7th International Conference on Nuclear Engineering Tokyo, Japan, April 19-23,1999 ICONE- Two-phase Flow Instability in a Double-Channel Natural Circulation Loop with Equal or Unequal Heating Powers W.L. Chen, S.B. Wang, C.R. Chung, S.S. Twu and Chin Pan Department of Engineering and System Science National fiing Hua University, Hsinchu, Taiwan 30043, ROC Fax: 886-3-5720724 Tel : 886-3-5725363 E-mail: cf,an~~fucuty.I-l fhu. edu. tw Abstract This study investigated experimentally double channel natural circulation the glass channel the two-phase flow instability phenomena loop with equal or unequal heating powers. The results of heated with a power of 5kW, and with power ratios between channel and glass channel of 1, 0.822, 0.676, 0.524 and 0.448, respectively, discussed. Two-phase is varied. in a flows with periodic or aperiodic oscillations the steel are reported and appear as the power ratio The mean flow rate in the glass channel, which is heated with a constant power of 5 kW for all the cases in this study, decreases with higher heating power in the steel channel. The ratio of mean flow rate between steel and glass channel increases linearly with increase the power ratio. Keyword: two-phase flow, instability, multiple channel, natural circulation loop l.INTRODUCTION Two-phase natural circulation process thermosiphon loops have many industrial reboiler, waste heat recovery applications, and next generation e.g., chemical nuclear reactors, because of its simplicity, high heat transfer capability and inherently passive nature. Because of safety considerations, operation of the industrial the unstable thermal-hydraulic conditions are not permissible in system. A designer or operator needs to know the conditions at which stability can be loose and what the behavior of the loop system will be. known that several different kinds of instability can arise in a two-phase 1 It is well natural circulation Copyright 01999 by JSME loop under certain conditions( Aritomi et al., 1992; Furuya et al., 1997; Hsieh et al., 1997; Jiang et al.,1 997; Wang and Pan, 1998). The instability in a multiple channel natural circulation could be more complicated due to channel-to-channel interaction. of unequal heating powers on the unstable thermal-hydraulic significant importance into the effect behavior is of great interest and for the design, operation and control of a system. The heating power is one of the key parameter influencing experimental Investigation loop the stability of a two-phase flow system. Most previous studies on double-channel natural circulation loops(eg. Aritomi et al., 1992) were conducted with equal power dissipated in both channels. It is not clear what the effect of unequal power dissipation The primary objective phenomena in the two channels on two-phase natural circulation phenomena of the present study is to investigate the two-phase is. flow instability in a double channel natural circulation loop with equal or unequal heating powers. 2. EXPERIMENTAL The experimental our previous loop employed study. The experimental APPARATUS AND PROCEDURE in this study, shown in Fig. 1, is a modified loop consists of two parallel channels heated flow passage and adiabatic riser, condenser, downcomer, version of with annular upper and lower horizontal sections (Chen et a1.1998) . One of the two parallel channels is a stainless steel tube and the other is a transparent Pyrex glass tube that allows direct flow visualization. double channel is designed with an annular passageway electric heated respectively. rods in the stainless for fluid. The heater of the The outer diameters of the steel and Pyrex tubes are 10.6 mm and 11.2 mm, The outer diameters of the annular for both channels are 20.2 mm. The total length of the heated region is l.lm but only the top 0.92m is actually heated. The length of riser is 3.65m. The two channels have common upper plenum and lower mixing tanks. The working fluid used was distilled water. A single DC power supply device controls the heating power. A special circuit was designed such that a DC power supply may provide two channels with equal or unequal heating powers. The glass channel was heated with a power of 5kW, and the power ratios between the steel channel and glass channel were 1, 0.822, 0.676, 0.524 and 0.448, respectively. The outputs recorded of thermocouples, and analyzed flowmeters by two data acquisition and differential systems: pressure transducer HP3852 and ARllOOA. are Both are connected with a personal computer. The former provides 96 channels for recording the data with a sampling rate of 0,5Hz, while the later has four channels only and can have a much higher sampling rate. The AR1 1OOA was used during parts of transient and quasi-steady intervals which are of significant interest for the measurements of the two channels, riser fluid temperature The signals of LEDs are used to synchronize state of the mass flow rate in each in the glass tube and voltage outputs of the LED. the thermal hydraulic data with the risers(Hsieh et a1.,1997) Copyright 01999 by JSME To enable independent flow measurement, section, in which an orifice flowmeter both channels have their own lower horizontal is installed. Both flowmeters forward and reversed flows. The uncertainties in measurement were calibrated for both in the low flow rate region, at which the uncertainly is the highest, for forward and reversed 4.9kg/hr, respectively, in the glass channel and &4.5kg/hr flow are rfr8.4kg/hr and+ and &7.3kg/hr, respectively, in the stainless steel channel. The purpose of this study is to investigate the effect of unequal heating power in the two channels on natural circulation behavior. Thus, the loop gage pressure is maintained m Hz0 in the joint temperature station of the condenser is kept at the exit on the primary side of the condenser the flow rate and inlet temperature of the secondary referred to as the inlet fluid temperature at the inlets is also oscillating cooling The working fluid at 60 + I “c by adjusting water. This temperature is as it is not much different from that at both channel inlets for the cases with steady circulation. controllable and the downcomer. at 6 +0.5 Under oscillation conditions, due to possible reversed the fluid temperature flow and cannot be treated as a variable. The heating powers in both channels are first gradually increased to 5 kW. Care was exercised to maneuver the operation such that the steady state is obtained with the inlet temperature and loop pressure pressurizer is disconnected of gas in the pressurizer state has maintained Thereafter, at the desired levels. At that stage, the lower with the loop, which might evil effect of the compressible on the loop thermal - hydraulic volume instability. After the quasi-steady for about 1 hour, AR1 1OOA was turned on to record the data for 52 min. the power in the steel channel is decreased to the next desired level and the process as for the case with 5kW in each channel is repeated. 3. EXPERIMENTAL Two-phase RESULTS AND DISCUSSION flow prevails when both channels are operated with equal power of 5kW. This value is about the power for the incipience of two-phase flow in the loop under the given conditions (Chen et a1.,1997). Moreover early out-of-phase channels as shown in Fig.2. The oscillation flows. The oscillation periodic flows persist in the two is also characterized by the presence of reversed frequency is about 0.06Hz. The peak forward flow rate, in the range of 500 to 600 kg/hr, is usually accompanied with slug or churn flows in the riser. The very existence of such flow patterns indicates that the void fraction in the channel is high and the buoyancy force is significant Chen et al.( 1997) provides occurrence an explanation for the out-of-phase oscillation and the of reversed flows. The reversed flow is initially caused by the hot fluid entering from top pushed by the large flow in the other channel. The significant reversed flow ‘peak’ at the end of reversed flow period results from the rapid evaporation heated anuulus of the channel. Such a rapid vapor expansion 3 and vapor expansion in the expels the liquid both upward Copyright 01999 by JSME and downward and causes the reversed flow peak. On the other hand, this expanded vapor blanket provides a very large buoyancy large force and makes the fluid flow forward again and rises sharply to its positivepeak. At this stage, the flow patterns in the riser is typically slug on churn flow. Therefore, temperature the thermocouple in the lower riser will sense the vapor and a spike is recorded, as show in Fig.2(a) (Chen et al.,l997). Such a out-of- phase periodic flow persists when the heating power in the steel channel is reduced to 4.1 lkW, as show in Fig.3. The glass channel is heated with a power higher than that in the steel channel and should have a potential to oscillate more frequently. However, the power in the steel channel is not much lower than in the glass channel and the close interaction between two channels makes them oscillate together with a single frequency. Such a nice cooperation between two channels no longer exists when the heating power in the steel channel is reduced channel. Figure 4 illustrates oscillates more frequently conceivable to 3.38kW, i.e., 0.676 times of heating power in the glass the chaotic-like oscillations in the two channels. in the glass channel though the oscillation The flow is not periodic. This is in that the heating power in the glass channel is higher than that in the steel one. In addition to the high temperature peaks”, low temperature time. This indicates that part of the low temperature glass channel, at which the riser fluid temperature spikes also appear form time to fluid may be able to reach the riser of the is measured. Such an irregular oscillation continues to display when the heating power in the steel channel is further decreased to 2.62 kW, i.e., the power ratio between steel and glass channel Although the time plots look like periodic oscillations, is 0.524, as shown in Fig.5. the power spectrum of the flow rate in the glass channel presents no distinct peaks. The intensity of the one distinct peak at 0.04Hz for the flow rate in the steel channel is relatively low compared to those in Fig.2 or Fig.3. Therefore, the oscillation is considered irregular. The flow becomes periodic again when the heating power in the steel channel is further reduced to 2.24kW, i.e., 0.448 times of that in the glass channel, which is apparent in Fig.6 The flow in glass channel is of double frequencies, i.e., 0.035 and 0.071H2, while the flow in the steel channel is of single frequency of 0.035 Hz. The low frequency oscillation channel due to its relatively low heating power is able to match the fundamental the glass channel. Consequently, the periodic oscillation in the steel frequency of in the system may be resumed. It is also worth noticing that the flow in the steel channel is characterized by the presence of a relatively long-last reversed flow. During the reversed flow interval in the steel channel there are two forward flow excursions in the glass channel. The power ratio between two channels also has a significant effect on the mean flow rate, as shown in Fig.7. The mean flow rate in the glass channel, which is heated with a heating power of 5 kW for all the cases in this study, decreases with increase in the heating power in the steel channel. This is because the reversed flow interval in the glass channel is 4 Copyright 01999 by JSME increased with increase in the heating power of the steel channel The mean flow rate in the steel channel increases with increase in heating power as expected resulting in the elevation of total loop flow rate with increase in total heating power and the mean flow rate ratio between two channels increases approximately As shown in linearly with increase in power ratios. Fig7c, the ratio of flow rate in the two channels can be expressed by the following equation: QC-- 0.91 -.--L- E=2.31 QA where FC / FA is mass flow rate ratio between steel and glass channel; Qc /Q, is heating power ratio between steel and glass channel. 4. CONCLUSION This study explores experimentally double channel natural circulation demonstrate the significant circulation phenomena. the two-phase flow instability of power ratio between If a power ratio results in oscillation frequency power of 5 kW, periodic oscillation in a loop with equal and unequal heating powers. The results effect matches the fundamental phenomena two chartnel frequency on the natural in the steel channel in the glass channel which is heated with a constant will appear in the system. If not, aperiodic prevails in the two channel. Such a significant interest for the operation of a muti-channel oscillations effect of heating power ratio is of significant natural circulation loop. ACKNOWLEDGEMENTS This work was supported by the National Science Council of Republic of China under Contract No. Nsc84-2212-E-007-068. REFERENCES 1. M. Aritomi, J.H. Chang and M. Mori, 1992, Fundamental Studies on Safety-Related Thermo-Hydraulics of Natural Circulation Boiling Parallel Channel Flow Systems Under Startup Conditions (Mechanisms of Geysering in Parallel Channels),” Nuclear Safety, Vol.33, pp. 170-182. 2. M. Furuya, F. Inada and A. Yasuo, 1997, “Thermohydraulic Circulation Loop with a Chimney (Part 2. Experimental Instability in Detail),” Heat Transfer-Japanese 3. C.C. Hsieh, S.B. Wang and C. Pan, 1997, Instability of Boiling Natural Approach to Clarify the Flow Research, Vol. 24, pp. 577-588. Dynamic Visualization of Two-Phase Flow Patterns in a Natural Circulation Loop,” Int. J. Multiphase Flow, Vol.98 pp. 616-622. 4. S.Y. Jiang, X.X. Wu, S.R. Wu, H.H. Bo, Y.J. Zhang and P, Han, 1997, 5 Copyright Experimental 01999 by JSME Study on Flashing Concerned S.B. Wang and C. Pan, 1998, Circulation in a Natural Kerntechnek, Heating Reactor Conditions,” 5. Instability Circulation system at Nuclear Vol. 62, pp. 148-153. Two-phase Flow Instability Loop Using the Taguchi Method”, Experimental Experiment in a Natural Thermal and Fluid Science, Vol. 17, pp. 189-201. 6. W.L. Chen, S.B. Wang, S.S. Twu, C.R. Chung and C. Pan, 1998, Double Channel Engineering, Natural Circulation Loop,” Proc. of @ Int. Hysteresis Conf. Effect in a On Nuclear (in CD-ROM), San Diego, CA. May 10-14, 1998. Fig. 1 The double channel natural circulation loop 6 Copyright 01999 by JSME 0 50 100 150 TIME (see) 125 ,~~~ 0.000.05o.*o0.15 0.20 Frequency 250 200 _ 0.25 (Hz) 300 - ~. 0.35 0.30 0.40 Fiq. 2 Time evolution of thermal hydraulics for both channels operated at 5.0 kW. (a) riser fluid temperature in the glass channel (b) flow rate in the glass channel(FA) (c) flow rate in the steel channel(FC) (d) power spectrum of FA (e) power spectrum of FC. Q078TS02 0 50 100 3,s 0.00 0.05 Fiq. 3 Time evolution l6m 0.10 60’1: 150 TIM f (see] 0.15 0.20 Frequency of thermal 4.113 250 200 0.25 (Hz) kW) 0.35 0.30 hydraulics 300 0.40 for the power ratio (steel / glass) of 0.822. (a) riser fluid temperature in the glass channel (b) flow rate in the glass channel (FA) (c) flow rate in the steel channel (FC) (d) power spectrum of FA (e) power spectrum of FC. 7 Copyright 01999 by JSME Ii6 fi, ii ; ~. 0.00 0.05 Fiq. 4 Time evolution 0.10 0.15 0.20 0.25 Frequency(Hr) 0.30 of thermal hydraulics 0.35 0.40 for the power ratio (steel / glass) of 0.676. (a) riser fluid temperature in the glass channel (b) flow rate in the glass channel (FA) (c) flow rate in the steel channel (IX) (d) power spectrum of FA (e) power spectrum of FC. 1078SSO2 0.00 Fiq. 5 Time evolution 0.05 0.90 (6m 60% 0.15 0.20 Frequency 2.62:6 0.25 (Hz) of thermal hydraulics kW) 0.30 0.35 0.40 for the power ratio (steel / glass) of 0.524. (a) riser fluid temperature in the glass channel (b) flow rate in the glass channel (FA) (c) flow rate in the steel channel (FC) (d) power spectrum of FA (e) power spectrum of FC. 8 Copyright 01999 by JSME 1000 $ 500 01 5. Y O 0 9 -q Qk 2 9 g Q 100 150 100 50 250 200 300 TlME (WC) - 75 50 25 100 80 60 ~~ 40 0.00 0.05 0.10 0.15 0.25 0.20 Freq*ncy 0-w 0.30 0.35 0.40 Fiq. 6 Time evolution of thermal hydraulics for the power ratio (steel / glass) of 0.448. (a) riser fluid temperature in the glass channel (b) flow rate in the glass channel (FA) (c) flow rate in the steel channel (FC) (d) power spectrum of FA (e) power spectrum of FC. 9 Copyright 02999 by JSME 0 + FA - - - - FC 9.0 8.0 Total Power (kW) 0.40 0.50 0.60 0.70 0.80 0.90 1 .oo Power Ratio (QJQA) Fiq. 7 Effect of power ratio (steel / glass) on mean flow rate. 10 Copyright 01999 by JSME