Int. J. Multiphase Flow Vol. 14, No. 2, pp. 175-187,1988 0301-9322/88 $3.00+0.00 Printed in Great Britain.All rightsreserved Copyright © 1988Pergamon Press/Elsevier THE INFLUENCE OF NON-CONDENSIBLE GAS ON TWO-PHASE CRITICAL FLOW G. P. CELATA, M. CUMO, F. D'ANNIBALE and G. E. FARELLO ENEA Casaccia TERM/ISP Heat Transfer Laboratory, Via Anguillarese 301, 00060 Rome, Italy (Received 7 April 1987; in revised form 25 October 1987) Abstract--The present paper deals with a two-phase steam-water critical flow experimentin long tubes, in which known air flow rates are injected into the stagnation region. The aim of the experiment is to detect the influenceof non-condensiblegas on the two-phase critical mass flux as well as to establish the limit, in terms of air concentration, beyond which the critical flow is affectedby the presenceof the gas. The test section is a vertical, circular channel with i.d. 4.6 mm and a length of 1500 mm (LID = 325). Results of experiments with initially subcooled liquid (together with some data from saturated liquid discharges), up to pressures of 1.5 MPa are reported together with the analysis of the effects of the non-condensibles under the different stagnation conditions. 1. INTRODUCTION In nuclear reactor safety analyses, the most serious accident sequences are represented by those caused by a break in the pressure boundary of the primary loop. During the last few years, both experimental and theoretical studies concerning this kind of accident have been intensified because of the growing attention paid to the so-called "small break" LOCAs (SB-LOCA), especially after the 1979 Three Mile Island accident. An SB-LOCA, in comparison to a "large break" LOCA (LB-LOCA), has, among others, two main disadvantages which make it more important for consideration: (a) an SB-LOCA has a potential occurrence frequency much higher than that of an LB-LOCA because of the larger number of small diameter pipes; (b) since human actions to mitigate the consequences of the accident are possible, an in-depth knowledge of the transient following the break is required. Research in the broad area devoted to the analysis of the above-mentioned problems has been reported by Ardron (1978), Dobran (1985), Flinta (1983), Henry (1968, 1970a, b), Henry & Fauske (1971), Ilic et al. (1986), Lackm6 (1982), Moody (1965, 1966), Reocreux (1974, 1976), Weigand et al. (1983) and in various Proceedings (1981, 1983, 1985) as far as the international situation is concerned. ENEA theoretical and experimental contributions on two-phase steam-water critical flows have been reported by Celata et al. (1983a, b, 1985, 1986a, b). Very little information is available in the literature about the influence of non-condensible gas on the critical two-phase flow behaviour with respect to a reference situation (degassed liquids). On the other hand, in a nuclear reactor, under severe accident conditions, the rate at which non-condensibles can be released into the coolant is significant. The situation is also of great interest for safety considerations in the chemical industry. In the present work, an experiment on critical two-phase flows with air injected into the two-phase steam-water mixture is reported and the results are discussed. 2. THE EXPERIMENTAL SETUP The experimental tests were carried out with the steam-water loop sketched in figure 1. The loop consists of two cylindrical pressure vessels, each having a capacity of 1001., an electric heater (10 kW) for water heating and a centrifugal pump for the recirculation of the liquid from the main vessel (St) through the electric heater. Vessel S~, filled with demineralized water, simulates the reference pressure vessel, whilst vessel $2 allows volumetric expansion of the water during heating and acts as a pressurizer: it is partially filled with cold water and pressurized with nitrogen. 175