62 CHAPTER 5 CONVECTIVE HEAT TRANSFER COEFFICIENT 5.1 INTRODUCTION The primary objective of this work is to investigate the convective heat transfer characteristics of silver/water nanofluid. In order to do this, a convective heat transfer experimental test facility is designed and constructed. The test facility consists of a tube-in-tube heat exchanger test section with high accuracy measuring devices for temperature, flow rate and pressure. A multi channel data logger is used to record the signals automatically from these devices. The post processing of the data is done using a MATLAB programme for the estimation of parameters, such as the heat transfer coefficient and pressure drop. The test facility is validated using the experimental results with de-ionized water against the published results and correlations. This chapter deals with the working principle of the test facility, measurement devices used in the facility, repeatability, long term stability study of the nanoparticles in the base fluid, actual experimental procedure, data reduction and the uncertainty analysis of the study. 5.2 EXPERIMENTAL FACILITY The test facility consists of two flow circuits, one for the cooling water and the other for nanofluid as shown in Figure 5.1. The nanofluid circuit consists of several units such as a pump, a flow meter, a constant temperature cooling water bath, a nanofluid bath with a temperature controller.
Nanofluid Circuit Cooling system Cooling Water Circuit Flow meter (Cooling water) W 2 W 1 By-pass valve Constant Temperature Bath Stirrer W 3 T Water in Flow control valve (cooling water) Pump (cooling water) Differential pressure transmitter Heater P T W 1 T W2 TW 3 T W 4 T W5 T W 6 T Nano out T Nano in N 4 N 3 T N6 T N5 T N4 T N3 T N2 T N1 Nanofluid Tank N 1 By-pass valve N 2 Mass flow meter (nanofluid) T Water out W 4 Heater Pump (nanofluid) Flow control valve (nanofluid) Figure 5.1 Schematic diagram of the experimental test facility 63
64 The test section is a 2.94 m long counter flow horizontal tube-intube heat exchanger with nanofluid flowing in the inner tube while the cooling water flows in the annular side. In the test section, the temperature of the nanofluid and the cooling water are measured at regular intervals. The test section configuration is explained in section 5.3. In the nanofluid circuit, the silver/water nanofluid is pumped into the inlet of the test section at different temperatures and flow rate. It is cooled by the cooling water which flows in the annulus in the counter flow direction. Similarly, in the cooling water circuit, the heated water leaving the test section flows into a constant temperature bath, where the heated water is cooled by an auxiliary cooling system and pumped back into the test section using a centrifugal pump. The test section (counter flow horizontal tube-in-tube heat exchanger) is thermally insulated to its full length by ceramic wool, in order to minimize the heat loss to the surroundings. The heat infiltration in the test setup for the extreme conditions of T = 40 o C and T Wi = 25 o C is theoretically estimated to be 215 W/m 2. This is just 4% of the minimum heat flux (5237 W/m 2 ) encountered in the study. Thus the possible infiltration could be ignored. The mass flow rate of the nanofluid is measured using a Coriolis mass flow meter with ± 0.1 % accuracy. The pressure drop of the nanofluid is measured between the inlet and exit of the test section using an absolute piezo-resistive differential pressure transmitter with ± 0.075 % accuracy. The PT100 thin film temperature detectors (RTD s) with ± 0.15 o C accuracy are used to measure the temperature at all the necessary points. The mass flow rate of the cooling water is measured using a suitable glass tube rota-meter with ± 2 % accuracy. The heat transfer test section which consists of a tubein-tube heat exchanger is placed inside an insulated wooden box casing on a horizontal work bench. The differential pressure transmitter indicator, the nanofluid and the cooling water temperature controller, and the switching circuits are placed beside the test section for easy access. The temperature sensors are carefully drawn out from the test section and are connected to the data logger (Agilent) which is placed near the test section. A photograph of the same is shown in Figure 5.2.
Condensing unit Constant temperature bath Nanofluid tank Pump Rota-meter Mass flow meter Computer Differential pressure transmitter Data logger PID Controllers TEST SECTION Figure 5.2 Photographic view of the experimental test facility 65
66 5.3 TEST SECTION The test section consists of a tube-in-tube counter flow heat exchanger made up of bright annealed hardened copper tubes. The inner tube has an inner diameter of 4.3 mm and an outer diameter of 6.3 mm, while the outer tube has an inner diameter of 10.5 mm and an outer diameter of 12.7 mm. The total effective length of 2.94 m is arranged horizontally into five equal subsections as shown in Figure 5.3. Cooling water tube Nanofluid tube Figure 5.3 Photographic view of the heat transfer test section kept inside the wooden box before packing with insulation material A differential pressure transmitter and two RTDs are mounted at both ends of the test section to measure the pressure drop between the inlet and outlet, and the bulk temperatures of the nanofluid, respectively. Six RTDs are mounted at different locations on the inner and outer tube surface of the wall to measure the temperature distributions along the length of the heat exchanger test section. The locations for the temperature measurement of the nanofluid and cooling water are shown schematically in Figure 5.4 and a photographic view of the same is shown in Figure 5.5. At the beginning of every subsection, the cooling water temperature (T W ) and outer wall temperature of the inner tube (T N ) carrying the nanofluid are measured using PT100 RTD sensors. The entire length of the test section is initially wound
67 with 2 of asbestos rope insulation of thermal conductivity k = 0.1 W m -1 K -1. The insulated test section is then mounted inside a wooden box of (300 cm 30 cm 30 cm). Thermocole insulation of k = 0.5 W m -1 K -1 with 5 cm thickness is provided along the inner periphery of the box. The remaining space inside the box is packed with high quality glass wool and ceramic wool of k = 0.3 W m -1 K -1 as shown in Figure 5.6. This ensures that heat lost by the hot fluid is equal to the heat gained by the cold fluid and there is no heat interaction with the atmosphere. Both the inner and outer tubes are thoroughly cleaned for any traces of dirt in it. To ensure that the test section is leak free, high pressure nitrogen test and vacuum test are performed. T W1 (Measured) T W2 (Measured) T N Cooling water Nanofluid 0.42 m T N2 (Measured) T N1 (Measured) Figure 5.4 Locations of temperature measurement for nanofluid and cooling water Cooling water RTD Nanofluid RTD Figure 5.5 Photographic view of the temperature sensor locations
68 Glass wool insulation to the test section Figure 5.6 Photographic view of the glass wool insulated test section 5.4 NANOFLUID CIRCUIT The nanofluid tank with a 0.5 kw heater controlled with a proportional integral differential controller (PID-C) is used, in order to keep the nanofluid temperature constant. The controlled flow rate of the silver/water nanofluid at required temperature is used, to provide the desired heat load to the cooling water for a particular operating condition. The nanofluid is heated in an insulated cylindrical tank 16 cm long and 12 cm in diameter. This tank is deliberately made small (1 litre) so that the volume concentration of nanoparticles can be varied from 0.3% to 0.9% with a maximum nanofluid mass of 93 grams. The cost involved in procuring the nanoparticles is a constraint in deciding the tank size. The temperature inside the nanofluid tank is measured using a sheathed thermocouple with ± 1 o C accuracy, which gives the necessary feedback signals to the temperature controller. The heated nanofluid is tapped out from the bottom of the tank and pumped into the test section in the counter flow direction to the cooling water. The nanofluid flow rate is measured using a Coriolis type mass flow meter connected in the circuit just after the pump. In the test section the nanofluid loses heat to the cooling water and is returned to the tank at a lower temperature where it is again heated to the required temperature level. The nanofluid entry temperature is varied from 50 o C to 90 o C during the
69 experimentation. The flow rate of the nanofluid is varied by using a flow control valve, which is placed prior to the mass flow meter. All the inter connecting tubes in the nanofluid circuit are provided with necessary insulation to avoid heat loss to the surroundings. 5.5 COOLING WATER CIRCUIT A controlled flow rate of cooling water and at required temperature is used to provide the desired cooling load to the nanofluid for a particular flow condition. The water is cooled using an auxiliary cooling system which includes a 0.5 TR capacity condensing unit and a 1 kw heater coupled with a PID controller in order to keep the temperature of the cooling water constant in an insulated rectangular vessel of dimension 50 cm 50 cm 40 cm. The temperature inside the bath is measured using a sheathed PT100 RTD with ± 0.15 o C accuracy, which gives necessary signals to the temperature controller. A stirrer motor operating at 1500 rpm is mounted to ensure uniform temperature inside the cooling water bath. The cooling water is tapped out from the bottom of the tank and pumped through a glass tube rota-meter into the test section in the counter current direction to the nanofluid. The cooling water gains heat from the nanofluid and returns to the bath at a higher temperature. The flow rate of the cooling water can be varied by using a flow control valve, and the diversion valve that returns the excess cooling water back into the cooling water bath. However, in the present study, the cooling water flow rate of 16 g s -1 is kept for all the testing conditions. The maximum testing temperature of the cooling water during experimentation is maintained below 35 o C. All the interconnecting tubes in the cooling water circuit are provided with necessary insulation to avoid heat infiltration into the system.
70 5.6 MEASUREMENT DEVICES The accuracy of the heat transfer coefficient depends on the accuracy level of the measuring instruments. The complete details of the devices are listed in Appendix 2. The details of the measuring devices are explained below. 5.6.1 Temperature sensors To measure the temperature, 2 wire calibrated PT100 thin film temperature sensors of 2 m wire length are used. The sensors are fixed to the surface of the copper tube with a suitable tape, properly glued and wound with a Teflon tape to secure its location from moisture entry, if any. The level of uncertainties in the sensors according to the manufacturer is ± 0.15 o C. 5.6.1.1 Accuracy confirmation test The temperature is one of the main inputs for the estimation of the heat transfer coefficient. Hence, it is necessary to measure the global residual error of all the sensors. To achieve this, the following procedure is used: (i) The system is brought to a steady state operating condition as explained in section 5.7. (ii) The test section valves N3, N4, W3, and W4 (Figure 5.1) are closed to arrest the flow of cooling water and nanofluid across the test section. The test facility is completely switched OFF. (iii) The temperature values are observed in the data logger through a computer interface to see whether all the temperature values from the sensors are uniform (i.e. T W = T WO ) after waiting for steady state.
71 (iv) The temperatures are recorded from this point for a span of 2 hours. The mean deviation of temperature indicated by all the sensors with respect to time, is found out. After 2 hours, the test revealed that the maximum and minimum deviation of T W and T WO is ± 0.08 o C and ± 0.13 o C respectively. This confirms the accuracy of the temperature measuring system to be within ± 0.15 o C. 5.6.2 Differential pressure transmitter To measure the pressure difference across the nanofluid heat transfer test section, piezo-resistive differential pressure transmitter (0-2bar) is used. The pressure transmitter gives the differential pressure between the inlet and outlet of the test section with ±0.075 % accuracy and the output from the pressure transmitter is 4-20 ma. 5.6.3 Mass flow meter To measure the mass flow rate of the nanofluid in the circuit accurately, a Coriolis mass flow meter (3 50 g s -1 ) of ± 0.1 % accuracy is used. This meter is capable of measuring the mass flow rate and density of the liquid that passes through it. The output from the meter is 4-20 ma and the diameter of the coriolis tube is 2mm. The inlet and outlet of the meter are flared to the circuit tubing through a VCO coupling. 5.6.4 Cooling water flow meter A calibrated glass tube rota-meter of ± 2 % accuracy is used to measure the water flow rate in the cooling water circuit. The Pyrex glass tube rota-meter with an inner diameter of 12.7 mm and a stem length of 300 mm is used in the present study.
72 5.7 PROCEDURE TO START THE EXPERIMENTAL FACILITY To bring the system to a steady state condition for the experimentation, it is necessary to follow a certain procedure. This is given below: The required quantity of the silver/water nanofluid with the chosen volume concentration is prepared and fed into the nanofluid bath and is heated there to the required temperature. The auxiliary cooling system is switched ON to cool the water to the required temperature in order to let it flow in the annulus of the test section in the counter current direction to the nanofluid. The stirrer motor is switched ON to ensure a uniform temperature inside the cooling water bath. The pump in the nanofluid circuit is switched ON and the flow rate of the nanofluid is set by adjusting the flow control valve. Throughout the test section loop, all valves except the bypass valves are opened. Similarly, the pump in the cooling water circuit is switched ON and the flow rate of the cooling water is set by adjusting the flow control valve for all the testing conditions. The PID temperature controllers in the nanofluid and the cooling water circuits are switched ON. The inlet temperatures of the nanofluid and the cooling water are set by using the PID controllers as per the required conditions. The PID controllers in both the circuits are programmed to automatically regulate the heater and cooler based on the initial set conditions.
73 The test facility is made to run for a sufficient time until the steady state condition is reached. The steady state condition is manually confirmed, by checking the uniformity in the temperature indicated by all the temperature sensors in the test section, in accordance with the test conditions. Then the required parameters are logged into the data logger and saved in the computer for data processing. 5.8 EXPERIMENTATION PROCEDURE The system is brought to a steady state condition as described in the previous section. The parameters which are varied in the experimentation are the mass flow rate, and the inlet temperature of the nanofluid and cooling water. The mass flow rate and the temperature at the entry of the test section are the major design parameters, which are varied to set the desired operating condition in the setup. Thus, to achieve a certain condition, the nanofluid with a specific concentration is first filled in the nanofluid tank. The pump for the nanofluid and that for the cooling system are started. The inlet temperature of the nanofluid is varied from 50 o C to 90 o C, whereas the inlet temperature of the chilled water is varied from 25 o C to 35 o C. Similarly, the mass flow rate of the nanofluid is varied from 2 g s -1 to 23 g s -1, while the cooling water flow rate of 16 g s -1 is kept constant for all the test conditions. The heated nanofluid passing through the heat exchanger test section is cooled by the cooling water. The test section temperature distribution and pressure drop are obtained after the system reaches the steady state condition. The tests are repeated for different inlet temperatures of the nanofluid, mass flow rates and volume concentrations of the silver nanoparticles as shown in the test matrix in Table 5.1. These flow conditions are selected so that the study could cover a full laminar-turbulent regime heat transfer. For this experiment the steady
74 state condition is arrived after 1½ hours. During the experimental runs, the pressure difference of the nanofluid and the tube wall temperature at different positions along the axial direction of the test section are measured. Each experiment is repeated at least two times to get the average values. The measured temperatures and pressures are used to calculate the heat transfer coefficients. The same procedure is repeated with different mass flow rates, inlet temperatures and volumetric concentrations of the silver nanoparticles. It is noticed that a certain level of fluctuation in the parameter to be measured is usually present in the measuring instruments. In the case of the mass flow rate of the nanofluid (m N ), a fluctuation of ± 0.01 g s -1 is observed. For example, if the required m N = 8 g s -1, then the readings are logged between 7.99 g s -1 to 8.01 g s -1, which is considered as a steady state condition for m N = 8 g s -1. Similarly, the fluctuation in the differential pressure of the nanofluid is observed to be within ± 0.1 kpa. However, a fluctuation in the nanofluid and cooling water temperature are practically not seen. 5.8.1 Test matrix Based on the literature survey the variables for the experimentation are fixed and the test matrix is shown in Table 5.1. The survey has revealed that 4.3 mm tubes are often used in the small size heat exchangers in thermal industries for building heating, industrial process heating and for solar heating applications. The nanoparticles concentration is limited to be less than 1% in order to avoid excessive pressure drop due to viscosity rise. The mass flow rate chosen corresponds to the flow condition from the laminar, transition and turbulent regimes for the 4.3 mm diameter tube that pertain to the Reynolds number ranging from 800 to 12,000. The temperature of cooling water in normal conditions range from 25 o C to 35 o C and hence 25 o C, 30 o C and 35 o C are considered as the cooling water inlet temperature with a flow rate of 16 g s -1 for all conditions. The nanofluid inlet temperature is to be studied for 50 o C to 90 o C range. However, since the nanofluid returning from the test section is
75 also collected in the same tank (1 litre capacity) the nanofluid entry temperature to the test section is also indirectly influenced by the temperature of cooling water circulated. Thus the temperature of the nanofluid at steady state is noted for all combinations of mass flow rates and cooling water temperatures. The observed steady state nanofluid inlet temperatures ranged from 40 o C to 65 o C for the cooling water temperatures of 25 o C, 30 o C and 35 o C respectively. Needless to say, that the heaters in both the tanks are also regulated to achieve these conditions. Thus, the parameters in the test matrix are rationally fixed based on the application temperature and the test facility considerations. Table 5.1 Test matrix for experimentation Volume Concentration (%) T (N1) ( o C) T (W1) ( o C) m N (g s -1 ) Observations Derived Parameters 0.3 0.6 0.9 48 25 2 3 4 5 8 1.Temperature variations along the test section (T N, T W ) 2.Pressure drop across the inlet and outlet of the test section P) 1.Heat transfer coefficient 2.Nusselt number 3.Heat flux 4.Friction factor 11 14 17 20 23 57 30 do 65 35 do 5.8.2 MATLAB EXCEL interface for Data Reduction The wall temperature of the nanofluid (T N ), the cooling water temperature (T W ) measured across each subsection, the pressure difference at the inlet and exit of the test section ( P), and the mass flow rate of the working fluids (m N and m W ) are noted. The data reduction procedure to estimate the heat transfer coefficient and pressure drop is shown in Figure 5.7.
76 START Q W =mwc W (TW1 -T W2) U = Q A LMTD i W i 1 1 ln(d/d) o i 1 = + + UA h A 2kL h A i i N i W o h = 0.023 Re W Pr 0.8 0.3 k D h = N 1 1 2kAiL 1 A - - U ln(d d ) h A i o i W o i STOP Figure 5.7 Flow chart for data reduction The heat load is calculated by estimating the change in internal energy of the cooling water and balancing the energy transfer exchange between the nanofluid and the cooling water. Then the overall heat transfer coefficient is calculated by the logarithmic mean temperature difference between the nanofluid and the cooling water. As the test section is insulated, the cooling load applied to a known length of the test section is assumed to be reflected proportionally in the change in temperature of the nanofluid across the same length of the test section. The annular side heat transfer coefficient is calculated from the Dittus-Boelter correlation (1962). Hence, finally, the heat transfer coefficient of the nanofluid is calculated. For every test condition, the temperature and pressure data are logged in continuously to check for
77 attaining steady state operation. Once the test section stabilizes the data are logged in for further processing. The experiment for each condition is repeated for at least 2 times. The average results of the scans are taken as the steady state reading for a given condition. The output spread sheet from the interface is programmed to estimate and store the required results, such as the heat transfer coefficient, the Reynolds number, heat flux, the Nusselt number and the pressure drop etc. The MATLAB routine is also programmed to compare the experimental results with the well known correlations, and present the statistical information. Thus, a versatile software interface is established between the experimental database, the MATLAB and EXCEL. 5.9 UNCERTAINTY ANALYSIS The accuracy of the heat transfer coefficient depends on the accuracy level of the measuring instruments. Therefore, the uncertainty of the estimated heat transfer coefficient is to be ascertained. The error analysis for the heat transfer coefficient is carried out by applying the uncertainty analysis suggested by Moffat (1988). The heat flux (q) and its uncertainty are estimated by using equations (5.1) and (5.2). The mass flow rate and temperature of the nanofluid are the parameters considered for the uncertainty analysis. The uncertainty of the heat transfer coefficient is calculated using equations (5.3) and (5.4). For the operating conditions shown in Table 5.1, the range of uncertainty is shown in Table 5.2. A sample of the model calculation for uncertainty analysis is given in the Appendix 2. q= m C T N pn N D L (5.1) 2 2 2 2 Uq Um N U TN1 U TN2 q m T T T T N N1 N2 N1 N2 (5.2)
78 h N T N q T WI (5.3) 2 2 2 Uh U N Uq h N T h WI N h h q h T N N N WI (5.4) Table 5.2 Uncertainties of the measured and derived quantities Parameters Uncertainty Mass flow rate (m N ) ± 0.1 % Temperature (T N ) ± 0.15 C Heat flux (q N ) ± 1.18 % to ± 4.86 % Heat transfer coefficient (h N ) ± 4.44 % to ± 7.72 % 5.9.1 Repeatability test To check the repeatability of the experimental setup, tests are performed as prescribed in Ross et al (1987). Many repeatability tests are conducted to further ascertain the credibility of the test facility, and the results are compared on a day to day basis, as performed by Ross et al (1987). A comparison of the silver/water nanofluid with 0.6 % volume fraction for three mass flow rate conditions, namely, 14, 17 and 20 g s -1 is shown in Figure 5.8. The deviation in the heat transfer coefficient between day1 and day 2 lies within ± 1.5 %. This confirms that the repeatability and accuracy of the test facility are acceptable.
79 10 Day 2 HTC (kw m -2 K -1 ) 9 8 7 Day 1 +10% -10% 6 5 5 6 7 8 9 10 HTC (kw m -2 K -1 ) Figure 5.8 Repeatability test for 0.6 % volume concentration of silver/water nanofluid 5.10 LONG TERM STABILITY TEST A long term stability test is conducted to ensure that the nanoparticles are stably suspended in the base fluid for a longer period. The test is conducted for the mass flow rates of 14, 17 and 20 g s -1 with 0.6 % volume fraction of silver/water nanofluid. The tests are repeated for the same operating conditions at regular intervals. After conducting the experiment with 0.6 % volume concentration of the silver/water nanofluid, the test facility is switched OFF and kept for 15 days without any disturbance. After 15 days, the experiments are repeated for the same test conditions. Then, the experiments are repeated for the same condition after 30 days and 45 days. Further, the measurements are taken continuously one after the other without leaving any time delay between the first and the second reading for a given inlet test conditions. Finally, the results from the initially conducted
80 experiments (immediately after the preparation of the silver/water nanofluid sample) are compared with those taken continuously, after 15 days after 30 days and after 45 days. The variations observed in the heat transfer coefficient against the Reynolds number over this period are shown in Figure 5.9. HTC (kw m -2 K -1 ) 8 7.5 7 6.5 6 5.5 Initial After 15 Days After 30 Days After 45 Days Continuous1 Continuous2 5 6000 6500 7000 7500 8000 8500 Reynolds number Figure 5.9 Long term stability test for 0.6 % volume concentration of silver/water nanofluid The results show that a deviation of 4.5 % is observed in the heat transfer coefficients between the initially conducted experiments and the reading taken after 15 days. Similarly, a deviation of 5.4% is observed for readings taken after 30 days when compared with those of the initial ones and 4.7% deviation is observed for reading taken after 45 days and 1.07 % deviation is observed in the continuously taken readings. The deviation observed is less than 6%, which is normally accepted due to experimental errors. Thus, from the obtained results it is clearly seen that the nanoparticles
81 are stably suspended in the base fluid over a longer period of time and hence the heat transfer characteristics are also stable. Based on the uncertainty analysis, repeatability and long term stability tests, it is confirmed that the experimental facility has been fabricated as per the standards and the procedure adopted for conducting the experiments is also acceptable. Therefore, the experimental observation results from this facility are considered for further data reduction to evolve the heat transfer coefficient correlation, which could be used as a tool for designing heat exchangers. A detailed discussion of the obtained results from the experiments, comparisons of the measured data with the published results and the mechanisms involved in the enhancement of heat transfer are presented in chapter 7.