【正文】 creases heat transfer enhancement because of spallation. They also reported that cavities with longer axial length produced higher heat transfer enhancement on the flat surface. The present study simulates the effect of TBC spallation on a cylindrical leadingedge model. The spallation sizes were based on scaling from real engine spalls and TBC thicknesses from landbased gasturbine engines. Heat transfer tests were conducted at three freestream turbulence levels for four spallation locations on the leading edge. Detailed heat transfer enhancement distributions are presented within and around the spallation. A transient liquidcrystal technique was used to obtain the detailed heat transfer distributions on the leadingedge surface. Two different spallation depths were studied at each spallation location to understand the effect of spallation depth. The highresolution results help identify local high and lowheat transfer regions caused by flow reattachment and separation, respectively.Test Apparatus and InstrumentationA schematic of the wind tunnel is shown in Fig. 1. The suctiontype wind tunnel was designed to avoid uncontrolled turbulence at the inlet of the test section. A flow straightener is followed by a contraction inlet nozzle. The nozzle has a contraction ratio of 3:1. The test tunnel is cm X cm in cross section and 183 cm long with the cylindrical model placed cm downstream of the nozzle exit. A tailboard was placed at the rear of the cylinder to reduce the downstream wake effects on upstream heat transfer. A grid is placed at the nozzle exit for the generation of higher freestream turbulence. The imageprocessing system for analyzing the liquidcrystal color changes consists of a personal puter with a color framegrabber board, a red—greenblue (RGB) camera, a color monitor, and customized software. The color image processing system is described in detail by Ekkad et Two turbulence grids were fabricated to generate different levels of freestream turbulence intensity. Grid 1 is made of stainlesssteel bars, is cm square in cross section, and is cm in pitch. Grid 2 is made of hollow brass tubes, is cm square in cross section, and is cm in pitch. The width and height of the grids are the same as those of the test tunnel. Both grids have an open area of 54%. The turbulence grids generate turbulence intensity levels close to actual engine conditions. Hotwire probes were inserted through slots to measure the oning flow velocity and turbulent fluctuations. A calibrated singlewire probe was used to measure mainstream flow behavior. It was connected to a fourchannel TSI IFA 100 hotwire anemometer. The analog signal was digitized using a 250kHz A/D board in a personal puter.Figure 2 shows the cylindrical test model. The cylinder, cm in diameter and cm long, is hollow with six cartridge heaters embedded along the circumference to heat the outer surface uniformly. The cartridge heaters, cm long and cm in diameter, conduct to a copper cylinder that is in contact with the black polycarbonate exterior surface. There was no airspace between the polycarbonate and copper layers. The polycarbonate layer is cm thick and has low thermal conductivity and diffusivity. The front half of the polycarbonate exterior can be changed for different spallation locations. Spallations were machined on each piece at a different location. A spallation occupies about a 20 deg width on the cylinder surface. Spallations were placed at 020 deg (SI), 10 30 deg (S2), 2040 deg (S3), and 3555 deg (S4). All of the Spallations are cm wide (W) and cm long (L). Two different spallation depths of and cm were tested for each spallation location. Figure 2 also illustrates a typical spallation location on the cylinder test model. The real TBC spallation is of irregular shape and it is difficult to estimate the real effect of the spallation under actual engine conditions. The real TBC spallation is simulated as a rectangular slot with rounded edges in the present study. Also, the real engine blade TBC thickness that spalls is in the range between 5 and 20 jjim. The leadingedge model in this study is about 510 larger scale than the actual engine blade, whereas the mainstream velocity is much lower to simulate flow Reynolds numbers around the leadingedge region of the turbine blade. Also, there is a need to also scale the spallation depth, size, and shape to simulate the actual However, the present study only pares the depth and location effects for the same spallation size and shape. The measurement region is also indicated in Fig. 2. The measured region is limited to one side of the front half of the cylinder from stagnation (0 deg) to about 70 deg from stagnation.Theory and ProcedureThe heat transfer measurement is obtained by the transient liquidcrystal technique. The test surface is made of polycarbonate,which is a good thermal insulator. The test surface is heated to a uniform temperature using the cartridge heaters. When a uniform surface temperature is achieved, the cartridge heaters are shut off and the transient test is initiated by blowing the mainstream air over the test surface. The transient liquidcrystal technique1339。53附錄：外文翻譯資料Heat Transfer Distributions on a Cylinder with Simulated Thermal Barrier Coating SpallationSrinath and JeChin of thermophysics and heat ,13(1):7681.IntroductionHigher inlet temperatures in modern gasturbine engines provide higher power output and higher thermal efficiency. Hightemperature gases have a damaging effect on hot gas path ponents in the engine. Thermal barrier coatings (TBC) are used to protect engine ponent metal surfaces from these hightemperature gases. The TBC coating is a heatresistant ceramic layer sprayed onto the hot gas side of the turbine airfoil. The TBC application process is either done by air plasma spraying (APS) or the plasma vapor deposition (PVD) technique. The TBCs cons