HEAT TRANSFER CAPABILITY OF A THERMOSYPHON HEAT TRANSPORT DEVICE WITH EXPERIMENTAL AND CFD STUDIES B.M. Lingade a*, Elizabeth Raju b, A Borgohain a, N.K. Maheshwari a, P.K.Vijayan a a Reactor Engineering Division, Bhabha Atomic Research Centre, Trombay, Mumbai-400085 India *Email: lingadeb@barc.gov.in b Manipal Institute of Technology, Manipal University, India ABSTRACT Advanced nuclear reactor designs envisage use of passive means of heat transfer during normal operation and accidental conditions. Thermosyphon device is a passive device that can be used for heat exchange applications where heat source and heat sink are separated by some distance. Thermosyphon device operates on natural circulation principle in which the flow is driven by thermally generated density gradients without use of pumps. This paper deals with studies on thermal hydraulic behaviour of a thermosyphon device. Experiments are performed in the thermosyphon device using water as working fluid. CFD simulations with FLUENT are also carried out to simulate experimental conditions. The results obtained from CFD studies are compared with experimental data. Keywords: Nuclear reactor, thermosyphon, CFD, experiment, thermal hydraulics. INTRODUCTION The conventional nuclear reactors have seen an extensive use of active engineering safety systems for reactor control and protection. The active systems rely on operator action and actuation of active components to mitigate the consequences of accident. Since the reliability of active systems cannot be improved beyond a threshold, advanced reactor envisage extensive use of passive safety systems. The Indian Advanced Heavy Water Reactor (AHWR) have been designed and developed to achieve large scale use of thorium for the generation of commercial nuclear power. The AHWR incorporates a number of passive safety features and is associated with a closed fuel cycle having reduced environmental impact. Passive safety systems operate based on natural physical laws such as gravity, buoyancy etc. and do not require external source of energy or operator action. Since potential cause of failure of active system such as lack of human action or power failures do not exist when passive systems are employed. The proposed thermosyphon device is a passive device that can be used in heat transfer applications of reactor. The device consists of heated and cooling sections through which density gradient is generated. This gradient establishes natural circulation (thermosyphon effect) in the device. Thermosyphon is employed in nuclear field for effective heat removal as driving head increases with power. The experiments are carried out to study heat transfer behaviour of thermosyphon device. CFD analysis has been performed using FLUENT. Results of CFD analysis are compared with experimental data. LITERATURE REVIEW Over the last few years many studies have been devoted to heat transfer devices based on a recirculating mass which may be accompanied with and without phase change such as, heat pipes and single-and two phase closed or open thermosyphon. Grover [1] introduced the term heat pipe in 1964. The two-phase closed thermosyphon used in this study was a gravity-assisted wickless heat pipe, which was very efficient for the transport of heat with a small temperature difference via the phase change of the working fluid. It consisted of an evacuated-closed tube filled with a certain amount of a suitable pure working fluid. Park et al. [2] studied the heat transfer characteristics in thermosyphon depending on the amount of working fluid and when the operation limits occur. Noie [3] presented in his work an experimental study of a two phase thermosyphon of (980 mm length and 25 mm internal diameter) made of smooth copper tube, with distilled water as a working fluid. They determined the temperature distribution at the inside wall of thermosyphon device. Mohammed et al. [4] studied heat
transfer performance of two phase thermosyphon experimentally with different cross section shapes for the thermosyphon tube. A thermosyphon with three different cross section shape (circular, square and rectangular) having the same hydraulic diameter and length were used. Methanol was used as the working fluid. Their results showed that the heat transfer coefficient increases with the increase of input power and thermal resistance is indirectly proportional to the input power. Vijayan et al. [5] studied the steady state and stability characteristics of single phase natural circulation in a rectangular loop with different heater and cooler orientations. The various performance parameters of a natural circulation system could be understood from his studies. THERMOSYPHON DEVICE DESCRIPTION The experimental setup for the thermosyphon device is made up of borosilicate glass as shown in Fig. 1. (a): Schematic of the facility (b): Photograph of the facility Fig. 1: Experimental setup of thermosyphon device The device consists of heater section at bottom and cooler section at top. It consists of two concentric annular tubes for circulating primary fluid. The bottom cap was insulated to avoid heat gain from that portion of device. Cooler is annular type and has one inlet and one outlet. Water is used as the working fluid inside the device and coolant inside the cooling jacket. The mass flow rate of coolant inside the cooler is measured by allowing a known volume of water to pass through it in a specified time. The inlet and outlet temperature of coolant is measured with the aid of k type thermocouples. Thermocouple TC-1 measures heater surface temperature and TC-2, TC-3 measures primary medium temperatures along the height. EXPERIMENTAL STUDIES Experiments have been performed by maintaining heater section at constant temperature. System was allowed to achieve steady state condition by observing values of process parameters. The water inside the thermosyphon device transfers the heat from heater wall and rises through the annulus. The heated fluid enters the cooler portion towards the top and transfers heat to coolant in the jacket. After getting cooled water comes down from the central core of the device and thus make recirculating path. Further, experiments have been repeated with higher water temperature at heated section. Water is
used as cooling medium during experiments. Table 1 gives various measured and calculated parameters at steady state condition of experiments. Maximum heater surface temperature is limited below boiling point of heating water as indicated by thermocouple TC-1 in table 1. Heat removal rate of the thermosyphon device is calculated from heat balance across cooler using equation (1). (1) In one dimensional analysis the only coordinate, z, runs around the loop with origin at the inlet of the heater. The governing momentum equation for incompressible fluid can be written as, ( ) (2) Neglecting fluid acceleration, steady state solution of momentum equation is obtained by cyclic integration ( ) (3) Where, N p and N b are number of pipe segments and bends in the loop respectively. Cyclic integral is evaluated with assumption of Boussinesq approximation to be valid so that the density can be expressed as [ ] (4) Table 1 gives the experimental data collected for the device. From equation (3) and (4) mass flow rate of primary medium in thermosyphon device is calculated and is given in Table 1. Pressure drop in thermosyphon device is calculated as summation of pressure drop in all section of device. Table 1: Steady state results of five experimental cases Experi -ment No. Cooling water flow rate (ml/s) Cooling water temperature TC-1 TC-2 TC-3 Primary medium flow rate (kg/s) Total pressure loss (pa) Heat removal rate (W) Inlet Outlet TH-1 7 34.0 34.5 60.4 44.4 43.6 0.0018 11.4 14.63 TH-2 7 34.8 35.5 65.4 48.5 46.0 0.0021 14.5 20.78 TH-3 7 35.0 35.9 76.1 53.3 52.8 0.0025 18.6 26.33 TH-4 7 35.0 36.2 88.5 62.0 61.4 0.0026 19.4 35.11 TH-5 7 35.0 36.5 95.2 68.4 65 0.0028 22.1 43.89 From table 1 it is clear that heat removal capacity and coolant flow rate increases with increasing heat surface temperature (TC-1). While difference between heater surface temperature and coolant temperature is also increases with heater temperature. The maximum heat removal capacity of thermosyphon device is observed to be 43.89 W at heater temperature 95.2 o C. Flow is turbulent in all cases studied because minimum value of Reynolds number was calculated to be 903. CFD ANALYSIS FOR THERMOSYPHON DEVICE The numerical studies have been carried out which can predict experimental behaviour of the device. The geometric model of the device was generated and meshed with the aid of GAMBIT 2.4.6. Structured hexahedral mesh was used for the annulus space and tetrahedral mesh for the fluid inside the device. Since flow is governed by buoyancy force, Boussinesq approximation is assumed to consider density variation due to temperature change. Buoyancy force direction is along Y axis, whereas coolant inlet and outlet are within X-Y plane in the CFD modelling. Ansys FLUENT 15 is used for the analysis of the device. Analysis is carried out with SIMPLE algorithm and k-epsilon turbulent model. Relaxation factors are taken to be default values. Convergence criteria for continuity
Temperature (deg C) Pressure drop (pa) Thorium Energy Conference 2015 and energy are set to 1 10-6. For analysis, in the heater section, temperature boundary condition is specified. The cooling water inlet temperature and mass flow rate is specified in the analysis. As device was not insulated heat loss to ambient was also considered in the analysis. The temperatures at thermocouple locations are predicted. Further, the heat removal capacity and pressure drop of the device is calculated. RESULTS AND DISCUSSION Steady state CFD analysis was carried out to find heat transfer capability and thermal hydraulic behaviour of thermosyphon device. The steady state results predicted using CFD has been compared with the experimental data as shown in Fig. 2(a). Thermocouples TC-2 and TC-3 show temperatures along the elevation of device in annular space. It can be seen that the temperature predicted in CFD analysis matches well with the experimental data. A maximum deviation is estimated to be about 4.35%. Figure 2 (a) further shows that with large increase in heater temperature there is small increase in cooling water outlet temperature. This may be due to increase in thermosyphon flow with increase in heater section temperature. Figure 2 (b) shows comparison of total pressure drop between CFD results and calculated by equation (2). 80 70 60 TC-2 (Exp) TC-2 (CFD) TC-3 (Exp) TC-3 (CFD) Coolant outlet (Exp) Coolant outlet (CFD) 35 28 21 Analytical CFD 50 14 40 30 7 20 50 60 70 80 90 100 0 50 60 70 80 90 100 Heater temperature (deg C) Heater temperature (deg C) (a): Temperature prediction (b): Pressure drop prediction Fig. 2: Comparison of experimental and CFD results Figure 3 shows velocity vectors and temperature contour for experiment number TH-1. For two different perpendicular planes, velocity vectors are shown in annular space above heater section. Vector plot shows that primary medium is flowing in two different directions (i.e. bidirectional flow) in two different planes. It can also be seen from temperature contour plot given in Fig. 3 (a). Temperatures are non-uniform across cross section of the device. At low heater surface temperature, buoyancy force is not sufficiently strong. This may lead to local recirculation in annular space of primary medium. Further heat transfer is better at inlet and outlet lines of cooling water (Inlet and outlet lines of cooling water was in XY plane) due to which velocity is in +Y direction. Due to lower buoyancy force and local recirculation, the flow is in downward direction (-Y direction) in annular region in ZY plane. This kind of behaviour is not observed with high heater surface temperature and high power conditions as shown in Fig. 4 (a). At this condition primary fluid, is flowing in unidirectional (upward direction) manner in annulus as buoyancy force is strong. Figure 4(a) shows that uniform temperature contours observed within annulus and core. Primary medium is flowing in upward direction within annulus and flowing downward within core region. Figure 4(b) shows temperature contours in Y-Z plane for experiment number TH-5. For tube in tube type of thermosyphon device with smaller annular space flow is expected to take place in upward direction in annular region under sufficient buoyancy head.
(a): Temperature contour (X-Z plane) (From bottom (b): Velocity vectors in X-Y and Z-Y plane 0.06 m above) Fig. 3: Temperature and velocity contours at cross section above the heated section (a): Temperature contour at height 0.06m (X-Z plane) (b): Temperature contours at mid section of device Fig. 4: Steady state CFD results for experiment no. TH-5 CONCLUSIONS In glass type thermosyphon device, experiments with water have been performed at atmospheric pressure and below boiling temperature. Steady state CFD analysis is also carried out and the temperatures predicted using CFD is in good agreement with experimental results. Bidirectional flow was observed at low heater surface temperature and unidirectional flow observed at high heater surface temperature. NOMENCLATURE Symbol Description [Unit] Greek Symbols C p Specific heat [kj/ o C] Density [kg/m 3 ] D Diameter (m) μ Dynamic viscosity (Pa.s) H Elevation difference between heater and Thermal expansion coefficient cooler section (m) (k -1 ) K Local loss coefficent Dynamic Viscosity (m 2 /s) k Thermal conductivity (W/m o k) φ Inclination angle of the pipe
L Length (m) with respect to vertical direction Mass flow rate (kg/s) SUBSCRIPTS P Power (W) c Cooling water Pr Prandlt number h Hydraulic diameter p Pressure (pa.s) p Primary medium T Temperature (k) s Surface u Velocity (m/s) REFERENCES [1] Grover, G. M. (1964). Evaporation-condensation heat transfer device. US Patent No. 3229759 [2] Park J., Kang K., Kim J., Heat transfer characteristics of a two-phase closed thermosyphon to the fill charge ratio, International Journal of Heat and Mass Transfer; 2002, Vol.45, (23), pp. 4655-4661 [3] Noie S.H, Heat transfer characteristics of a two-phase closed thermosyphon, Applied Thermal Engineering; 2005, Vol.25 (4), pp. 495-506. [4] Mohammed M. I. Hammad, Jasem H. ALsuwaidi, Experimental investigation of the eat transfer coefficient of the thermosyphon cross section shape, Int. Journal of Engineering Research and Applications; ISSN : 2248-9622, Vol. 5, Issue 3, ( Part -4) March 2015, pp.57-66 [5] Vijayan, P.K., Experimental observations on the general trends of the steady state and stability behaviour of single phase natural circulation loops, Nuclear Engineering & Design, vol.215, 2002, 139-152.