An experimental analysis of an oscillating heat pipe consisting of three interconnected columns at various operating pressures

Heat pipes are devices that working with the principle of heat convection and phase change of the refrigerant fluid. Because of the simplicity of construction, having no moving part and high capacity of heat transfer, these devices are preferred in several heat exchanger systems. Oscillating heat pipes (OHP) are compact and efficient heat pipes that benefit from the oscillating flow. OHPs were invented in late 20th century and they are especially used in electronics. In this study, an alternative model of oscillating heat pipe is investigated. Design fundamentals of oscillating heat pipes are depend on various parameters. Refrigerant fluid, fluid flow in heat pipe (diameter of heat pipe), total number of meandering, orientation of heat pipe, filling ratio of heat pipe and total heat flux to condensation zone etc. effects thermal efficiency. Due to the complex characteristics of heat transfer and dynamic oscillating behavior inside oscillating heat pipes, experimental studies are needed for determining overall heat transfer coefficient of OHPs. In this study, experimental setup was installed first to determine the overall heat transfer coefficient between condensation and evaporation zones. By changing the working pressures and heat flux, thermal and mechanical behavior of OHP is investigated both experimentally and theoretically. A mathematical model was applied cases in this study. The mathematical model was developed before by previous graduate thesis supervised by Dr. Özdemir. Minor changes was made to upgrade the mathematical model. The experimental setup used in this study, is upgraded version of the setup in ITU Faculty of Mechanical Engineering, Research Laboratory of Heat and Mass Transfer. Experimental setup has similarities with closed end loop heat pipes. It has additional third liquid column which lead oscillating flow in this study. Setup has three interconnected columns called evaporator column, condenser column and compression column. Columns are made of heat-resistant glass of 35.2 mm in inner diameter. The glass columns are interconnected by brass tubes of 35.8 mm inner diameter. Heat input provided by DC power. A DC power supply is used to supply the electrical heater inside brass tube in evaporator column. Condensation is provided by the cooler tube, which is made of two concentric copper tube be supplied by constant temperature bath. Cooling water enters the inner tube, and leaves from the outer tube. Experiment installation, the liquid column makes it different from other third loop heat pipe. Compression column which differs the experimental setup from loop heat pipes, connected to a compression tank. Pressure of the compression tank can be adjusted by vacuum pump. Experimental setup also consists 21 thermocouples for measuring temperatures on surfaces and cross sections. Furthermore, there is a piezo resistive pressure transmitter between condensation and evaporation columns in order to measure instantaneous pressure on vapor zone. Also there is a digital barometer to control pressure on compression tank. Phase transitions should be continuous on conventional heat pipes. So that their operating range is not flexible. If it exceeds operating range, all of working fluid turns into steam or fluid which makes heat pipe unable to work. In this study, due to the fact that tank pressure can be adjusted, it widen the operating range and differs this study from all other studies in the literature. Main goal of this thesis is investigating performance of oscillating heat pipe under different working both experimentally and theoretically. Furthermore, building a Nusselt correlation also depends on working pressure of OHP. Because of phase lag between the evaporation and condensation processes, liquid in the columns starts to oscillate. Oscillation causes heat transfer by convection. Amplitude of liquid level oscillation is in the order of 10 cm. In the scope of this study, 18 experiments were made. Experiments were made under six different heat load and tank pressure. Three of tank pressure are below atmospheric pressure and rest of them are above atmospheric pressure. Three different heat load applied each pressure conditions during experiments. Filling ratio of the system was kept constant. However, as lack of DC supply some of liquid discharged in high pressured experiments. All temperature and pressure data was collected by Keithley 2700 multimeter data acquisition system and saved by computer. A video camera record oscillating motions to measure instantaneous level of liquid columns. Videos were shot 33 frame per second. Liquid levels were determined for every frame with the help of a video tracker program. Oscillation curves of liquid columns, pressure curves of vapor zone, overall temperature of heater and cooler units, probe temperatures of liquid and vapor zones and overall heat transfer coefficients are determined for all experiments. Experiment #1 is chosen for sample experiment and all calculations are shown detailed in chapters. Rest of the results are represented in tables and charts. Fitted curves for liquid motion is determined with Fast Fourier Transform (FFT) analysis. The result of the FFT analysis, there is one frequency for oscillation motion. Within 18 experiments, frequencies are varied between 4,78 and 4,28 (rad/s). Largest frequency is obtained from highest heater power load (230 W) and smallest frequency is obtained from highest tank pressure (130 kPa). Fitted curves for pressure oscillations is also obtained by FFT analysis. Unlike the liquid columns, oscillation of the vapor pressure has two frequencies. Primary frequencies are obtained close to liquid oscillation frequency and secondary frequencies are obtained in the order of half value of primary frequencies. Variation of overall heat transfer with frequencies, tank pressure, overall temperature of heater and cooler units, probe temperatures investigated. Due to the fact that, all those parameters couldn't have given enough information for overall heat transfer coefficient, dimensional analysis was applied. 11 dimensionless number obtained. By algebraic manipulations with these numbers, number of dimensionless parameters reduced to 6 which are, Jakob, Weber, Bond, Prandtl, Equivalent Grashof and Kinetic Reynoldss numbers. These 6 number are considered important for the case because first 5 numbers are related with boiling and Kinetic Reynoldss number is related with oscillating motion. Variation of 5 dimensionless parameters with Nusselt number is investigated and it was seemed that Nusselt number can be written as a function of the other dimensionless numbers. By multi-parameterized regression analysis, Nusselt correlation was found. Validation of Nusselt correlation with experimental data is found highly successful. Finally. The mathematical model was applied all cases for this study and results are compared with experimental data. Coherent results were obtained from mathematical model.