Foam on Tile Impact Modeling for the STS-107 Investigation

Following the breakup of the Space Shuttle Columbia during reentry a NASA/Contractor investigation team was formed to examine the probable damage inflicted on Orbiter Thermal Protection System elements by impact of External Tank insulating foam projectiles. The authors formed a working subgroup within the larger team to apply the Smooth Particle Hydrodynamics code SPHC to the damage estimation problem. Numerical models of the Orbiter's tiles and of the Tank's foam were constructed and used as inputs into the code. Material properties needed to properly model the tiles and foam were obtained from other working subgroups who performed tests on these items for this purpose. Two- and three-dimensional models of the tiles were constructed, including the glass outer layer, the main body of LI-900 insulation, the densified lower layer of LI-900, the Nomex felt mounting layer, and the Aluminum 2024 vehicle skin. A model for the BX-250 foam including porous compression, elastic rebound, and surface erosion was developed. Code results for the tile damage and foam behavior were extensively validated through comparison with Southwest Research Institute foam-on-tile impact experiments carried out in 1999. These tests involved small projectiles striking individual tiles and small tile arrays. Following code and model validation we simulated impacts of larger foam projectiles on the examples of tile systems used on the Orbiter. Results for impacts on the main landing gear door are presented in this paper, including effects of impacts at several angles, and of rapidly rotating projectiles. General results suggest that foam impacts on tiles at about 500 mph could cause appreciable damage if the impact angle is greater than about 20 degrees. Some variations of the foam properties, such as increased brittleness or increased density could increase damage in some cases. Rotation up to 17 rps failed to increase the damage for the two cases considered. This does not rule out other cases in which the rotational energy might lead to an increase in tile damage, but suggests that in most cases rotation will not be an important factor.