Classification Of Flow Regimes Research Articles

Artificial Neural Networks for Gas‐Liquid Flow Regime Classification in Small Channels

AbstractThe reliable design of multiphase micro‐structured apparatus requires a precise knowledge of the internal flow regime. Previous research indicated that classifiers based on artificial neural networks (ANN) are relatively simple to develop and provide a reasonable accuracy when trained with data for specific inlet designs. This paper introduces advanced ANN classifiers capable of predicting all relevant flow regimes regardless of the inlet design with a recall of 94 % and above for Taylor, churn, dispersed, rivulet, and parallel flows, between 89 % and 94 % for annular and bubbly flows, and 83 % for Taylor‐annular flow. These classifiers were trained and validated by using more than 13,000 experimental data points extracted from 97 flow maps.

A deep learning approach to classifying flow-induced vibration response regimes of an elliptical cylinder

This study proposes a new approach that leverages deep learning to the study of flow-induced vibration (FIV), specifically to automate flow regime classification and to visualize the transitions between these regimes. Using previously obtained data on the amplitude response of an elastically mounted cylinder as a function of reduced velocity for a range of structural damping ratios, the time trace of the body displacement and fluid driving forces are first converted into a frequency-time representation using continuous wavelet transforms before being input to several pre-trained convolutional neural networks for feature extraction. When utilizing the outputs of each convolutional neural network for regime classification, we found that almost all the machine learning approaches had high cross-validation accuracy that was statistically insignificant from each other. The five best-performing classifiers were then used as an ensemble method, yielding a weighted accuracy of 99.1% on the test data. The FIV response regimes were further investigated by projecting the outputs of the pre-trained convolutional neural networks onto the first three modes identified with principal component analysis (PCA). The PCA plots indicated that, among all the models considered, Xception showed superior capability in delineating distinct locations for different FIV response regimes, based on the lock-in frequency and the presence of harmonics in the driving fluid forces. Moreover, the PCA plots also showed that increasing the structural damping ratio resulted in a diminished disparity in the dynamics of the identified FIV response regimes, leading to a less discernible separation between the regimes in the plots.

Flow regime classification using various dimensionality reduction methods and AutoML

Accurate identification of flow regimes is paramount in several industries, especially in chemical and hydrocarbon sectors. This paper describes a comprehensive data-driven workflow for flow regime identification. The workflow encompasses: i) the collection of dynamic pressure signals using an experimentally verified numerical two-phase flow model for three different flow regimes: stratified, slug and annular flow, ii) feature extraction from pressure signals using Discrete Wavelet Transformation (DWT), iii) Evaluation and testing of 12 different Dimensionality Reduction (DR) techniques, iv) the application of an AutoML framework for automated Machine Learning classifier selection among K-Nearest Neighbors, Artificial Neural Networks, Support Vector Machines, Gradient Boosting, Random Forest, and Logistic Regression, with hyper-parameter tuning. Kernel Fisher Discriminant Analysis (KFDA) is the best DR technique, exhibiting superior goodness of clustering, while KNN proved to be the top classifier with an accuracy of 92.5 % and excellent repeatability. The combination of DWT, KFDA and KNN was used to produce a virtual flow regime map. The proposed workflow represents a significant step forward in automating flow regime identification and enhancing the interpretability of ML classifiers, allowing its application to opaque pipes fitted with pressure sensors for achieving flow assurance and automatic monitoring of two-phase flow in various process industries.

Buoyant miscible viscoplastic displacements in vertical pipes: Flow regimes and their characterizations

We experimentally study miscible displacement flows of a light Newtonian fluid by a heavy viscoplastic fluid, in a vertical pipe with a large aspect ratio (δ−1≫1). We use camera imaging, laser-induced fluorescence, and ultrasound Doppler velocimetry techniques, to capture and process data. Four dimensionless parameters, namely, the Reynolds (Re), Bingham (B), viscosity ratio (M), and densimetric Froude (Fr) numbers (or their combinations), mainly govern the flow dynamics. We identify and characterize three distinct flow regimes, including plug, separation, and mixing regimes, while we describe each regime's dynamics in detail, particularly in terms of the velocity and concentration fields as well as the displacement front velocity. In addition, we analyze the plug regime concerning the residual wall layers, the separation regime in terms of the separation dynamics, spatiotemporal separation zone, and viscoplastic layer thinning, and the mixing regime regarding the mixing index and macroscopic diffusion. Finally, we develop a simplified model to help delineate the flow regime classification, in the plane of Re/Fr2 and M.

Classification of watersheds in the conterminous United States using shape-based time-series clustering and Random Forests

Watershed classification is considered necessary for various purposes, including improving the transferability of streamflow information and allowing the generalization of hydrologic theories. In this paper, we proposed a classification method using shape-based time-series clustering techniques to define hydrologically homogeneous classes for a set of 638 gauged watersheds in the conterminous United States. We defined 15 distinct flow-regime classes based on standardized weekly-step mean annual hydrographs of the watersheds analyzed. Most classes showed regionality, though to various degrees. Classes in the Atlantic Coast region showed strong geographical contiguity indicating that spatial proximity may be an appropriate indicator of flow-regime similarity; but on the other hand, nearby watersheds may exhibit different intra-annual variabilities and watersheds far apart from one another may exhibit a similar flow regime. Such cases show that spatial proximity may not be used as a universal indicator of flow-regime similarity. To expand the flow-regime classification to ungauged watersheds, where no streamflow data are available, we proposed a physical-characteristic classification method using the Random Forests algorithm based on a set of physical-climatic attributes. Based on the Boruta feature-selection algorithm, attributes related to snowiness, precipitation seasonality, and green vegetation coverage stood out as the most relevant controls on regime-class memberships. By analyzing the within-class variability of the watershed attributes, we found that watersheds of the same flow-regime class may be very different in terms of individual attribute values indicating that flow-regime dis/similarity may not be equated to dis/similarity in individual watershed attributes; instead, it is the interplay among attributes that determines the streamflow behavior. Overall, methods proposed for the classification of gauged and ungauged watersheds are believed to be of value in a range of applications including selecting gauged donor watersheds for estimating historical streamflow records at ungauged watersheds.

Linear logistic regression with weight thresholding for flow regime classification of a stratified wake

• Build a DNS database with O(100) realizations for stratified wake. • Develop a machine learning tool to classify the flow regimes in a stratified wake. • The physics governing stratified wake is universal with respect to Re and Fr. • The classifier captures the universal physics and generalize well. A stratified wake has multiple flow regimes, and exhibits different behaviors in these regimes due to the competing physical effects of momentum and buoyancy. This work aims at automated classification of the weakly and the strongly stratified turbulence regimes based on information available in a full Reynolds stress model. First, we generate a direct numerical simulation database with Reynolds numbers from 10000 to 50000 and Froude numbers from 2 to 50. Order (100) independent realizations of temporally evolving wakes are computed to get converged statistics. Second, we train a linear logistic regression classifier with weight thresholding for automated flow regime classification. The classifier is designed to identify the physics critical to classification. Trained against data at one flow condition, the classifier is found to generalize well to other Reynolds and Froude numbers. The results show that the physics governing wake evolution is universal, and that the classifier captures that physics.

Development of Gas-Liquid Flow Regimes Identification Using a Noninvasive Ultrasonic Sensor, Belt-Shape Features, and Convolutional Neural Network in an S-Shaped Riser.

The problem of classifying gas-liquid two-phase flow regimes from ultrasonic signals is considered. A new method, belt-shaped features (BSFs), is proposed for performing feature extraction on the preprocessed data. A convolutional neural network (CNN/ConvNet)-based classifier is then applied to categorize into one of the four flow regimes: 1) annular; 2) churn; 3) slug; or 4) bubbly. The proposed ConvNet classifier includes multiple stages of convolution and pooling layers, which both decrease the dimension and learn the classification features. Using experimental data collected from an industrial-scale multiphase flow facility, the proposed ConvNet classifier achieved 97.40%, 94.57%, and 94.94% accuracy, respectively, for the training set, testing set, and validation set. These results demonstrate the applicability of the BSF features and the ConvNet classifier for flow regime classification in industrial applications.

Aerodynamic performance of a sinusoidal wavy cylinder at subcritical and critical Reynolds numbers

Previous studies have shown that in the subcritical regime, wavy circular cylinders exhibit less wind-induced drag than standard circular cylinders. However, the aerodynamic performance of wavy cylinders at higher Reynolds numbers is still unknown. Hence, a series of wind tunnel tests including measurement of the surface pressure, vortex-induced vibration (VIV) and dry galloping were conducted for a wavy cylinder and a corresponding standard cylinder at Re = 1.00 × 105-3.86 × 105. The results indicate that the aerodynamic characteristics, including the mean and fluctuating pressure distributions, mean and fluctuating aerodynamic force coefficients, Strouhal number, flow regime classification, and non-Gaussian characteristics, are significantly modified when shifting from the standard cylinder to the wavy cylinder. Furthermore, these modifications highly depend on the Reynolds number. Regarding the wavy cylinder, related parameters obviously vary along the spanwise direction, indicating three-dimensional aerodynamic characteristics. Moreover, the maximum VIV amplitude of the wavy cylinder is 40% lower than that of the standard cylinder. The velocity corresponding to the maximum amplitude is increased by 11% for the wavy cylinder. Additionally, the maximum dry galloping amplitude is reduced by 11%.

Accurate Flow Regime Classification and Void Fraction Measurement in Two-Phase Flowmeters Using Frequency-Domain Feature Extraction and Neural Networks

Two-phase flow is very important in many areas of science, engineering, and industry. Two-phase flow comprising gas and liquid phases is a common occurrence in oil and gas related industries. This study considers three flow regimes, including homogeneous, annular, and stratified regimes ranging from 5–90% of void fractions simulated via the Mont Carlo N-Particle (MCNP) Code. In the proposed model, two NaI detectors were used for recording the emitted photons of a cesium 137 source that pass through the pipe. Following that, fast Fourier transform (FFT), which aims to transfer recorded signals to frequency domain, was adopted. By analyzing signals in the frequency domain, it is possible to extract some hidden features that are not visible in the time domain analysis. Four distinctive features of registered signals, including average value, the amplitude of dominant frequency, standard deviation (STD), and skewness were extracted. These features were compared to each other to determine the best feature that can offer the best separation. Furthermore, artificial neural network (ANN) was utilized to increase the efficiency of two-phase flowmeters. Additionally, two multi-layer perceptron (MLP) neural networks were adopted for classifying the considered regimes and estimating the volumetric percentages. Applying the proposed model, the outlined flow regimes were accurately classified, resulting in volumetric percentages with a low root mean square error (RMSE) of 1.1%.

Experimental investigation of two-phase flow regimes in rectangular micro-channel with two mixer types

This study primarily aimed to explore the two-phase flow characteristics in a rectangular channel with a hydraulic diameter of 1.33 mm. High-speed video imaging techniques were used to capture two-phase flow behavior corresponding to different superficial gas and liquid velocities. To examine the effect of the gas–liquid mixer configuration on two-phase flow patterns, experiments were conducted using two types of mixers. The results show that the mixer configuration can affect the flow structure only in the slug flow regime. The present study provides a more detailed classification of flow regimes than previous studies. The flow regimes observed were classified into seven categories: slug, aerated-slug, transition, multiple, wavy-annular, smooth-annular, and dispersed-annular flows. Moreover, the quantitative characterization of flow regimes and transition boundaries in micro-channel flows using the standard deviation of differential pressure fluctuations was attempted. This study comprehensively reviews and assesses previous two-phase flow regime maps for both macro-channels and mini/micro-channels. It is shown that the two-phase flow pattern in mini/micro-channels is influenced by various factors, such as the channel hydraulic diameter, the channel cross-sectional shape, the gas–liquid mixer configuration, the difference in the hydraulic diameters between the mixer outlet and test channel, and the thermophysical properties of the working fluids. The comparison between the present experimental data and predictions of the previous flow regime maps demonstrates the necessity of developing a new flow regime map to accurately predict two-phase flow regimes in micro-channels.

Self-organizing maps for efficient classification of flow regimes from gamma densitometry time series in three-phase fluidized beds

The potential of artificial neural networks as a tool to classify and identify a change in the flow regime of a three-phase fluidized bed is studied. Particularly, the suitability of self-organizing maps (SOMs), unsupervised neural networks that visualize the data in a lower dimension, is evaluated. Statistical features of experimental time series determined in a three-phase (granulated carbon-air-water) fluidized bed are extracted as inputs to train the SOM. Photon-count time series are obtained along the fluidized bed vertical axis by gamma-densitometry at different operative conditions. Then, they are analyzed to determine the underlying flow regime indexes. When each input data is presented to the SOMs, a neuron is activated, giving a visual representation of the data. The resulting models show three different regions on the map for the homogenous, transition, and heterogeneous flow regimes. Once these regions are delimited, the map can quickly classify the equipment operating conditions. The ability of the SOMs to diagnose a flow transition is verified against visual observation and gas hold-up trends. The conclusions are tested for their sensitivity to alternative axial positions of the radiation source used for the densitometry.

Aerodynamic Interactions of Side-by-Side Rotors in Ground Proximity

An experimental investigation is conducted to study the aerodynamic behavior of a two-rotor system in ground proximity. The counter-rotating rotors are placed side-by-side in the hovering condition. The time-averaged and unsteady flow behavior is studied when the rotor-to-rotor lateral distance and the distance between the rotors and the ground are varied. The experiments are performed using three-dimensional large-scale volumetric velocimetry with helium-filled soap bubbles as tracers, tracked by the particle motion analysis technique “Shake-The-Box.” The mean velocity field reveals the wake deflection due to the ground plane and the formation of toroidal-shape regions of separated flow below each rotor. The interaction of the wall jets formed by slipstream deflection results in a separation line with the flow emerging from the wall in a fountain-like pattern. Regimes of flow re-ingestion occur when the rotors are sufficiently far apart. The flowfield exhibits the tendency toward asymmetric states, during which the fountain flow column and the domain of re-ingestion shift closer to one of the rotors. A generic classification of flow regimes is proposed in relation to the behavior of two rotors in ground effect.

Computer Vision-Based Classification of Flow Regime and Vapor Quality in Vertical Two-Phase Flow.

This paper presents a method to classify flow regime and vapor quality in vertical two-phase (vapor-liquid) flow, using a video of the flow as the input; this represents the first high-performing and entirely camera image-based method for the classification of a vertical flow regime (which is effective across a wide range of regimes) and the first image-based tool for estimating vapor quality. The approach makes use of computer vision techniques and deep learning to train a convolutional neural network (CNN), which is used for individual frame classification and image feature extraction, and a deep long short-term memory (LSTM) network, used to capture temporal information present in a sequence of image feature sets and to make a final vapor quality or flow regime classification. This novel architecture for two-phase flow studies achieves accurate flow regime and vapor quality classifications in a practical application to two-phase CO2 flow in vertical tubes, based on offline data and an online prototype implementation, developed as a proof of concept for the use of these models within a feedback control loop. The use of automatically selected image features, produced by a CNN architecture in three distinct tasks comprising flow-image classification, flow-regime classification, and vapor quality prediction, confirms that these features are robust and useful, and offer a viable alternative to manually extracting image features for image-based flow studies. The successful application of the LSTM network reveals the significance of temporal information for image-based studies of two-phase flow.

Study on Flow Regime Classification of Gas-Liquid Two-Phase Flow with Bidirectional LSTM Networks

Study on Flow Regime Classification of Gas-Liquid Two-Phase Flow with Bidirectional LSTM Networks

Identification of flow regimes in boiling flows in a vertical annulus channel with machine learning techniques

Machine learning techniques are applied to identify the flow regimes of boiling flows in a vertical annulus channel using conductivity probe signals as the input data. The boiling dataset used in this work spans a system pressure of 198.7–907.2 kPa, a heat flux of 68–285 kW/m2, an inlet subcooling of 3.8–11.4 °C, a mass flux of 642.6–2465.1 kg/m2 s, an area-averaged gas superficial velocity of 0.01–2.83 m/s, and an area-averaged liquid superficial velocity of 0.68–2.69 m/s. A two-step approach that uses the unsupervised self-organizing map (SOM) for identifying the local and global flow regimes is proposed. The cumulative distribution functions (CDFs) of bubble durations are used as the input of the SOM. The combinations of local flow regimes that form the global regimes are extensively analyzed. The radial profiles of the global flow regimes are studied where the effects of the heated inner wall and local subcooling are discussed. Furthermore, the supervised support vector machine (SVM) and the K-nearest neighbors (KNN) algorithms are trained with features extracted from the conductivity probe signals for flow regime classification, with the global flow regimes identified by the SOM as the reference. Two sets of features are generated where the first consists of four statistical features of bubble durations and the second consists of six flow features that describe the collective bubble behaviors in the flow. Using the global flow regimes identified by the SOM as the reference and the flow features dataset as the input data, the trained models are able to classify the flow regimes to an accuracy of more than 90%.

Classification of flow regimes using a neural network and a non-invasive ultrasonic sensor in an S-shaped pipeline-riser system

A method for classifying flow regimes is proposed that employs a neural network with inputs of extracted features from Doppler ultrasonic signals of flows using either the Discrete Wavelet Transform (DWT) or the Power Spectral Density (PSD). The flow regimes are classified into four types: annular, churn, slug, and bubbly flow regimes. The neural network used in this work is a feedforward network with 20 hidden neurons. The network comprises four output neurons, each of which corresponds to the target vector's element number. 13 and 40 inputs are used for features extracted from PSD and DWT respectively. Experimental data were collected from an industrial-scale multiphase flow facility. Using the PSD features, the neural network classifier misclassified 3 out of 31 test datasets in the classification and gave 90.3% accuracy, while only one dataset was misclassified with the DWT features, yielding an accuracy of 95.8%, thus showing the superiority of the DWT in feature extraction of flow regime classification. The approach demonstrates the applicability of a neural network and DWT for flow regime classification in industrial applications using a clamp-on Doppler ultrasonic sensor. The scheme has significant advantages over other techniques as only a non-radioactive and non-intrusive sensor is used. To the best of our knowledge, this is the first known successful attempt for the classification of liquid-gas flow regimes in an S-shape riser system using an ultrasonic sensor, PSD-DWTs features, and a neural network.

Flow pattern observations and flow pattern map for adiabatic two-phase flow of carbon dioxide in vertical upward and downward direction

A test facility to investigate flow pattern transitions of vertical two-phase flow of CO2 has been built within the scope of the high-luminosity detector upgrades at the European Organization for Nuclear Research (CERN). Adiabatic flow pattern observations for both vertical up- and downflow are recorded with high-speed imaging in tubes of 8 mm inner diameter, with saturation temperatures in the range of −25°C to +5°C and mass velocities ranging from 100kgm−2s−1 to 450kgm−2s−1. A database of 431 flow pattern observations in upward and 123 in downward direction has been compiled. The recorded data have been analysed with machine learning techniques and a previously trained Frame- and Flow-Regime-Classifier is used for the flow regime classification. The observed two-phase flow pattern transitions did not match the transition lines of existing flow pattern maps. As a consequence, new transition lines for the bubbly-to-slug, slug-to-churn and churn-to-annular transitions have been developed for both vertical upflow and downflow respectively and condensed into new flow pattern maps. It is concluded, that the flow regime transitions are strongly depended on vapour quality, mass velocity, the flow direction and the fluid properties. Compared to horizontal flow, a dryout region is not observed and the liquid film of the annular flow regime dries out symmetrically at vapour qualities close to x=1.

Application of Wavelet Feature Extraction and Artificial Neural Networks for Improving the Performance of Gas-Liquid Two-Phase Flow Meters Used in Oil and Petrochemical Industries.

Measuring fluid characteristics is of high importance in various industries such as the polymer, petroleum, and petrochemical industries, etc. Flow regime classification and void fraction measurement are essential for predicting the performance of many systems. The efficiency of multiphase flow meters strongly depends on the flow parameters. In this study, MCNP (Monte Carlo N-Particle) code was employed to simulate annular, stratified, and homogeneous regimes. In this approach, two detectors (NaI) were utilized to detect the emitted photons from a cesium-137 source. The registered signals of both detectors were decomposed using a discrete wavelet transform (DWT). Following this, the low-frequency (approximation) and high-frequency (detail) components of the signals were calculated. Finally, various features of the approximation signals were extracted, using the average value, kurtosis, standard deviation (STD), and root mean square (RMS). The extracted features were thoroughly analyzed to find those features which could classify the flow regimes and be utilized as the inputs to a network for improving the efficiency of flow meters. Two different networks were implemented for flow regime classification and void fraction prediction. In the current study, using the wavelet transform and feature extraction approach, the considered flow regimes were classified correctly, and the void fraction percentages were calculated with a mean relative error (MRE) of 0.4%. Although the system presented in this study is proposed for measuring the characteristics of petroleum fluids, it can be easily used for other types of fluids such as polymeric fluids.

Classification and Prediction of Natural Streamflow Regimes in Arid Regions of the USA

Understanding how natural variation in flow regimes influences stream ecosystem structure and function is critical to the development of effective stream management policies. Spatial variation in flow regimes among streams is reasonably well understood for streams in mesic regions, but a more robust characterization of flow regimes in arid regions is needed, especially to support biological monitoring and assessment programs. In this paper, we used long-term (41 years) records of mean daily streamflow from 287 stream reaches in the arid and semi-arid western USA to develop and compare several alternative flow-regime classifications. We also evaluated how accurately we could predict the flow-regime classes of ungauged reaches. Over the 41-year record examined (water years 1972–2013), the gauged reaches varied continuously from always having flow > zero to seldom having flow. We predicted ephemeral and perennial reaches with less error than reaches with an intermediate number of zero-flow days or years. We illustrate application of our approach by predicting the flow-regime classes at ungauged reaches in Arizona, USA. Maps based on these predictions were generally consistent with qualitative expectations of how flow regimes vary spatially across Arizona. These results represent a promising step toward more effective assessment and management of streams in arid regions.

Development of flow regime classification method with k-means clustering and application to the high-pressure two-phase flow

Development of flow regime classification method with k-means clustering and application to the high-pressure two-phase flow