Two-phase flow regime prediction is of great importance for designing evaporators and condensers because the influence of the heat transfer coefficients is strongly related to the flow regimes. These flow regimes are often presented in flow pattern maps. As most flow pattern maps are based on visual observations or transition models fitted to data obtained by visual observations, these maps still lack of objectivity in defining the flow regime transitions. In order to refine the flow regime maps and to add objective flow characteristics to the transitions boundaries, a two-phase flow sensor was developed. The sensor measures the capacitance of the two-phase flow. Because of the difference in dielectric constant of liquid and vapour and the dependency of the capacitance to the internal distribution of liquid and vapour in the cross-section of the tube, the sensor is able to characterize two-phase flow regimes. Measures were taken to improve the accuracy and reliability of the measurements. A charge/discharge transducer with a fast response was built to dynamically measure capacitance differences in the picofarad range. A large number of experiments was done with air-water flow. The setup was able to cover all flow regimes for horizontal flow in a 9mm tube. The sensor can be used as a flow regime detector. Important for obtaining good classification results, information about vapour-liquid distribution in the cross-section of the tube should be combined with time-dependent information at the measurement location. To obtain both spatial and time information, statistical parameters of the probability density function and the power spectral distributions of the signals were selected to build up a statistical classification model. Decision trees and support vector machines were used for this purpose. A high-speed camera was used as a comparison for the results of the flow detector. More than 90% of the test runs were correctly classified by both statistical techniques.
[1]
S. Borg,et al.
CYRANO: a computational model for the detailed design of plate-fin-and-tube heat exchangers using pure and mixed refrigerants
,
1997
.
[2]
Brian E. Williams,et al.
Comparison study of a cluster of four dynamic flow pattern discrimination techniques for multi-phase flow
,
1999
.
[3]
John R. Thome.
Two-Phase Heat Transfer Using No-Phase Flow Models?
,
2004
.
[4]
Anthony Widjaja,et al.
Learning with Kernels: Support Vector Machines, Regularization, Optimization, and Beyond
,
2003,
IEEE Transactions on Neural Networks.
[5]
D. Mewes,et al.
Multielectrode capacitance sensors for the visualization of transient two-phase flows
,
1997
.
[6]
Vikrant Aute,et al.
CoilDesigner: a general-purpose simulation and design tool for air-to-refrigerant heat exchangers
,
2006
.
[7]
Leo Breiman,et al.
Classification and Regression Trees
,
1984
.
[8]
Wuqiang Yang,et al.
A portable stray-immune capacitance meter
,
2002
.
[9]
Soushan Wu,et al.
Credit rating analysis with support vector machines and neural networks: a market comparative study
,
2004,
Decis. Support Syst..
[10]
René Westhovens,et al.
Prediction of dose escalation for rheumatoid arthritis patients under infliximab treatment
,
2006,
Eng. Appl. Artif. Intell..