2-FLUID MODEL STUDIES FOR HIGH-DENSITY 2-PHASE LIQUID-METAL VERTICAL FLOWS

Citation
P. Satyamurthy et al., 2-FLUID MODEL STUDIES FOR HIGH-DENSITY 2-PHASE LIQUID-METAL VERTICAL FLOWS, International journal of multiphase flow, 24(5), 1998, pp. 721-737
Citations number
23
Categorie Soggetti
Mechanics
ISSN journal
03019322
Volume
24
Issue
5
Year of publication
1998
Pages
721 - 737
Database
ISI
SICI code
0301-9322(1998)24:5<721:2MSFH2>2.0.ZU;2-K
Abstract
Liquid metal magnetohydrodynamic power converters (LMMHD PC) have been recently proposed for electrical power generation. These systems cont ain two-phase vertical flows consisting of high density liquid metals and suitable gas-vapor. Optimum design of LMMHD power plants require a ccurate modeling of two-phase flows in the riser. A two-fluid model ha s been developed for this purpose. One-dimensional, steady state two-f luid flow equations consisting of conservation of mass, momentum of ea ch phase along with auxiliary relations have been solved numerically b y the Runge-Kutta method. Interfacial drag force corresponding to mult i-bubble, churn turbulent and slug flow based on Ishii et al. and Tait el classification has been used. Effect of variation of void fraction and phase velocities of the fluids across the cross section of the pip e has been studied based on Ishii et al. model by modifying relative v elocity and incorporating appropriate coefficients in the conservation equations. Bubble size at the mixer orifice exit has been calculated using the equations of Kumar et al. In order to verify the accuracy of the model, a nitrogen-mercury experimental system has been set up. Vo id profiles have been measured using gamma-ray attenuation method. Voi d fraction, slip and pressure at different locations were determined f or the mass fluxes varying from 0.125 to 2.302 kg/sm(2) for nitrogen a nd 5.52 x 10(3) to 12.26 x 10(3) kg/sm(2) for mercury. The predicted v alues have been compared with the experimental data. The void fraction values matched well with the experimental data within 10% and within 20% when cross-sectional effects were included. The over all pressure values were within 13% and 8%, respectively, while the slip values dev iated within 25% and 27%, respectively. In general, the model matched better with experimental data when the cross-sectional effects were no t included. This is due to the high density of the liquid metal and re latively larger pipe diameter. (C) 1998 Elsevier Science Ltd. All righ ts reserved.