Steady-State, Regional Average Heat Transfer Measurement Technique
This technique is one of the most widely employed measurement methods, for convective heat transfer, used by researchers. The main advantage of this technique is that it is a reliable method which can give repeatable results. The relative uncertainty of this method, if correctly applied, can be controlled to be around 5%. In fact, data obtained by using this technique has been used by several researchers as a basis for comparing and calibrating more advanced techniques such as liquid crystal thermography, mass transfer techniques, and infrared thermography.
The primary disadvantage of this measurement technique is its inability to obtain a true local distribution of the heat transfer coefficient on the test surface. This disadvantage can be partly overcome by resolving the test surface into several smaller regions and using a copper plate for each region. Either a single heater, supplying a uniform heat flux over the heated area, or individual heaters for each region can be used to heat the test surface. By using individual heaters, the copper plate temperature, or the wall temperature, can be maintained equal for each region resulting in a uniform wall temperature boundary condition. A disadvantage of using many heaters is that variable transformers equal to the number of regions are needed to control the power input to each heater. This can be operationally inconvenient for the experimentalist, as well as expensive. It should be noted that when using a copper plate grid as a test surface, each region should be isolated from the adjacent regions using an insulating material such as rubber or balsa wood to prevent conduction losses.
Multiple Copper Plates with Heaters and Thermocouples
This method involves using copper plates with embedded thermocouples as the test wall, with one side of the plate in contact with a heater while the other side is exposed to the convective fluid. The reason copper is used for the test surface is due to its very high thermal conductivity (~400 W/mK) which results in low Biot number. If the Biot number for the copper plate is less than 0.1, the wall temperature can be assumed to be uniform in the plate volume due to high conduction within the copper plate. Thus, accurate, average heat transfer coefficients can be measured with this method. Aluminum can be used for the surface if the Biot numbers are less than 0.1.
Ideally, several thermocouples should be embedded in each copper plate so that an accurate average wall temperature can be obtained. If the Biot number is less than 0.1, all thermocouples should read the same temperature. Tiny holes can be drilled in the copper plate symmetrically for the thermocouples. An adhesive, such as a thermally conductive epoxy, can be used to attach the thermocouples to the copper plate. It should be noted that the holes should be drilled in such a manner as to leave the heat transfer surface smooth. If the thermocouples are to be inserted from the heater side, grooves may be cut in the copper plate to allow passage for the thermocouple wires. This is much more convenient than cutting the heater to allow the wires to pass through it. A data acquisition system can monitor the temperatures of the copper plates as measured by the thermocouples.
Power supplied to the heater raises the temperature of the copper plate and is generally controlled through a variable transformer which can regulate the applied voltage. A resistance wire, flexible heater or a foil heater can be used. If the test surface is circular, an axially placed rod heater can be used to heat the wall. If one side of the heater is exposed, it should be covered with a suitable insulation material (e.g., fiberglass, wood, etc.) to prevent any extraneous heat losses. The minimum insulation thickness needed can be approximated by applying heat conduction principles between the heated wall and the ambient surroundings.
The heater power input can be predetermined before the experiment by estimating the heat transfer coefficients on the surface using a suitable empirical correlation. For example, the Dittus-Boelter correlation can be used for turbulent flow in a duct to determine the heat transfer coefficients for a specified Reynolds number. If tabulators are to be studied, a suitable enhancement factor can also be employed on the heat transfer coefficient. By using this magnitude for the heat transfer coefficient, the power input can be estimated for the given test surface area and temperature difference between the wall and fluid.
The heated wall exposed to the flowing fluid will attain a steady-state (unchanging wall temperature for a specified heat flux) after some time has elapsed. The wall temperature should be maintained approximately 20°C higher than the fluid temperature. By measuring the power input to the heater and the related fluid and wall temperatures, the heat transfer coefficients can be calculated. The heater power input should be corrected for any extraneous heat losses in the system.