Machine Learning Algorithms for Flow Pattern Classification in Pulsating Heat Pipes (original) (raw)
Related papers
Predicting the thermal performance of pulsating heat pipes using artificial neural networks
Zenodo (CERN European Organization for Nuclear Research), 2022
The present work proposes a promising approach to predict the thermal performance of Pulsating Heat Pipes (PHPs) using Artificial Neural Networks (ANN). The available database corresponds to 1097 experimental records with nine distinct geometries conceived in the frame of the Clean Sky2 project PHP2. According to the transient data, for 12.5% of cases, PHPs failed to operate in a pulsating mode and instead exhibited conduction behavior. Thus, the proposed approach consists of separating the database into 'Conduction' and 'Pulsation' and training three ANN models. First, a classification model is constructed to predict the operating mode. The F1-score, used to examine the accuracy of this model, resulted in 96.35 % for the pulsation category. Then, a regression model is created for each mode to predict the thermal resistance. The mean relative error of regression models resulted in an acceptable error of 11.48% and 40.93% for pulsating and conduction modes, respectively.
Machine Learning for Prediction of Heat Pipe Effectiveness
Energies
This paper details the selection of machine learning models for predicting the effectiveness of a heat pipe system in a concentric tube exchanger. Heat exchanger experiments with methanol as the working fluid were conducted. The value of the angle varied from 0° to 90°, values of temperature varied from 50 °C to 70 °C, and the flow rate varied from 40 to 120 litres per min. Multiple experiments were conducted at different combinations of the input parameters and the effectiveness was measured for each trial. Multiple machine learning algorithms were taken into consideration for prediction. Experimental data were divided into subsets and the performance of the machine learning model was analysed for each of the subsets. For the overall analysis, which included all the three parameters, the random forest algorithm returned the best results with a mean average error of 1.176 and root-mean-square-error of 1.542.
Machine learning applications to predict two-phase flow patterns
2021
Recent advances in artificial intelligence with traditional machine learning algorithms and deep learning architectures solve complex classification problems. This work presents the performance of different artificial intelligence models to classify two-phase flow patterns, showing the best alternatives for this specific classification problem using two-phase flow regimes (liquid and gas) in pipes. Flow patterns are affected by physical variables such as superficial velocity, viscosity, density, and superficial tension. They also depend on the construction characteristics of the pipe, such as the angle of inclination and the diameter. We selected 12 databases (9,029 samples) to train and test machine learning models, considering these variables that influence the flow patterns. The primary dataset is Shoham (1982), containing 5,675 samples with six different flow patterns. An extensive set of metrics validated the results obtained. The most relevant characteristics for training the models using Shoham (1982) dataset are gas and liquid superficial velocities, angle of inclination, and diameter. Regarding the algorithms, the Extra Trees model classifies the flow patterns with the highest degree of fidelity, achieving an accuracy of 98.8%.
Scientific Reports
In the current study, Artificial Intelligence (AI) approach was used for the learning of a physical system. We applied four inputs and one output in the learning process of AI. In the learning process, the inputs are space locations of a BCR (bubble column reactor), which are x, y, and z coordinate as well as the amount of gas fraction in BCR. The liquid velocity is also considered as output. A variety of functions were used in learning, such as gbellmf and gaussmf functions, to examine which functions can give the best learning. At the end of the study, all of the results were compared to CFD (computational fluid dynamics). A three-dimensional (3D) BCR was used in this research, and we studied simulation by CFD as well as AI. The data from CFD in a 3D BCR was studied in the AI domain. In AI, we tuned for various parameters to achieve the best intelligence in the system. For instance, different inputs, different membership functions, different numbers of membership functions were us...
Classification of machine learning frameworks for data-driven thermal fluid models
International Journal of Thermal Sciences
Thermal fluid processes are inherently multi-physics and multi-scale, involving mass-momentumenergy transport phenomena at multiple scales. Thermal fluid simulation (TFS) is based on solving conservative equations, for which-except for "first-principles" direct numerical simulationclosure relations (CRs) are required to provide microscopic interactions or so-called sub-grid-scale physics. In practice, TFS is realized through reduced-order modeling, and its CRs as low-fidelity models can be informed by observations and data from relevant and adequately evaluated experiments and high-fidelity simulations. This paper is focused on data-driven TFS models, specifically on their development using machine learning (ML). Five ML frameworks are introduced including physics-separated ML (PSML or Type I ML), physics-evaluated ML (PEML or Type II ML), physics-integrated ML (PIML or Type III ML), physics-recovered (PRML or Type IV ML), and physics-discovered ML (PDML or Type V ML). The frameworks vary in their performance for different applications depending on the level of knowledge of governing physics, source, type, amount and quality of available data for training. Notably, outlined for the first time in this paper, Type III models present stringent requirements on modeling, substantial computing resources for training, and high potential in extracting value from "big data" in thermal fluid research. The current paper demonstrates and investigates ML frameworks in three examples. First, we utilize the heat diffusion equation with a nonlinear conductivity model formulated by convolutional neural networks (CNNs) and feedforward neural networks (FNNs) to illustrate the applications of Type I, Type II, Type III, and Type V ML. The results indicate a preference for Type II ML under deficient data support. Type III ML can effectively utilize field data, potentially generating more robust predictions than Type I and Type II ML. CNN-based closures exhibit more predictability than FNN-based closures, but CNN-based closures require more training data to obtain accurate predictions. Second, we illustrate how to employ Type I ML and Type II ML frameworks for data-driven turbulence modeling using reference works. Third, we demonstrate Type I ML by building a deep FNN-based slip closure for two-phase flow modeling. The results show that deep FNN-based closures exhibit a bounded error in the prediction domain.
Application of NARX neural networks in thermal dynamics identification of a pulsating heat pipe
Energy Conversion and Management, 2009
The pulsating heat pipe (PHP) receiving much attention in industries is a novel type of cooling device. The distinguishing feature of PHPs is the unsteady flow oscillations formed by the passing non-uniform distributions of vapour plugs and liquid slugs. This study introduces a methodology of a non-linear autoregressive with exogenous (NARX) neural network to analyze the thermal dynamics of a PHP in both the time and frequency domains. Three heating powers: 30, 70, and 110 W are tested, and all the predicted results are presented in quite good agreement with the measured results. Herein, the harmonic analysis of the non-linear structure can be equivalently conducted with generalized frequency response functions (GFRFs). Based on the non-linear coupling between the various input spectral components, the interpretations of the higher order GFRFs have been extensively presented for demonstrating the non-linear effects on the heat transfer of a PHP at different operating conditions. Crown
Application of Machine Learning to Predict Thermal Performances of Heat Sinks
Proceedings of the 8th World Congress on Mechanical, Chemical, and Material Engineering
In the present study, the capabilities of two machine learning (ML) regression methods, support vector regression (SVR) and kernel ridge regression (KRR), to predict heat transfer coefficients (HTCs) in air-cooled heat sinks (HSs) are evaluated. Within the laminar regime, HSs with different geometrical parameters and at five different Reynolds numbers are considered for the simulations. Since the focus of the present study is the proof-of-concept, the ML-based models are developed using limited numbers of input data. The input data are prepared by solving three-dimensional equations of continuity, momentum, and energy inside the channels of HSs. Results indicate that both SVR and KRR predict HTCs with excellent accuracy and within ±1.9% of simulated values. The present study suggests that both SVR and KRR are promising design tools to predict hydrothermal performances of thermal systems using sufficiently large and accurate input data. Such precise ML-based models will be excellent alternatives to expensive experimental and computational efforts that are required to develop physics-based correlations for predicting hydrothermal performances of engineering systems.
Scientific Reports
For understanding the complex behavior of fluids in a multiphase chemical bubble column reactor, a combination of the computational fluid dynamic (CFD) method and the adaptive network-based fuzzy inference system (ANFIS) method is used to predict bubble flow inside a reactor based on the function of column height. In this study, the Euler–Euler model is employed as a CFD method. In the Eulerian method, continuity and momentum governing equations are mathematically computed for each phase, while the equations are connected together by source terms. After calculating the flow pattern and turbulence flow in the reactor, all data sets are used to prepare a fully artificial method for further prediction. This algorithm contains different learning dimensions such as learning in different directions of reactor or large amount of input parameters and data set representing “big data”. The ANFIS method was evaluated in three steps by using one, two, and three inputs in each one to predict the...
Mathematical Modeling of Closed Loop Pulsating Heat Pipe by Using Artificial Neural Networks
International Journal of Heat and Technology, 2021
Pulsating heat pipe is one of the prominent technology for thermal management of electronic devices. It consists of three sections namely evaporator, adiabatic and condenser section. PHP is a two phase passive device having efficient and quick ability of transferring heat from evaporator section to condenser section. At first an 8 turn pulsating heat pipe of closed loop ends (CLPHP) with copper tube capillary dimensions is investigated experimentally for different fill ratios and for different inclinations by varying range of heat inputs. Different working fluids viz Water, Acetone, Ethanol and Methanol are considered for the experimentation. One of the recent analytical technology for modelling of CLPHPs is Artificial Neural Network (ANN) approach. The analytical models are having limited scope of applicability and they are simple in nature. The present paper describes Validation of experimental data by training prediction model ANN with available data. Three input nodes such as in...
Sādhanā
Pulsating heat pipe (PHP) is one of the prominent research areas in the family of heat pipes. Heat transfer and fluid flow mechanism associated with PHP are quite involved. The analytical models are simple in nature and limited in scope and applicability. The regression models and Artificial Neural Network (ANN) are also limited to a number of input parameters, their ranges and accuracy. The present paper discusses the thermal performance prediction models of a PHP based on ANN and RCA approach. Totally 1652 experimental data are collected from the literature (2003-2017). Nine major influencing input variables are considered for the first time to develop the prediction models. Feed-forward back-propagation neural network is developed and verified. Backward regression analysis is used in RCA-based regression model. Linear and power-law regression correlations are developed for input heat flux in terms of dimensionless Kutateladze (Ku) number, which is a function of Jakob number (Ja), Morton number (Mo), Bond number (Bo), Prandtl number (Pr) and geometry of a PHP. The prediction accuracy of present regression models (R 2 = 0.95) is observed to be better as compared with literature-based correlations.