Demonstration of advanced life-prediction and state-awareness technologies necessary for prognosis of turbine engine disks Stephan M. Russ a , Andrew H. Rosenberger a , James M. Larsen a , Robert B. Berkley b , David Carroll b , Bradford A. Cowles b , Richard A. Holmes b , Jerrol W. Littles Jr. b , Richard G. Pettit b , and John J. Schirra b a Air Force Research Laboratory, 2230 Tenth St, Wright-Patterson AFB, OH, USA 45433; b Pratt & Whitney, 400 Main St, East Hartford, CT, USA 06108 ABSTRACT This paper summarizes a spin test of an IN100 minidisk that demonstrated advanced fatigue crack growth predictive tools under dwell fatigue and the ability to infer the damage state from state-awareness sensed data. The test was performed at elevated temperature in a partial vacuum and was a major success. The advantages of employing 3D fracture mechanics tools were clearly demonstrated when the prediction using the advanced analysis technique produced a crack growth lifetime 10 times greater than the standard 2D tools and well within a factor of two of the test result. However, when the effect of the partial vacuum was also taken into account, the predictions more closely resembled the test results. Another success was demonstrated when the “blade-tip” time-of-arrival sensors detected deflections of 50 microns and greater at temperature. Using the 3D analysis tools, a transfer function was created that related blade-tip deflection to crack size. The crack size was tracked in near real time for the final 150 cycles. This test represented significant demonstrations of both 3D fracture mechanics tools and state-awareness sensing. Both are advanced technologies that will eventually impact the life management of turbine engines. Keywords: prognosis, life prediction, state-awareness sensing, time-of-arrival, 3D fracture mechanics, dwell fatigue, crack growth, transfer function 1. INTRODUCTION Diagnostic tools by design are used to interrogate damage or detect a structure or system in the process of failing. The distinguishing difference between diagnosis and prognosis is that prognosis implies the prediction of a future state. Thus, to accomplish prognosis requires both diagnostic and predictive tools, the former to sense the current state of damage and the latter to predict the future state based on projected usage and applicable life-prediction routines. Prognosis systems are, therefore, most readily applied to structures and systems that undergo gradual wear-out modes; where the failure process can be detected at an early stage, monitored, and accurately predicted such that the part or system can be serviced at a later stage of damage prior to catastrophic failure. Structures and/or structural components that are subject to fatigue loading lend themselves well to this philosophy, where a crack will initiate and grow over some period of time prior to fracture. Many turbine engine components are designed and life managed based on low cycle fatigue behavior. 1 USAF engines are required under the engine structural integrity program (ENSIP) to meet both crack initiation (safe-life) and fatigue crack growth (damage tolerant) design criteria. 2 Currently the USAF mandates fluorescent penetrant and eddy current inspections of many fracture-critical components, such as turbine engine disks, to interrogate for relatively small fatigue cracks. These nondestructive inspections are performed to ensure that a crack does not exist which could lead to fracture of a disk in the subsequent service interval. Unfortunately, these NDE inspections are both time-consuming and costly, because they drive engines off wing for complete disassembly in order to allow access to the components. Thus, turbine engine disks provide motivation for developing diagnostic and prognostic tools applicable for turbine engines. A new initiative by the U.S. Defense Advanced Research Projects Agency (DARPA) is developing and demonstrating technologies for materials damage prognosis. 3 The vision is to enable real-time management of materials and Health Monitoring and Smart Nondestructive Evaluation of Structural and Biological Systems III, edited by Tribikram Kundu, Proc. of SPIE Vol. 5394 (SPIE, Bellingham, WA, 2004) · 0277-786X/04/$15 · doi: 10.1117/12.540404 23