Artificial Neural Network Enhanced Modeling of Convection Driven Heat Transfer During Volumetric Freezing of Binary Eutectic Systems

Variation in entropies of fusion driven by mixing in binary glass forming eutectics

Engineering Applications of Bio-Inspired Artificial Neural Networks

In eutectic, a lamellar microstructure offers better tableting than that of the nonreacted physical mixture. However, bulk deformation remains elusive in two binary eutectics. We hypothesized that the binary eutectic of a drug with different components, having different H-bonding dimensionalities and crystal structure, shall allow the understanding of the structural integrity in the bulk deformation behavior. The shearing molecular solid (FXT Q) shared a common composition with the viscoelastic crystal (ASP I) and brittle (PCM I), forming EM-1 (ϕ1 = 41.27:58.73% w/w) and EM-2 (ϕ2 = 41.10:58.90% w/w), respectively. The excess thermodynamic functions were contributed by high energy microstructures (nonbonding interactions) along incoherent phase boundaries (visualized under CLSM). The energy dispersive analysis enabled the recognition of the relative distribution of higher atoms over the heterogeneous surface. EM-1 (FXT Q-ASP I) demonstrated higher compressibility, tensile strength, and compactibility (CTC profile) compared to those of EM-2 (FXT Q-PCM I) over a range of applied compaction pressures. The lower true yield strength (σ0(EM-1) = 138.66 MPa) of EM-1 as compared to that of EM-2 (σ0(EM-2) = 166.66 MPa) suggested a better deformation performance and incipient plasticity quantified from the "out-of-die" Heckel analysis. From Ryshkewitch analysis, the tensile strength at zero porosity (τ01 = 3.83 MPa) was predicted to be higher for EM-1 than EM-2 (τ02 = 2.54 MPa). The higher bonding strength of EM-1 was contributed to the additional influence of true density and isotropic van der Waals interactions of ASP I (0D). In contrast, EM-2 demonstrated lower compressibility and compactibility, having herringbone molecular packing of PCM I (1D) with a common shearing component (FXT Q (1D)). This study confirmed that the intrinsic deformational and chemical nature of the second component defined the compressibility and compactibility tendency to a greater extent in the tableting performance of conglomerates of crystalline solid solution.

PurposeThe purpose of this paper is to study experimentally enhancement of heat transfer in a tube with axial swirling‐flow promoters. The geometric features of flow geometry to improve heat transfer can be selected in order to yield the maximum opposite reduction in heat exchange flow irreversibility by using exergy‐destruction method. The paper seeks to illustrate the use of neural network approach to analyze heat transfer enhancement data for further study in the scope of the experimental program.Design/methodology/approachFor this purpose, 402 experimental measurements are collected. About 225 of those are used as training data for neural networks, the rest is used for testing. Then, these testing results of artificial neural network (ANN) and experimental data are compared. A formula for presenting exergy loses in a tubular heat exchanger is derived first and then the thermodynamic optimum instead of economic optimum is found by minimizing the exergy losses in the system.FindingsResults from all configurations studied show that the heat transfer rate of the heated increases when the swirling‐flow promoter is inserted. From the heat transfer improvement number defined, it is observed that about 100 percent increase in heat transfer rate and five times increase in the pressure drop can be achieved under the condition of constant flow for the single promoter which has three blades, its blade angle is 30° and its location is in the middle of the tube length.Research limitations/implicationsThe back‐propagation (BP) algorithm was selected as the neural network algorithm, which uses the generalized delta learning rule. The training time of BP algorithm is considerably long. However, the testing of our neural network is real‐time.Practical implicationsThe experimental setup is established to collect the experimental data. It consists of an entrance region, test region (heat exchanger and steam generator), and, flow measurement and control. Also, a software program of neural networks trained BP is written by using Pascal high‐level languages.Originality/valueAn alternative and new approach is proposed in the paper to find optimum flow geometry for a pipe flow with an axial swirling‐flow promoter inserts. It is too difficult to predict the response of a complex physical system that cannot be easily modeled mathematically. The result thus obtained compare well with experimental results, but the computational effort of the ANN and time required in the analysis is much faster as compared. These results show that the ANN can be used efficiently for prediction.

Numerical modeling of solidification and convection in a viscous pure binary eutectic system

Preparation and thermal performance of phase change material with high latent heat and thermal conductivity based on novel binary inorganic eutectic system

Purpose This study aims to analyze the mixed convection of Williamson fluid through a vertical channel, taking into account the effects of both the Soret phenomenon and a magnetic field along with a first-order chemical reaction. An “artificial neural network” (ANN) is used to carry out the investigation. The aim is to examine the impact of various physical parameters on velocity, temperature and concentration profiles as well as on skin friction, heat and mass transfer rates. Design/methodology/approach ANNs are used to solve the flow problem. A multilayer perceptron neural network with tunable parameters is used for the trial functions. These parameters are adjusted to find the best solution. The Adam optimization algorithm (adaptive moment estimation) is applied to adjust the parameters of the trial solution. Findings The authors assess the convergence and accuracy of the findings by comparing them with the exact spectral quasi linearization method, resulting in satisfactory outcomes. The graphs demonstrate how changes in different parameter factors affect the velocity, temperature and concentration curves. The outcome demonstrates that the axial velocity and concentration profiles decrease as the magnetic parameter value increases. The axial velocity and concentration trends amplify as the Hall parameter increases. Meanwhile, a spike in the Williamson and Soret parameters enhances these profiles, except for the temperature profile. The results indicate that raising the value of the chemical reaction parameter drops the velocity, temperature and concentration profiles. The skin friction coefficient and heat transfer amount increase with the rise of the Soret number. Originality/value The current finding from this study can be applied in various engineering and industrial processes where non-Newtonian fluids are involved, such as in chemical processing, petroleum engineering and materials manufacturing. The use of a feed-forward multilayer perceptron neural network, along with the ADAM optimization technique, presents a novel methodology for addressing complex nonlinear equations contributing to the advancements in computational fluid dynamics.

Purpose An innovative model is proposed to analyze the boundary layer flow of a Carreau fluid, including bioconvection effects from gyrotactic microorganisms, and thermal radiation, specifically focusing on a wedge geometry. This paper aims to understand the complex interplay of these factors on fluid dynamics and heat transfer. Design/methodology/approach Artificial neural networks (ANNs)-based technique, specifically backpropagation neural networks, are being used to analyze the bioconvective wedge flow. These techniques are used to understand the intricate thermal and momentum interactions within the flow. To train the ANN, the MATLAB built-in function bvp4c solver is used to generate a reference solution. Findings Artificial Intelligence-based neural networks are indeed powerful tools for enhancing simulation accuracy, particularly in complex flow scenarios, by learning intricate patterns from data and offering more efficient solutions than traditional methods. The training of these networks involves iterating through epochs, where each epoch represents a full pass of the training data, allowing the model to adjust its parameters (weights and biases). The optimal number of epochs is determined by monitoring performance metrics, such as accuracy and error, as the network trains. In the specific context of nanoparticle movement due to thermophoresis, neural networks can be trained to predict this behavior, and their performance is evaluated through metrics like linear regression and fitness. Validation is achieved by comparing the results with existing literature, treating the special cases as a validation method. Originality/value The simultaneous effects of bioconvection, gyrotactic microorganisms, magnetohydrodynamics, Carreau fluid, wedge flow, thermal radiation and the application of ANN with regression analysis in a single study appear to be a novel research area, as it is not explicitly documented in the existing literature.

In much of the public literature on pin-fin heat transfer, Nusselt number is presented as a function of Reynolds number using a power-law correlation. Power-law correlations typically have an accuracy of 20% while the experimental uncertainty of such measurements is typically between 5% and 10%. Additionally, the use of power-law correlations may require many sets of empirical constants to fully characterize heat transfer for different geometrical arrangements. In the present work, artificial neural networks were used to predict heat transfer as a function of streamwise spacing, spanwise spacing, pin-fin height, Reynolds number, and row position. When predicting experimental heat transfer data, the neural network was able to predict 73% of array-averaged heat transfer data to within 10% accuracy while published power-law correlations predicted 48% of the data to within 10% accuracy. Similarly, the neural network predicted 81% of row-averaged data to within 10% accuracy while 52% of the data was predicted to within 10% accuracy using power-law correlations. The present work shows that first-order heat transfer predictions may be simplified by using a single neural network model rather than combining or interpolating between power-law correlations. Furthermore, the neural network may be expanded to include additional pin-fin features of interest such as fillets, duct rotation, pin shape, pin inclination angle, and more making neural networks expandable and adaptable models for predicting pin-fin heat transfer.

In much of the public literature on pin-fin heat transfer, the Nusselt number is presented as a function of Reynolds number using a power-law correlation. Power-law correlations typically have an accuracy of 20% while the experimental uncertainty of such measurements is typically between 5% and 10%. Additionally, the use of power-law correlations may require many sets of empirical constants to fully characterize heat transfer for different geometrical arrangements. In the present work, artificial neural networks were used to predict heat transfer as a function of streamwise spacing, spanwise spacing, pin-fin height, Reynolds number, and row position. When predicting experimental heat transfer data, the neural network was able to predict 73% of array-averaged heat transfer data to within 10% accuracy while published power-law correlations predicted 48% of the data to within 10% accuracy. Similarly, the neural network predicted 81% of row-averaged data to within 10% accuracy while 52% of the data was predicted to within 10% accuracy using power-law correlations. The present work shows that first-order heat transfer predictions may be simplified by using a single neural network model rather than combining or interpolating between power-law correlations. Furthermore, the neural network may be expanded to include additional pin-fin features of interest such as fillets, duct rotation, pin shape, pin inclination angle, and more making neural networks expandable and adaptable models for predicting pin-fin heat transfer.

In the present study, deep learning neural network model has been employed in many engineering problems including heat transfer prediction. The main consideration of this document is to predict the performance of the boiling heat transfer in helical coils under terrestrial gravity conditions and compare with actual experimental data. Total of 877 data sample has been used in the present neural model. Artificial new Neural Network (ANN) model developed in Python environment with Multi-layer Perceptron (MLP) using four parameters (helical coils dimensions, mass flow rate, heating power, inlet temperature) and one parameter (outlet temperature) has been used in the input layer and output layer in order. Levenberg-Marquardt (LM) algorithm using L2 Regularization to find out the optimal model. A typical feed-forward neural network model composed of three layers, with 30 numbers of neurons in each hidden layer, has been found as optimal based on statistical error analysis. The 4-30-30-1 neural model predicts the characteristics of the helical coil with the accuracy of 98.16 percent in the training stage and 96.68 percent in the testing stage. The result indicated that the proposed ANN model successfully predicts the heat transfer performance in helical coils and can be applied for others operation concerned with heat transfer prediction for future works

The preparation of deep eutectic systems (DES) is a priori a simple procedure. By definition, two or more components are mixed together at a given molar ratio to form a DES. However, from our experience in the laboratory, there is a need to standardize the procedure to prepare, characterize and report the methodologies followed by different researchers, so that the results published can be reproduced. In this work, we test different approaches reported in the literature to prepare eutectic systems and evaluated the importance of water in the successful preparation of liquid systems at room temperature. These published eutectic systems were composed of citric acid, glucose, sucrose, malic acid, β-alanine, L-tartaric acid and betaine and not all of preparation methods described could be reproduced. However, in some cases, it was possible to reproduce the systems described, with the inclusion of water as a third component of the eutectic mixture.

Convergent Temperature Representations in Artificial and Biological Neural Networks

Performance predictions of laminar and turbulent heat transfer and fluid flow of heat exchangers having large tube-diameter and large tube-row by artificial neural networks

The binary eutectic mixtures of fatty acid esters are promising phase change materials for energy storage application. However, the low thermal conductivity which is a common problem for organic phase change materials restricts their further and better applications. In order to solve the problem, a novel composite phase change material (CPCM) was prepared in this research by using methyl palmitate-methyl stearate (MP-MS), a typical binary eutectic mixture of fatty acid esters, as phase change material and expanded graphite (EG) as heat transfer enhancer. The heat transfer performance of MP-MS/EG CPCM was numerical simulated by finite element analysis software ABAQUS. Numerical simulation results revealed that EG could notably enhance the heat transfer performance of MP-MS eutectic mixture. The heat transfer rate and phase change reaction rate of MP-MS/EG CPCM were 14 times and 3 times that of MP-MS eutectic mixture, respectively.