Convolutional neural network for flow boiling patterns classification

This work combines fuzzy logic and a support vector machine (SVM) with a principal component analysis (PCA) to create an artificial-intelligence system that identifies nanofluid gas-liquid two-phase flow states in a vertical mini-channel. Flow-pattern recognition requires finding the operational details of the process and doing computer simulations and image processing can be used to automate the description of flow patterns in nanofluid gas-liquid two-phase flow. This work uses fuzzy logic and a SVM with PCA to improve the accuracy with which the flow pattern of a nanofluid gas-liquid two-phase flow is identified. To acquire images of nanofluid gas-liquid two-phase flow patterns of flow boiling, a high-speed digital camera was used to record four different types of flow-pattern images, namely annular flow, bubbly flow, churn flow, and slug flow. The textural features extracted by processing the images of nanofluid gas–liquid two-phase flow patterns are used as inputs to various identification schemes such as fuzzy logic, SVM, and SVM with PCA to identify the type of flow pattern. The results indicate that the SVM with reduced characteristics of PCA provides the best identification accuracy and requires less calculation time than the other two schemes. The data reported herein should be very useful for the design and operation of industrial applications.

The behaviour of two-phase gas-liquid flows in a stopper-rod controlled submerged entry nozzle (SEN) is investigated in water model experiments. The observed two-phase flow patterns can be classified into either bubble coring or bubbly slug. The scaling of the two-phase flows by means of similarity parameters is discussed. In the experiments, it is found that the liquid flow rates depend strongly on the two-phase flow patterns. Additionally, the influence of swirl on the flow patterns is investigated in detail. It is shown that swirl has a marked impact on the transition from bubble coring to bubbly slug. Finally, an estimation of the two-phase argon-steel flow patterns in industrialscale SEN flows is given.

Liqnet: A real-time monitoring network for two-phase flow patterns

The air-water two-phase flow plays an important role in many applications of industry fields. Usually, a 90-degree bend is used to connect pipes for changing the direction of flow which influences the two-phase flow pattern. In this paper, the effect of 90-degree bend under different ranges of gas and liquid superficial velocities on the two-phase flow patterns in the horizontal pipe located after the bend was experimentally investigated, and then results were presented and compared in a two-phase flow pattern map. Also, tomographic images and probability density functions were used to capture the cross-section void fraction and its distribution for the two-phase flow patterns. The results revealed that at low liquid and gas flow rates, a stratified-wavy flow pattern was observed as a dominant flow pattern. While the wavy-annular and semiannular flow patterns were observed at a high range of gas flow rates in the horizontal pipe. The results also showed that at the high range of liquid flow rate, bubbly, plug, slug, stratified-wavy, and wavy-annular flow patterns were observed in the horizontal pipe when the gas flow increased. The tomographic images and probability density functions gave good agreement with the experimental observations and results.

Two-Phase Flow Patterns in Microchannel Vaporization of CO2 at Near-Critical Pressure

Two-phase flow patterns and transition characteristics for in-tube condensation with different surface inclinations

Two-phase flow patterns and heat transfer characteristics of R134a refrigerant during flow boiling in a single rectangular micro-channel

3D simulations of the impact of two-phase flow on PEM fuel cell performance

Two-phase flow patterns in short horizontal rectangular microchannels

Subcooled flow boiling of seawater in a vertical annulus

Numerical simulation for heat transfer and velocity field characteristics of two-phase flow systems in axially rotating horizontal cans

A numerical and experimental study is reported for two-phase flow and heat transfer in a horizontal porous formation with water through flow and partial heating from below. Based on a newly developed two-phase mixture model, numerical results of the temperature distribution, liquid saturation, liquid and vapor phase velocity fields are presented for three representative cases with varying inlet velocities. It is found that the resulting two-phase structure and flow patterns are strongly dependent upon the water inlet velocity and the bottom heat flux. The former parameter measures the flow along the horizontal direction, while the latter creates a relative motion between the phases in the vertical direction. Experiments are also performed to measure temperature distributions and to visualize the two-phase flow patterns. Qualitative agreement between experiments and numerical predictions is achieved. Overall, this combined experimental and numerical study has provided new insight into conjugate single- and two-phase flow and heat transfer in porous media, although future research is required if accurate modeling of these complex problems is to be accomplished.

Summary Of significant importance to the multiphase flow, the gas/liquid two-phase flow patterns and the transition boundaries between them are extremely difficult to be characterized inside the centrifugal pumps because the steel-made nontransparent casing shrouds the impeller’s flow passages. How to visualize the internal flow patterns within the rotating impeller has drawn considerable attention. In this study, an indirect approach proposed by Gamboa and Prado (2011) to recognize the gas/liquid flow patterns in the centrifugal impeller via inflection characteristics of H-Q performance curves is used and verified by both experimental tests and numerical simulations. To further validate the indirect flow pattern recognition approach to a low-specific-speed centrifugal pump, the acrylic-made transparent impeller was adopted to conduct experimental testing, in which the internal two-phase flow pattern can be visualized by a high-speed camera. The 3D numerical simulation was performed and compared with corresponding visualization experiments to illustrate the two-phase flow structures. The preliminary results reveal that the inflection characteristics on the H-Q performance curves of the centrifugal pump can be regarded as the flow pattern transition boundaries under various flow conditions. The experiments and simulations conducted for this paper further validated an existing mechanistic model to predict flow pattern transitions.

To study two-phase flows occurring in the process of coaxially electrospinning core-shell composite nanofibers, a mathematic physical model was proposed with shear-thinning fluid properties, surface tension, and electric fields specially considered. Numerical results showed three flow patterns: slug, slug-annular transitional, and annular flows. According to a flow pattern model based on the Weber number, slug-annular transitional flows were predicted to occur when the Weber number was between 0.5 and 26, whereas the slug and annular flows were to be expected when the Weber number was less than 0.5 and greater than 26, respectively. Body forces (induced by electric fields) and surface tension forces had more explicit effects on the flow patterns than forces due to viscosity and permeability gradients on interfaces. To form an annular flow pattern with smooth core-shell interfaces, which is favored in engineering practice, the applied electric field should be stronger than 100 kv/m.

In the paper, numerical methods of data analysis recurrence quantification analysis (RQA) and self-organizing map (SOM) have been used to analyse pressure drop oscillations during the flow boiling in minichannel. The performed analysis allows us to identify flow patterns based on the character of the pressure drop oscillations. The following two-phase flow patterns have been identified: liquid flow, liquid flow with small vapour bubble, slug flow, long slug flow and confined bubble flow. In the experiment, the open-loop boiling system in a circular horizontal minichannel with an inner diameter of 1 mm was investigated. The two-phase flow patterns at the outlet of the heated section were observed through the glass tube (with an inner diameter of 1 mm) and recorded by a high-speed camera Phantom v1610.