A. Teyssedou

Optical fiber probe to measure local void fraction profiles.

A fiber optic probe for measuring the local void fraction in a two-phase (air-water) flow has been developed. The liquid-gas and gas-liquid interfaces are detected by the normal reflections of IR laser light at the tip of a multimode fiber. The probe and its associated electronic circuitry were calibrated by using the averaged void fraction measured by a quick closing valve system. The void profiles obtained by this method were compared with those obtained by a conductivity local probe. The proposed optical system shows a better sensitivity at high void fractions.

The effect of turbulent mixing models on the predictions of subchannel codes

Abstract In this paper, the predictions of the COBRA-IV and ASSERT-4 subchannel codes have been compared with experimental data on void fraction, mass flow rate, and pressure drop obtained for two interconnected subchannels. COBRA-IV is based on a one-dimensional separated flow model with the turbulent intersubchannel mixing formulated as an extension of the single-phase mixing model, i.e. fluctuating equal mass exchange. ASSERT-4 is based on a drift flux model with the turbulent mixing modelled by assuming an exchange of equal volumes with different densities thus allowing a net fluctuating transverse mass flux from one subchannel to the other. This feature is implemented in the constitutive relationship for the relative velocity required by the conservation equations. It is observed that the predictions of ASSERT-4 follow the experimental trends better than COBRA-IV; therefore the approach of equal volume exchange constitutes an improvement over that of the equal mass exchange.

PRESSURE LOSSES CAUSED BY AREA CHANGES IN A SINGLE-CHANNEL FLOW UNDER TWO-PHASE FLOW CONDITIONS

Experimental data obtained on plate and smooth blockages under two-phase flow conditions in a square vertical channel have been presented and analysed using the Janssen-Kervinen and momentum-energy models. The vena contracta coefficients obtained for plate blockages using these models agree well up to about 30% void fraction. It is observed that the contraction coefficients for two-phase flow differ somewhat from those for single-phase flows. The irreversible pressure loss coefficients for plate and smooth blockages depend on blockage severity and void fraction.

Critical heat flux data in a vertical tube at low and medium pressures

Abstract AECL Research and Ecole Polytechnique have been cooperating on the validation of the critical heat flux (CHF) look-up table (D.C. Groeneveld et al., Heat Transfer Eng. 7(1–2) (1986) 46–62). For low and medium pressures the values in the table have been obtained by extrapolation and curve fitting; therefore, errors could be expected. To reduce these possible extrapolation errors, CHF experiments are being carried out in water cooled 8 mm internal diameter (ID) tubes, at conditions where the data are scarce. This paper presents some of the experimental CHF data obtained for vertical up flow in an 8 mm ID test section, for a wide range of exit qualities (5–70%) and the exit pressure ranging from 5 to 30 bar. The experiments were carried out for heated lengths of 0.75, 1, 1.4 and 1.8 m. In general, the collected data show parametric trends similar to those described in the open literature. However, it was observed that for low pressure conditions CHF depends on the heated length; this dependence begins to disappear for exit pressure of about 30 bar. The CHF data have also been compared with predictions of well-known correlations (L. Biasi et al., Energia Nucl. 14(9) (1967) 530–536; R. Bowring, Br. Report AEEW-R789, Winfrith, UK, 1972; Y. Khatto and H. Ohno, Int. J. Heat Mass Transfer 27 (1984) 1641–1648) and those of the look-up table given by Groeneveld et al. For low pressures and low mass fluxes the look-up table seems to yield better predictions of the CHF than the correlations. However, for medium pressures and mass fluxes the correlations perform better than the look-up table; among those tested, Katto and Ohnos correlation gives the best results.

Moderator Flow Simulation Around Calandria Tubes of Candu-6 Nuclear Reactors

Abstract CFD simulations of cross-flows along in-line and staggered tube bundles which emulate those encountered in the calandria of CANDU-6 reactors are presented. The knowledge of external wall temperature distributions around calandria tubes is a major concern during normal and off-normal operating conditions of CANDU reactors. Calculations are performed using the FLUENT software with several turbulence models using segregated and Coupled algorithms. It is observed that κ-based models are able to reproduce mean velocities in staggered bundles. In most cases, the Coupled algorithm yields convergence even if it requires a longer computational time. Based on this work, the standard κ-ε model is recommended to perform this kind of simulations. Improved κ-ε models do not lead to better results while the κ-ω model predicts very well the physics only around the first row but it is unable to predict the flow around tubes located far downstream in the bundle.

Critical heat flux in a vertical tube at low and medium pressures: Part I. Experimental results

Abstract This paper presents experimental CHF data obtained for vertical up flow in an 8 mm I.D. test section, for a wide range of exit qualities (5–75%) and exit pressures ranging from 5 to 40 bar. The experiments were carried out for heated lengths of 0.75, 1, 1.4, 1.8, 2.5 and 3.5 m. A number of different coordinate systems are used to present the experimental results, these include CHF as a function of the inlet subcooling, as a function of the outlet thermodynamic quality, and as a function of the boiling length. An analysis of these commonly used representations of the CHF, has shown, that for all of the cases studied, the critical heat flux is dependent in some manner on the heated length. This dependence greatly complicates the possibility of obtaining a universal correlation for the critical heat flux at low and medium pressures.

Implementation and Application of the Lattice Boltzmann Method Using Matlab

Abstract: The Bhatnagar-Gross-Krook version of the Lattice Boltzmann method on two-dimensional Cartesian meshes has been used to develop a computational program suitable for the Matlab environment. The basic algorithm is implemented with a grid refinement approach that includes an accurate boundary treatment to handle complex geometries in the lattice Boltzmann framework. Validation of the program was conducted on a steady flow over a staggered bundle of tubes for which data are available. After this preliminary step, the code was applied to the calculation of the characteristic aerodynamic parameters at steady state around clean and three ice-accreted NACA63–415 profiles at Re = 500 at α = 0° and 8°. Finally, the unsteady nature of the method was used to compute flows over this clean and ice-accreted airfoil for an angle of attack of α = 28°. For all configurations, results were compared with computational results carried out with the commercial package fluent.

A capacitive two-phase flow slug detection system

A capacitive-type two-phase slug detection and characterization system has been developed. The main goal of this work consisted of developing a measurement technique that satisfies the following requirements: to be able to detect a wide range of liquid film thickness from a few millimeters up to the case when the channel is completely filled with water, to have a fast time response, to be nonintrusive and movable, to have a linear response for void fraction of up to 70%, and to be insensitive to the structure of the test section and to proximity effects. In order to detect very small changes in capacity required to fulfill these requirements, a nonlinear twin-T type bridge is used as the front end stage of the instrument. Furthermore, the linearity and stability response criteria are achieved by the appropriate selection of the geometry and topology of the electrodes; the calibration of the system shows an almost linear response for void fractions of up to 70%. The instrument is presently used to determin...

An investigation of the constitutive relations for intersubchannel transfer mechanisms in horizontal flows as applied in the ASSERT-4 subchannel code

Abstract In this paper, the influence that the constitutive relations used to represent some of the intersubchannel transfer mechanisms has on the predictions of the ASSERT-4 subchannel code for horizontal flows is examined. In particular the choices made in the representation of the gravity driven phase separation phenomena are analyzed. This is done by comparing the predictions of the ASSERT subchannel code with experimental data on void fraction and mass flow rate, obtained for two horizontal interconnected subchannels. ASSERT uses a drift flux model which allows the two phases to have different velocities. In particular ASSERT contains models for the buoyancy effects which cause phase separation between adjacent subchannels in horizontal flows. This feature, which is of great importance in the subchannel analysis of CANDU reactors, is implemented in the constitutive relationship for the relative velocity. In order to isolate different intersubchannel transfer mechanisms, three different subchannel orientations are analyzed. These are the two subchannels at the same elevation, the high void subchannel below the low void subchannel, and the high void subchannel above the low void subchannel. It is observed that for all three subchannel orientations ASSERT does a reasonably good job of predicting the experimental trends. However, certain modifications to the representation of the gravitational phase separation effects which seem to improve the overall predictions are suggested.

Critical heat flux under choking flow conditions. Part I - Outlet pressure fluctuations

Under conditions of forced convective boiling at low pressures and high mass fluxes, beyond a certain quality, choking flow may occur at the exit of a heated channel. An experimental investigation of the influence of this phenomenon on the critical heat flux (CHF) has been carried out. An analysis of the pressure distribution downstream of the heated length as well as the behaviour of the wall temperature during CHF has shown that the occurrence of choking flow, namely outlet pressure fluctuations, does not radically influence the values of the CHF. The physical reasons for the appearance of these outlet pressure fluctuations under conditions of diabatic choking flow have been analyzed. The experimental investigation has also shown that when choking flow occurs the outlet pressure cannot be lowered below a certain value that is fixed by the flow itself. The maximum values of flow parameters attained under choking flow conditions are presented in the Part II of this paper.