The development of engine health monitoring for gas turbine engine health and life management

Abstract

The development of any technology has been driven historically by a need to resolve a dilemma. The goal of that technology has sometimes never been clear or achievable because the described problem and need have not been fully understood. This paper will detail the approach that should be taken when trying to resolve an assumed problem through the development of technology. The paper will focus on describing the logical approach and direction that must be taken on the development of a gas turbine engine health monitoring (EHM) system for health and life management. The philosophical, methodological and technological approaches will be described, as well as the essential development of transferring data to information to knowledge to wisdom. This will result in a robust blueprint to achieve an EHM system that enables the 3Rs of health and life management i.e. The Right Action at the Right Time for the Right Reasons. The development of a sensing, learning and reasoning EHM system is the corner stone or more importantly the missing link between our current diagnostic confused state, and a status of a true diagnostic with prognostic (knowledge with wisdom) capability. The paper summarizes the current USAF/DoD advanced EHM activities that are being performed to achieve the 3Rs. It will attempt to provide clear examples of how data integrity, compression and fusion will feature as the first link towards achieving a truly cognitive ontogenetic (CO) (sensing and learning) EHM system. Additionally developing and applying a reasoning science like that provided by a probabilistic diagnostic and prognostic system (ProDAPS). Thus we will achieve an assemblage that can be described as truly artificially intelligent (AI) in structure, judgment and accomplishment, or more importantly capable of providing the guidance to apply 3Rs. MAINTENANCE FUNDAMENTALS The solution to any paradigm requires the understanding of the road that has already been taken before attempting to venture onwards. The development of EHM has come from 5 maintenance principles: Reactive Maintenance. Reactive Maintenance is only performed after parts have failed or an operational performance limit is not achievable. Known as unscheduled maintenance, also called fly to failure. Reactive maintenance is suitable for fail-safe designed components. Many parts are only used once and are discarded upon disassembly. Improper use of reactive maintenance can lead to flight safety, performance and reliability issues. Preventive___Maintenance. Preventive Maintenance is performed to a fixed maintenance schedule commonly called scheduled maintenance. Preventive maintenance is suitable for safe life designed components. Usually involves routine replacement of parts, fluids, and coatings. Critical lifed engine parts are removed based on operating hours or cycles, which dictates the engines removal and overhaul schedule. Safety and performance margins are restored with replacement of new or refurbished parts. Useful remaining life is based on the fleets statistical usage and not on an individual engines operational exposure/experience. Improper use of preventive maintenance can lead to replacement of parts with significant remaining useful life, reduction in operational capability or parts failing before a scheduled maintenance activity. Predictive Maintenance. _________________ Predictive Maintenance is a type of condition-based maintenance relying on performance or structural failure prediction using performance monitoring and/or nondestructive inspection (NDI) techniques. Engines remain installed until sufficient life i.e. heat cycles are accumulated. This determines the engine's overhaul schedule. Engines can be removed based on detected performance and structural deterioration trends from monitoring systems including onboard/offboard sensors, borescope, JOAP, and other NDI techniques. Detailed NDI inspections often accompany preventive overhauls and qualify as predictive maintenance for those parts. Overemphasis on predictive maintenance can lead to premature engine and parts removals and poor second-guessing of an event. Condition based maintenance can extend parts usage beyond fleet averages based on actual operational usage. Proactive Maintenance. _________________ Proactive Maintenance is another type of conditionbased maintenance emphasizing the routine detection and correction of root cause conditions that lead to performance changes and/or component failure. Conditions are corrected or parts are redesigned based on a root cause failure analysis. Common root causes are normally described to engine vibration, contamination, material specs, design, manufacturing, assembly or operational problems. Root causes such as vibration and contamination can be monitored automatically. Proactive Maintenance as root cause analysis can convert potential shop visits or flight safety incidents into continued safe use or the replacement of a less expensive part. Prognostic Maintenance. Prognostic Maintenance is the application of true oncondition maintenance. The use of prognostics within a health management system is known as Prognostic & Health Management (PHM). The application of PHM within an Autonomies Logistic System is the ultimate objective. It can be seen as a logical spin-off from failure prediction maintenance allowing the user to safely identify the remaining life of deteriorating components whilst in use, and to capitalize on that knowledge for operational and/or cost reasons. Engine operating data is used to produce information, which provides knowledge. The timely and accurate feedback within a PHM system provides the wisdom to perform the right action at the right time for the right reason; know as the 3Rs. There are no hard limits to forecast the remaining life of the component, as all components are individual event driven. The system provides a clear indication of when maintenance is due and automatically schedules the parts list, tooling and labor required to restore any event that is deteriorating towards a health, performance or safety limit. Prognostic maintenance is a short; medium and a long range mission planning tool which will significantly reduce operational costs, whilst enhancing operational capabilities at the same or improved safety margins. TECHNOLOGY STATUS QUO One can come to the conclusion that the aerospace industry has failed, in the past, to directly address the needs of it's military of civil customer for accurate maintenance information. This is not to say that all is doom, but we need to raise the awareness to 'true' EHM need as the answer to correct, accurate and prompt (CAP) maintenance. The genuine application of advanced EHM is the achievement of an artificial intelligent (AI) capability. The term AI is redundant because what we need is neither artificial nor intelligent. It can however be constructed around cognitive, ontogenetic and reasoning sciences. That is the connectivity of sensing, learning and deciding (fig 1). This can be structure as a Awareness, Learning and Reasoning System. The USAF has developed the first 2 parts and is now addressing the reasoning element. If we get this correct we provide the 3Rs and deliver CAP maintenance. Moreover we have the most important thing of all, which is event management. That is to say that an event has happened, is happening or is going to happen. Future aviation programs need CAP event management as a fundamental core requirement. To this end it is a reasoning