Javascript must be enabled to continue!
Modeling interfacial area concentration in bubbly flow
View through CrossRef
Accurate mass, momentum, and energy transfer predictions through gas and liquid interfaces are critical in simulating gas–liquid two-phase flow dynamics. Since the interfacial transfer rates depend on the interfacial area concentration (IAC), it is essential to model the IAC accurately for robust two-phase flow analyses. The inverse of the IAC characterizes bubble characteristic length, and thus, the IAC is related to the bubble Sauter mean diameter (SMD). This study proposed a method to use neural networks for systematically developing a one-dimensional SMD correlation, i.e., an IAC correlation for forced convective and bubble columns. First, an extensive literature survey provided 799 data points, consisting of 590 data points in forced convective bubbly flows and 209 in bubble columns. The neural networks were applied to 799 data points to determine the maximum attainable prediction accuracy of the hypothetical SMD correlations. Target mean absolute percentage errors (MAPEs) of 21.2% for IAC and 15.6% for SMD were established. The neural network could identify three critical non-dimensional parameters: liquid fraction, bubble Reynolds number, and viscosity number, enabling the development of the SMD correlation. With void fraction (VF) predicted by the drift-flux correlation, the SMD correlation could be converted to the IAC correlation, i.e., IAC = 6×VF/SMD. The newly developed SMD and IAC correlations demonstrated reliable predictive capabilities. For forced convective bubbly flows, MAPEs were 19.5% for IAC and 16.1% for SMD. For bubble columns, MAPEs were 45.6% for IAC and 26.9% for SMD. Across all of the datasets, the correlation achieved MAPEs of 26.3% for IAC and 19.0% for SMD. The developed correlations were successfully applied to a wide range of flow conditions: channel geometries (medium-to-large circular pipes, rectangular ducts, rod bundles, and annuli); flow directions (upward, downward, and horizontal); gas–liquid systems (air–water, nitrogen–water, and oxygen–sodium sulfite solution); hydraulic diameters (9.0 to 304 mm); and pressures (0.10 to 2.0 MPa).
Title: Modeling interfacial area concentration in bubbly flow
Description:
Accurate mass, momentum, and energy transfer predictions through gas and liquid interfaces are critical in simulating gas–liquid two-phase flow dynamics.
Since the interfacial transfer rates depend on the interfacial area concentration (IAC), it is essential to model the IAC accurately for robust two-phase flow analyses.
The inverse of the IAC characterizes bubble characteristic length, and thus, the IAC is related to the bubble Sauter mean diameter (SMD).
This study proposed a method to use neural networks for systematically developing a one-dimensional SMD correlation, i.
e.
, an IAC correlation for forced convective and bubble columns.
First, an extensive literature survey provided 799 data points, consisting of 590 data points in forced convective bubbly flows and 209 in bubble columns.
The neural networks were applied to 799 data points to determine the maximum attainable prediction accuracy of the hypothetical SMD correlations.
Target mean absolute percentage errors (MAPEs) of 21.
2% for IAC and 15.
6% for SMD were established.
The neural network could identify three critical non-dimensional parameters: liquid fraction, bubble Reynolds number, and viscosity number, enabling the development of the SMD correlation.
With void fraction (VF) predicted by the drift-flux correlation, the SMD correlation could be converted to the IAC correlation, i.
e.
, IAC = 6×VF/SMD.
The newly developed SMD and IAC correlations demonstrated reliable predictive capabilities.
For forced convective bubbly flows, MAPEs were 19.
5% for IAC and 16.
1% for SMD.
For bubble columns, MAPEs were 45.
6% for IAC and 26.
9% for SMD.
Across all of the datasets, the correlation achieved MAPEs of 26.
3% for IAC and 19.
0% for SMD.
The developed correlations were successfully applied to a wide range of flow conditions: channel geometries (medium-to-large circular pipes, rectangular ducts, rod bundles, and annuli); flow directions (upward, downward, and horizontal); gas–liquid systems (air–water, nitrogen–water, and oxygen–sodium sulfite solution); hydraulic diameters (9.
0 to 304 mm); and pressures (0.
10 to 2.
0 MPa).
Related Results
Interfacial thermal conductance of gallium nitride/graphene/diamond heterostructure based on molecular dynamics simulation
Interfacial thermal conductance of gallium nitride/graphene/diamond heterostructure based on molecular dynamics simulation
<sec>Gallium nitride chips are widely used in high-frequency and high-power devices. However, thermal management is a serious challenge for gallium nitride devices. To improv...
Interfacial Area Transport of Vertical Upward Annular Two-Phase Flow
Interfacial Area Transport of Vertical Upward Annular Two-Phase Flow
Accurate prediction of the interfacial area concentration is essential to successful development of the interfacial transfer terms in the two-fluid model. The interfacial area conc...
Interfacial Adhesion in Fibre-Polymer Composites
Interfacial Adhesion in Fibre-Polymer Composites
<p>The mechanical performance of a fibre-polymer composite is largely determined by the strength of interfacial adhesion across the fibre-polymer phase boundary. Therefore, a...
Distribution Parameter and Drift Velocity of Drift-Flux Model in Bubbly Flow
Distribution Parameter and Drift Velocity of Drift-Flux Model in Bubbly Flow
In view of the practical importance of the drift-flux model for two-phase flow analysis in general and in the analysis of nuclear-reactor transients and accidents in particular, th...
Multiphase Flow Metering:An Evaluation of Discharge Coefficients
Multiphase Flow Metering:An Evaluation of Discharge Coefficients
Abstract
The orifice discharge coefficient (CD) is the constant required to correct theoretical flow rate to actual flow rate. It is known that single phase orifi...
Interfacial Area Transport of Bubbly Flow Under Microgravity Environment
Interfacial Area Transport of Bubbly Flow Under Microgravity Environment
In relation to the development of the interfacial area transport equation, axial developments of one-dimensional void fraction, bubble number density, interfacial area concentratio...
Pressure Analysis of DST Flow Period Or Slug Flow For Horizontal Wells In Homogeneous Reservoir
Pressure Analysis of DST Flow Period Or Slug Flow For Horizontal Wells In Homogeneous Reservoir
Abstract
By the transient pressure for horizontal well with constant flow rate and Duhamel's principle, this paper presents the method to calculate the transient ...
Determinants of Cerebrovascular Reserve in Patients with Significant Carotid Stenosis
Determinants of Cerebrovascular Reserve in Patients with Significant Carotid Stenosis
AbstractIntroductionIn patients with 70% to 99% diameter carotid artery stenosis cerebral blood flow reserve may be protective of future ischemic cerebral events. Reserve cerebral ...

