undamaged surfaces (5,6) . Thus, the flow conditions in the damaged region must also be taken into account.Based on the background provided, the specific objectives of this study are: (a) to describe the experimental facility, and the experiments conducted to validate the FE thermomechanical models, and (b) to determine the heat flux on the damaged tile by considering the hypersonic flow past a cavity, and incorporate the refined thermal loading into the FE thermomechanical analysis in order to determine the effects of damage on TPS response. The numerical results obtained with the combined fluid and structural model, represent in more accurate manner the actual behaviour of the damaged TPS, and illustrate the importance of testing the TPS in a facility that can replicate the actual flow and its interactions with the cavity in the damaged region.
DESCRIPTION OF THE TEST FACILITY AND EXPERIMENTSThe experiments were conducted in the Thermal Structure Testing Laboratory at the Department of Aerospace Engineering of the University of Michigan. Specimens used for the experiments consisted of undamaged and damaged configurations with three different sizes of damage, D = 25·4mm, 38·1mm and 46·6mm. The damage is illustrasted schematically in Fig. 2, it is idealised as a cylindrical hole ending in a spherical cap. The total depth of the dmaged region D is equal to the diametrer of the spherical cap. Due to the limited number of specimens available, only two tests were conducted for each configuration.
Test facilityThe test facility resembles a somewhat similar facility at NASA Langley (7) . The precise re-entry conditions for testing TPS are difficult to duplicate. An alternative is to apply a transient temperature boundary condition that may exist during re-entry on the surface of the TPS test specimen, and simulate re-entry static pressure in the vacuum chamber. A major limitation of such a facility is the inability to account for the important interactions that exist between high speed flow and vehicle. An overview of the facility is shown in Fig. 3. It consists of the vacuum chamber, pressure control system, radiant heater system, and data acquisition system.The cylindrical steel vacuum chamber has a diameter of 851mm and a length of 940mm. The chamber is equipped with feed-through for power, gas, and instrumentation for 20 pairs of type K thermocouples and 12 pairs of strain gages. A vacuum of 5·33Pa can be achieved when equipped with a 0·0113m 3 /s dual-stage rotary vacuum pump. Nitrogen is bled into the chamber at a controlled rate to simulate re-entry pressures. This is accomplished by a control system which includes a control module and a servo-controlled leak