Status of Engine Rotor Burst Protection Program for Aircraft

CCOMPANYING today's use of turbomachiner y in aircraft is the associated danger that occasionally a highspeed rotor will fail and the fragments will damage equipment and threaten passenger safety. Such adverse effects can be reduced either by prevention of or protection against rotor bursts. A significant effort is being expended to improve methods for detecting defects during fabrication, operation, and maintenance. Despite such efforts, however, statistics continue to reflect the persistence of rotor failures. During the past 9 years, 170 uncontained failures of gas-turbine engine components were reported by U.S. commercial airlines. Accordingly, NASA has been sponsoring a research program with a view toward providing protection to the aircraft without imposing large weight penalties. This program is directed toward the development of criteria and methods for the design of practical devices to contain rotor fragments as well as to deflect them away from the critical aircraft components. The immediate aim is to develop meaningful tests and analytical methods to evaluate possible solutions. In order to generalize the results, experiments are closely coordinated with analytical methods. Tests, conducted in the Navy Containment Evaluation Facility,1"4 are compared with results of calculation methods developed by the Massachusetts Institute of Technology.5'6 Test rotors, modified to fail with predetermined sizes and shapes of fragments, are caused to impact containment/deflection devices in a spin chamber. The complex interaction between the fragments and the protective device is recorded by high-speed photography. Deformations of the containment/deflection device, measured on the photographs and related to the time reference, show the dynamic interaction. This paper highlights some of the accomplishments to date and presents future plans. Contents Simple tests were run first to support analysis. A blade was modified to fail at a specified speed and impact an aluminum alloy (2024-T4) freely supported containment ring (i.e., no restraints to motion in the plane of the ring). The response is show in Fig. 1 at the three elapsed times. Ring deflections around the circumference (at the alternate black and white marks) were recorded at approximately 30-jnsec intervals. Comparison of the measured and analytically predicted ring profiles after impact shows good agreement for this simple case. However, improvements in the calculation methods are being pursued to increase the accuracy of the predictions and to handle the interaction between more complex fragments and containment rings. The method of