required from the design phase to the development of model-based control strategies. In this paper, the experimental heat transfer coefficient and pressure drop during vaporization at typical temperatures for low quality waste heat recovery organic Rankine cycle systems are presented for the working fluids HFC-245fa and HFO-1233zd. The experiments were carried out at saturation temperatures of 100 °C, 115 °C and 130 °C and inlet and outlet qualities ranging between 0.1–0.4 and 0.5–1 respectively. The experimental heat transfer coefficients and frictional pressure drop were compared with well-known correlations and new ones are developed. The results indicated weak sensitivity of the heat transfer coefficients to the saturation temperature and were characterized by similar values for the two fluids. The frictional pressure drop showed a linear dependence with mean quality and increased as the saturation temperature decreased.
Abstract Present study deals with the flow boiling of R245fa, a commercial working fluid used in organic Rankine cycle, in brazed plate heat exchanger with chevron angle of 45 degree and 60 degree. The effects of the heat flux, mass flux rate of refrigerant, saturation temperature on convective heat transfer coefficients are investigated. The operating conditions of the experiment are as mass flux: 30–40 kg m−2 s−1 quality at evaporator inlet: 0.1–0.8, heat flux: 2–15 kW m−2. The heat transfer result suggests a nucleate boiling dominant process in the evaporator. The convective heat transfer coefficient showed a strong dependence on the heat flux and vapor quality at evaporator inlet. Moreover convective heat transfer coefficient show a linear relationship with mass flux of the refrigerant. It is worth mentioning that heat transfer coefficient is higher at higher saturation temperature and chevron angle. Based on the experimental data, empirical correlations were developed for the prediction of heat transfer coefficients and frictional pressure drop of refrigerant R245fa in brazed plate heat exchanger.
The optimal design of the evaporator is one of the key issues to improve the efficiency and economics of organic Rankine cycle units. The first step in studying the evaporator design is to understand the thermal- hydraulic performance of the working fluid in the evaporator of organic Rankine cycles. This paper is aimed at obtaining flow boiling heat transfer and pressure drop characteristics in a plate heat exchanger under the working conditions prevailing in the evaporator of organic Rankine cycle units. Two hydrofluoroolefins R1234yf and R1234ze, and one hydrofluorocarbon R134a, were selected as the working fluids. The heat transfer coefficients and pressure drops of the three working fluids were measured with varying saturation temperatures, mass fluxes, heat fluxes and outlet vapour qualities, which range from 60 C to 80 C, 86 kg/m2 s to 137 kg/m2 s, 9.8 kW/m2 to 36.8 kW/m2 and 0.5 to 1, respectively. The working conditions covered relatively high saturation temperatures (corresponding reduced pressures of 0.35–0.74), which are prevailing in organic Rankine cycles yet absent in the open literature. The experimental data were compared with existing correlations, and new correlations were developed that are more suitable for evaporation in organic Rankine cycles. The experimental results indicate that heat transfer coefficients are strongly dependent upon the heat flux and saturation temperature. Moreover, the results suggest better thermal-hydraulic performance for R1234yf than the other two working fluids at the same saturation temperatures. With the new heat transfer and pressure drop correlations, agreements within ±25% were obtained for experimental data in similar experiments with high saturation temperatures.
This paper presents the heat transfer coefficients and the pressure drops measured during HFC404A vaporisation inside a commercial BPHE and the comparison of this data with previous measurements carried out during HC290 (Propane) and HC1270 (Propylene) vaporisation inside the same BPHE and similar operating condition in order to assess the capability of HydroCarbon refrigerants as long-term low GWP substitutes for HFC404A in commercial and industrial refrigeration. Propane and Propylene exhibit boiling heat transfer coefficient very similar and frictional pressure drops higher than to those of HFC404A, therefore, taking into account also their good thermodynamic properties, they seems to be very promising as long-term low GWP substitutes for HFC404A. The HFC404A boiling heat transfer coefficients were also compared with a new model for refrigerant boiling inside BPHE (Longo et al., 2015): the mean absolute percentage deviation between calculated and experimental data is 6.0%. The heat transfer measurements were also complemented with an IR thermography analysis for a better understanding of refrigerant vaporisation heat transfer regime inside a BPHE.
This paper presents measurements of heat transfer coefficient obtained during flow boiling of R32 inside a brazed plate heat exchanger (BPHE). Although R32 is known as a very interesting refrigerant for its thermodynamic and thermophysical properties, very limited flow boiling data are published in the open literature for R32 working in brazed plate heat exchangers.
The present experimental data are measured to investigate the effect of refrigerant heat flux, mass velocity, inlet vapor quality and superheating at the outlet. The saturation temperature is kept constant at around 5 °C, which is a usual temperature level for evaporation in liquid coolers. As a significant result, differently from other studies on flow boiling with HFC refrigerants, mass flux is found to be very important, meaning a high contribution of the convective term on the heat transfer coefficient.
The present data are also analyzed to assess available correlations for flow boiling inside BPHEs, in order to provide useful information on the accuracy of predicting methods that can be used for evaporators with R32.
This paper investigates the effects of heat flux, saturation temperature, and outlet conditions on HFO1234ze(E) boiling inside a Brazed Plate Heat Exchanger (BPHE). The effect of the heat flux on the heat transfer coefficients was remarkable. Similar consideration applies for outlet condition effects whereas the impact of saturation temperature was found to be lower. The frictional pressure drop shows a linear dependence on the refrigerant kinetic energy per unit volume. The two-phase flow boiling heat transfer coefficients were compared with a new model for refrigerant boiling inside BPHE (Longo et al., 2015): the mean absolute percentage deviation between calculated and experimental data is 7.2%. The present data points were compared with those of HFC134a and HFO1234yf previously measured inside the same BPHE under the same operating conditions: HFO1234ze(E) exhibits heat transfer coefficients very similar to HFC134a and HFO1234yf and frictional pressure drops slightly higher than HFC134a and HFO1234yf. © 2016 Elsevier Ltd and International Institute of Refrigeration. All rights reserved.
The high ownership cost of mining equipment mean that downtimes are expensive and should be avoided with smart and efficient maintenance planning. Modern mines have large data sets on equipment performance and reliability, from dispatch and manufacturer health monitoring systems, that can be mined for more efficient maintenance planning. This study explores the application of classification and clustering approaches for pattern recognition and failure forecasting on mining shovels. The failure behaviour of a fleet of ten mining shovels during 1 year of operation was investigated using these techniques. The shovels were classified into four clusters using k-means clustering algorithms. Future failures were predicted using the support vector machine (SVM) classification technique. Historical failure and time to repair data were used to predict the next failure type for all shovels. The SVM technique was shown to be successful with prediction accuracy of over 75%. This is the first attempt (to the best of our knowledge) that the failure type is predicted based on historical failure/repair data for mining equipment. Clustering shovels based on their reliability can be used for equipment allocation and maintenance planning. These objectives cannot be achieved with traditional reliability modelling. Successful application of these techniques will be valuable input for decision-making during preventive maintenance scheduling.
With the increased regulations to move towards lower GWP refrigerants, natural fluids and their blends are becoming more favorable. Plate heat exchangers are being used in air-conditioning and refrigeration applications as well as a wide variety of other applications including food processing, chemical industry, and energy generation systems. Plate heat exchangers are favored because of their compactness, close approach temperature pure counter-flow operation, and enhanced heat transfer performance. Plate heat exchangers are increasingly utilized in two-phase flow operations due to their desirable characteristics. In order to achieve a better understanding of the current research status of two-phase flow in plate heat exchangers, this paper presents a literature review of available correlations for heat transfer and pressure drop calculations during evaporation in plate heat exchangers. Generally, research on evaporation heat transfer performance in plate heat exchangers is limited. Correlations on natural refrigerant mixtures are scarce. One recent correlationis proposed in literature for ammonia/water mixtures with specific conditions and mixture concentrations. A comparative evaluation of some of the existing correlations is presented in the light of their applicability to natural refrigerants. Overall, there is a significant gap in the literature regarding evaporation heat transfer and fluid flow characteristics of these types of exchangers.
This two-part paper presents an overview of evaporation heat transfer mechanisms, a review of the experimental and prediction methods and a creation of a consolidated multi-lab database of 3601 data points and provides a detailed comparison of all the prediction methods to this broad database and finally proposes new prediction methods for the local heat transfer coefficient and the frictional pressure gradient of flow boiling within plate heat exchangers. Specifically, in Part 1, a description of the complex geometry of plate heat exchangers and an introduction to their major applications are described, followed by an extensive literature survey of experimental studies and associated prediction methods. While many prediction methods are found to work in the literature, the results of this study show that these methods have only been compared to their original data, but have not been vetted against a large database covering many fluids, plate designs and test conditions.
In the second part of this study a sensitivity analysis on the prediction methods is performed to consider the effect of plate geometry on thermal–hydraulic performance and an extensive comparison of all the two-phase pressure drop and flow boiling heat transfer prediction methods available in the open literature are also provided versus the large diversified database presented in Part 1. The experimental databank, from numerous independent research studies, is then utilized to develop the new prediction methods to evaluate local heat transfer coefficients and pressure drops. These new methods were developed from 1903 heat transfer and 1513 frictional pressure drop data points (3416 total), respectively, and were proved to work better over a very wide range of operating conditions, plate designs and fluids (including ammonia). The prediction for flow boiling heat transfer coefficients was broken down into separate macro- and micro-scale methods.
Abstract This paper presents a new model for refrigerant boiling inside Brazed Plate Heat Exchangers (BPHEs) based on a set of 251 experimental data previously obtained by the authors which includes data points relative to HFC refrigerants (HFC236a, HFC134a, HFC410A), HC refrigerants (HC600a-Isobutane, HC290-Propane, HC1270-Propylene), and also a new low Global Warming Potential (GWP) HFO refrigerant (HFO1234yf). The new model includes specific equations for nucleate and convective boiling. The new model was compared against a set of 505 experimental data obtained by different laboratories, which includes HFC134a, HFC410A, HFC507A and HCFC22 data points with different plate geometries. The mean absolute percentage deviation between experimental and calculated data is around 20%.
This paper presents HFC32 average boiling heat transfer coefficients and pressure drops measured inside a small Brazed Plate Heat Exchanger (BPHE): the effects of heat flux, saturation temperature (pressure), and outlet conditions are investigated. The experimental tests were carried out at four different saturation temperatures (5, 10, 15, and 20°C) and four different evaporator outlet conditions (vapour quality around 0.80 and 1.00, vapour super-heating around 5 and 10 °C). The average heat transfer coefficients show great sensitivity to heat flux and outlet conditions and weak sensitivity to saturation temperature (pressure). The saturated boiling heat transfer coefficients were compared with a new model for refrigerant vaporisation inside BPHE (Longo et al., 2015): the mean absolute percentage deviation between calculated and experimental data is 4.7%. The heat transfer and pressure drop measurements are complemented with a IR thermography analysis for a better understanding of the vaporisation process inside a BPHE.
This paper presents the experimental heat transfer coefficients and pressure drop measured during HC-600a (isobutane), HC-290 (propane), and HC-1270 (propylene) vaporization inside a brazed plate heat exchanger (BPHE): the effects of heat flux, refrigerant mass flux, saturation temperature (pressure), evaporator outlet condition, and fluid properties are investigated. The experimental tests include 172 vaporization runs carried out at three different saturation temperatures (10, 15, and 20 degrees C) and four different evaporator outlet conditions (outlet vapor quality around 0.80 and 1.00, outlet vapor super-heating around 5 and 10 degrees C). The refrigerant mass flux ranges from 6.6 to 23.9 kg m(-2) s(-1) and the heat flux from 4.3 to 19.6 kW m(-2). The heat transfer and pressure drop measurements have been complemented with IR thermography analysis in order to quantify the portion of the heat transfer surface affected by vapor super-heating. The heat transfer coefficients show great sensitivity to heat flux, evaporator outlet condition and fluid properties and weak sensitivity to saturation temperature (pressure). The frictional pressure drop shows a linear dependence on the kinetic energy per unit volume of the refrigerant flow and therefore a quadratic dependence on refrigerant mass flux. HC-1270 exhibits heat transfer coefficients 6-12% higher than HC-290 and 35-50% higher than HC-600a and frictional pressure drops 5-10% lower than HC-290 and 60% lower than HC-600a. The experimental heat transfer coefficients are compared with two well-known correlations for nucleate boiling and a linear equation for frictional pressure drop is proposed. [DOI: 10.1115/1.4006817]
This paper presents the experimental heat transfer coefficients and pressure drop measured during vaporisation of the new low Global Warming Potential (GWP) refrigerant HFO1234yf inside a Brazed Plate Heat Exchanger (BPHE): the effects of heat flux, mass flux, saturation temperature (pressure) and outlet conditions are investigated. The heat transfer coefficients show great sensitivity to heat flux and outlet conditions and weak sensitivity to saturation temperature (pressure). The frictional pressure drop shows a linear dependence on the kinetic energy per unit volume of the refrigerant flow and therefore a quadratic dependence on refrigerant mass flux. The saturated boiling experimental heat transfer coefficients are reproduced by two well-known equations for nucleate boiling, Cooper (1984) and Gorenflo (1993), with reasonable agreement. The heat transfer and pressure drop measurements are complemented with IR thermography analysis in order to quantify the portion of the heat transfer surface affected by vapour super-heating.
Plate heat exchangers (PHE’s) are being used to an increasing extent as refrigerant evaporators but published information on their performance in this mode is rather limited. In this paper, two-phase heat transfer and pressure drop characteristics are presented for PHE’s when used as refrigerant liquid over-feed evaporators. Laboratory experiments were carried out with three industrial PHE’s having different chevron angle combinations, using refrigerant R134a and R507A. Measurements were made over ranges of mass flux, heat flux and corresponding outlet vapour qualities, and the effects of these parameters on the thermal and hydraulic performance of the evaporators were evaluated. Additional field test data of thermal performance were collected from ammonia and R12 water chillers, operating as thermosiphon evaporators. Based on all these data, empirical correlations are proposed for predicting the refrigerant boiling heat transfer coefficient and two-phase frictional pressure drop in PHE’s.
The evaporation heat transfer coefficient and pressure drop for refrigerant R-134a flowing in a plate heat exchanger were investigated experimentally in this study. Two vertical counterflow channels were formed in the exchanger by three plates of commercial geometry with a corrugated sine shape of a chevron angle of 60 deg. Upflow boiling of refrigerant R-134a in one channel receives heat from the hot down flow of water in the other channel. The effects of the mean vapor quality mass flux, heat flux, and pressure of R-134a on the el,evaporation heat transfer and pressure drop were explored. The quality change of R-134a between the inlet and outlet of the refrigerant channel ranges from 0.09 to 0.18. Even at a very low Reynolds number, the present flow visualization of evaporation in a plate heat exchanger with the transparent outer plate showed that the flow in the plate heat exchanger remains turbulent It is found that the evaporation heat transfer coefficient of R-134a in the plates is much higher than that in circular pipes and shows a very different variation with the vapor quality from that bl circular pipes, particularly in the convective evaporation dominated regime at high vapor quality. Relatively intense evaporation on the corrugated surface was seen from the flow visualization. Moreover, the present data showed that both the evaporation hear transfer coefficient and pressure drop increase with the vapor quality. At a higher mass flux the pressure drop is higher for the entire range of the vapor quality but the evaporation heat transfer is clearly better only at the high quality. Raising the imposed wall heat flux was found to slightly improve the heat transfer, while at a higher refrigerant pressure, both the heat transfer and pressure drop are slightly lower. Based an the present data, empirical correlations for the evaporation heat transfer coefficient and friction factor were proposed.
Experiments on the evaporative heat transfer and pressure drop in the brazed plate heat exchangers were performed with refrigerants R410A and R22. The plate heat exchangers with different 45°, 35°, and 20° chevron angles are used. Varying the mass flux of refrigerant (13–34 kg/m2s), the evaporating temperature (5, 10 and 15 °C), the vapor quality (0.9–0.15) and heat flux (2.5, 5.5 and 8.5 kW/m2), the evaporation heat transfer coefficients and pressure drops were measured. The heat transfer coefficient increases with increasing vapor quality and decreasing evaporating temperature at a given mass flux in all plate heat exchangers. The pressure drop increases with increasing mass flux and quality and with decreasing evaporating temperature and chevron angle. It is found that the heat transfer coefficients of R410A are larger than those of R22 and the pressure drops of R410A are less than those of R22. The empirical correlations of Nusselt number and friction factor are suggested for the tested PHEs. The deviations between correlations and experimental data are within ±25% for Nusselt number and ±15% for friction factor.
The characteristics of evaporation heat transfer and pressure drop for refrigerant R134a flowing in a plate heat exchanger were investigated experimentally in this study. Two vertical counter flow channels were formed in the exchanger by three plates of commercialized geometry with a corrugated sine shape of a chevron angle of 60°. Upflow boiling of refrigerant R134a in one channel receives heat from the hot downflow of water in the other channel. The effects of the heat flux, mass flux, quality and pressure of R134a on the evaporation heat transfer and pressure drop were explored. The preliminary measured data for the water to water single phase convection showed that the heat transfer coefficient in the plate heat exchanger is about 9 times of that in a circular pipe at the same Reynolds number. Even at a very low Reynolds number, the present flow visualization in a plate heat exchanger with the transparent outer plate showed that the flow in the plate heat exchanger remains turbulent. Data for the pressure drop were also examined in detail. It is found that the evaporation heat transfer coefficient of R134a in the plates is quite different from that in circular pipe, particularly in the convective evaporation dominated regime at high vapor quality. Relatively intense boiling on the corrugated surface was seen from the flow visualization.
More specifically, the present data showed that both the evaporation heat transfer coefficient and pressure drop increase with the vapor quality. At a higher mass flux the pressure drop is higher for the entire range of the vapor quality but the heat transfer is only better at high quality. Raising the imposed wall heat flux was found to slightly improve the heat transfer. While at a higher system pressure the heat transfer and pressure drop are both slightly lower.
Saturated flow boiling heat transfer and the associated frictional pressure drop of the ozone friendly refrigerant R-410A (a mixture of 50 wt% R-32 and 50 wt% R-125) flowing in a vertical plate heat exchanger (PHE) are investigated experimentally in the study. In the experiment two vertical counter flow channels are formed in the exchanger by three plates of commercial geometry with a corrugated sinusoidal shape of a chevron angle of 60°. Upflow boiling of saturated refrigerant R-410A in one channel receives heat from the downflow of hot water in the other channel. The experimental parameters in this study include the refrigerant R-410A mass flux ranging from 50 to 125 kg/m2 s and imposed heat flux from 5 to 35 kW/m2 for the system pressure fixed at 1.08, 1.25 and 1.44 MPa, which respectively correspond to the saturated temperatures of 10, 15 and 20 °C. The measured data showed that both the boiling heat transfer coefficient and frictional pressure drop increase almost linearly with the imposed heat flux. Furthermore, the refrigerant mass flux exhibits significant effect on the saturated flow boiling heat transfer coefficient only at higher imposed heat flux. For a rise of the refrigerant pressure from 1.08 to 1.44 MPa, the frictional pressure drops are found to be lower to a noticeable degree. However, the refrigerant pressure has very slight influences on the saturated flow boiling heat transfer coefficient. Finally, empirical correlations are proposed to correlate the present data for the saturated boiling heat transfer coefficients and friction factor in terms of the Boiling number and equivalent Reynolds number.
This paper presents the experimental heat transfer coefficients and pressure drop measured during HFC refrigerant 134a, 410A and 236fa vaporisation inside a small brazed plate heat exchanger: the effects of heat flux, refrigerant mass flux, saturation temperature, outlet conditions and fluid properties are investigated. The experimental results are reported in terms of refrigerant side heat transfer coefficients and frictional pressure drop. The heat transfer coefficients show great sensitivity to heat flux and outlet conditions and weak sensitivity to saturation temperature. The frictional pressure drop shows a linear dependence on the kinetic energy per unit volume of the refrigerant flow. HFC-410A shows heat transfer coefficients 40–50% higher than HFC-134a and 50–60% higher than HFC-236fa and frictional pressure drops 40–50% lower than HFC-134a and 50–60% lower than HFC-236fa. The experimental heat transfer coefficients are compared with two well-known equations for nucleate boiling [M.G. Cooper, Heat flows rates in saturated pool boiling – a wide ranging examination using reduced properties, Advanced Heat Transfer, Academic Press, Orlando, Florida, 1984, pp. 157–239; D. Gorenflo, Pool boiling, in: E.U. Schlünder (Ed.), VDI Heat Atlas, Dusseldorf, Germany, 1993, Ha1-25] and a correlation for frictional pressure drop is proposed.
This paper is published in the IX World Renewable Energy Congress, Florence, Italy. This paper presents a new method to optimise solar energy systems in order to maximise their economic benefits. The system is modelled with TRNSYS computer program. An artificial neural network is trained using a small number of annual TRNSYS simulation results, to learn the correlation of collector area and storage tank size on the auxiliary energy required by the system and thus on the net solar energy price. Subsequently a genetic algorithm is employed to estimate the optimum size of these two parameters, which maximise the net solar energy price, thus the design time is reduced substantially and the solution obtained is more accurate that the trial and error method used traditionally in these optimisations.
The objective of this work is to use artificial intelligence methods, like artificial neural-networks and genetic algorithms, to optimize a solar-energy system in order to maximize its economic benefits. The system is modeled using a TRNSYS computer program and the climatic conditions of Cyprus, included in a typical meteorological year (TMY) file. An artificial neural-network is trained using the results of a small number of TRNSYS simulations, to learn the correlation of collector area and storage-tank size on the auxiliary energy required by the system from which the life-cycle savings can be estimated. Subsequently, a genetic algorithm is employed to estimate the optimum size of these two parameters, for maximizing life-cycle savings: thus the design time is reduced substantially. As an example, the optimization of an industrial process heat-system employing flat-plate collectors is presented. The optimum solutions obtained from the present methodology give increased life-cycle savings of 4.9 and 3.1% when subsidized and non-subsidized fuel prices are used respectively, as compared to solutions obtained by the traditional trial-and-error method. The present method greatly reduces the time required by design engineers to find the optimum solution and in many cases reaches a solution that could not be easily obtained from simple modeling programs or by trial-and-error, which in most cases depends on the intuition of the engineer.
Šerbanovi ć , Mean heat transfer coefficient and pressure drop during the evaporation of 1,1,1,2 tetrafluoroethane (R-134a) in a plate heat exchanger
- E M Djordjedvi Ć
- S Kabelac
E.M. Djordjedvi ć, S. Kabelac, S.P. Šerbanovi ć, Mean heat transfer coefficient and
pressure drop during the evaporation of 1,1,1,2 tetrafluoroethane (R-134a) in a
plate heat exchanger, J. Chem. Serb. Soc. 72 (2007) 833-846.
NIST reference fluid thermodynamic and transport properties database -REFPROP, Version 10.0, National Institute of Standards and Technology, standard reference data program
- E W Lemmon
- I H Bell
- M L Huber
E.W. Lemmon, I.H. Bell, M.L. Huber, McLinden M.O, NIST reference fluid thermodynamic and transport properties database -REFPROP, Version 10.0, National Institute of Standards and Technology, standard reference data program,
Gaithersburg (2017).
A survey of correlations for heat transfer and pressure drop for evaporation and condensation in plate heat exchangers
- Eldeed
Mean heat transfer coefficient and pressure drop during the evaporation of 1,1,1,2 tetrafluoroethane (R-134a) in a plate heat exchanger
- Djordjedvić
NIST reference fluid thermodynamic and transport properties database - REFPROP, Version 10.0, National Institute of Standards and Technology, standard reference data program
- Lemmon