NASA NTRS2004
The technical foundation for the most probable damage scenario reported in the Columbia Accident Investigation Board's final report was largely derived from synergistic aerodynamic/aerothermodynamic wind tunnel measurements and inviscid predictions made at NASA Langley Research Center and later corroborated with engineering analysis, high fidelity numerical viscous simulations, and foam impact testing near the close of the investigation. This report provides an overview of the hypersonic aerothermodynamic wind tunnel program conducted at NASA Langley and illustrates how the ground-based heating measurements provided early insight that guided the direction and utilization of agency resources in support of the investigation. Global surface heat transfer mappings, surface streamline patterns, and shock shapes were measured on 0.0075 scale models of the Orbiter configuration with and without postulated damage to the thermal protection system. Test parametrics include angle of attack from 38 to 42 degs, sideslip angles of 38 to 42 degs, sideslip angles of plus or minus 1 deg, Reynolds numbers based upon model length from 0.05 x 10(exp 6) to 6.5 x 10(exp 6), and normal shock density ratios of 5 (Mach 6 Air) and 12 (Mach 6 CF4). The primary objective of the testing was to provide surface heating characteristics on scaled Orbiter models with outer mold line perturbations to simulate various forms of localized surface damage to the thermal protection system. Initial experimental testing conducted within two weeks of the accident simulated a broad spectrum of thermal protection system damage to the Orbiter windward surface and was used to refute several hypothesized forms of thermal protection system damage, which included gouges in the windward thermal protection system tiles, breaches through the wing new the main landing gear door, and protuberances along the wing leading edge that produced asymmetric boundary layer transition. As the forensic phase of the investigation developed and the condition of recovered debris was examined, increasing emphasis was placed on identifying wing leading edge damage (partially and fully missing reinforced carbon-carbon panels, and eventually holes in the wing leading edge with venting to the wing upper surface) that produced off-nominal heating trends consistent with extracted Orbiter flight recorder temperature data.