Document
A Roadmap for Aircraft Engine Life Extending Control Ten-Huei Gut NASA Glenn Research Center Cleveland, Ohio 44135 Abstract system, which consisted of a linear performance controller in the inner loop and a nonlinear damage The concept of Aircraft Engine Life Extending controller in the outer loop. The simulated LEC was able Control is introduced. A brief description of the tradeoffs to reduce the transient damage in the turbine blades of one between performance and engine life are first explained.
of the high-pressure turbo pumps by a factor of 35 while The overall goal of the life extending controller is to keeping the transient performance degradation very small reduce the engine operating cost by extending the on- [3,4,5]. The main elements of Life Extending Control wing engine life while improving operational safety. The application to air-breathing propulsion system are research results for NASA's Rocket Engine life extending illustrated in Figure 1 and discussed in the following control program are also briefly described. Major subsections.
building blocks of the Engine Life Extending Control architecture are examined. These blocks include: life Life Usage Model prediction models, engine operation models, stress and The calculation of the life of the engine components is thermal analysis tools, control schemes, and intelligent the centerpiece of Life Extending Control. Generally control systems. The technology areas that would likely speaking, current engine component life is calculated impact the successful implementation of an aircraft during the engine design phase. During the engine engine life extending control are also briefly described.
design, each engine part has to be analyzed using the Near, intermediate, and long term goals of NASA's maximum rated engine operating conditions. Although activities are also presented.
there are many parts with many different failure modes to be considered, the most important life-limited parts are 1. Introduction the rotating components in the hot section directly after Current aircraft engine controllers are designed and the burner. The typical failure modes for the hot section operated to provide both performance and stability engine components are low cycle fatigue (LCF), thermo- margins. NASA Glenn Research Center and its industrial mechanical fatigue (TMF), and creep/rupture. Most of and academic partners have been working together toward the lifing models are based on fracture mechanics and a new control concept that will include the consideration fatigue crack propagation, which take into account the of engine life usage to minimize overall operating costs.
material properties and the thermal stresses of the engine The new controller design will utilize damage models to components [3]. However, for safety reasons, the monitor the damage rate and damage accumulation of certified useful life of an engine component is set to be critical parts. The tradeoffs between performance and well before a crack would usually occur. This overly structural durability may be assessed for different levels conservative life limit recommendation practice has of mission requirements and engine life states. The goal resulted in engine parts being removed without evidence of the proposed controller is to reduce engine operating of degradation, wasting remaining usable life.
costs while improving operational safety.
More accurate lifing calculations of each component This paper will then present a roadmap for the life are needed to enable the implementation of LEC.
extending control program for aircraft engines. The Actually, better lifing calculations and life tracking will building blocks required for a successful program include also improve the operational safety of the engine. It is a high-fidelity engine operating model, temperature/stress necessary to have the engine life calculated during engine models for life-limited components, life models for life- operation and updated for every major cycle. An ideal limited components, mission profiles, sensors, real-time lifing model for LEC will be a package including the life tracking, advanced actuation, and optimum control for lifing calculations of all critical failure modes of all life- the integrated system. Details of each block will be limited components. It will be calculated in real-time discussed including technology readiness, implementation during engine operation and after each thermal stress issues, and optimum control integration.
cycle according to the operating conditions of the engine.
The capability of tracking component life usage is also 2. Elements of Engine Life Extending Control important because used parts with substantial life (time) The Life Extending Control (LEC) concept was first left are often re-used during the engine overhaul process.
introduced by Lorenzo et al. in the early 90s [1,2]. It was And, because the component life (usage) cannot be first applied to a simulation of the Space Shuttle Main measured directly, it is also very important to have the Engine (SSME). The open loop and closed loop lifing model verified experimentally (by factory applications of LEC were successfully demonstrated. A two-tier controller architecture was proposed for the LEC This is a preprint or reprint of a paper Intended for presentation at a conference. Because changes may be made before formal publication, this is made available with the understanding that it will not be cited or reproduced without the permission of the author.
E endurance testing) or empirically (through the study of Stress and Thermal Analysis field engine maintenance data).
Stress and thermal analysis is the tool bridging the gap between measured operating variables and life models.
Usually, engine sensors and the on-board model are able Performance Variables to provide values for rotor speeds, temperature, and pressure at select points. However, the life calculation
V
External i- . t .. I requires much more detailed information including metal I l-'enonnance Dynamics temperature, temperature gradient, mechanical load, and Engine C°mm_----_ I Control stress at each critical location on the part. Engine Commands/_k companies usually use tools such as finite-element Schedule | Measurements analysis (FEA) to calculate the thermal and stress conditions. Together with other variables such as shaft speeds, material properties and component geometries, Damage further analysis is also done to determine the loads and ,Loads, stresses of all possible failure locations at various Rate LifeUsage [ TM Model I Model l" _ Stress/Thermal 14 operating conditions. This information is then used to predict the life of engine components.
In Life Extending Control, it is unrealistic to use finite Figure 1. Life Extending Control Blocks element analysis for the stress and thermal analysis for an on-line application. This part of the calculation can be combined with the life model so that life usage can be Engine Model calculated directly from the engine parameters. It may In order to calculate the life usage of an engine also require a lot of model simplification and linearization component, it is necessary to obtain the operating in this process to increase the execution speed while conditions of the component. This can be accomplished preserving the fidelity of the life model.
by either direct sensor measurements or by an on-board Possible Control Schemes engine model simulation. Since the engine sensors are There are several control schemes that have the designed for control and safety monitoring purposes, they can only provide very limited information for engine life potential of increasing the life of an aircraft engine. Each calculations. During the design phase of an engine, an control scheme requires extensive study of its potential engine cycle model ("deck") can be used to determine the effect, sensor and actuator requirements, and engine component conditions during different flight implementation integration with the current control scenarios. The cycle deck models the engine with algorithm.
component efficiencies and performance maps as well as A life extending control scheme that requires no basic physics and is validated through various engine and engine modification is acceleration and deceleration logic flight tests. It provides a set of gaspath pressures, modification. During acceleration and deceleration, temperatures, and shaft speeds for each mission point in temperature gradients of many life-limited components the flight profile under different ambient conditions.
will build up. Thermo-mechanical fatigue (TMF) and low While it is adequate to use a cycle deck to predict the cycle fatigue (LCF) are functions of both temperature operating conditions of engine components, there are gradients and absolute metal temperatures. Also, metal some shortfalls when applying it to life extending control.
surface delamination is believed to be a strong function of First, the cycle deck is inherently large and it may not be the temperature gradient. It is possible to shape the feasible to run in a life extending control algorithm where acceleration and deceleration schedule of an engine to real-time computation may be necessary. Second, minimize absolute metal temperatures and temperature although the cycle model provides good overall gradients (and thus damage) while still meeting the simulation accuracy, it may not be detailed enough at the performance requirement of the engine [4].
locations that are critical for engine life calculation.
Another control scheme that does not require a An ideal engine model for life extending control is a hardware modification is model based adaptive control to simplified engine model that can give a fast but accurate optimize engine operation at cruise, taking into account overall engine operation results for both steady-state and cruise time, fuel consumption, and rupture/creep life of transient conditions. It shall also have the capability to turbine blades. In an ideal situation, the engine model update engine performance to account for degradation should be such that it can be updated frequently to match based on the sensor measurements. And, for each life- the operation of the deteriorated engine. This control limited component of interest a detailed model will be scheme will be used primary for the off-line flight path used to provide the additional variables required for life planning for establishing the cruise condition. Because of usage estimation.
the long duration of the cruise time, the potential benefit I in terms of engine life and fuel consumption can be very acceleration and deceleration cycles and temperature large while the control scheme is relatively easy to gradients of engine parts.
implement.
Since the life of an engine is directly related to the A multi-level life extending control system is component metal temperature, it is safe to assume that a illustrated in Figure 2. The proposed control architecture has three levels of decision and control. The traditional properly modulated cooling and heating system will be able to reduce component life consumption by reducing tracking control of engine performance is accomplished at the lower Execution Level. This level of control is both the absolute temperature and the temperature executed in real-time. The Coordination Level evaluates gradient of critical parts. However, a modulated cooling/heating system requires some hardware redesign the current health condition of the engine and engine to increase the cooling air capability and actuators that components, monitors the performance, and performs an can respond to the controller's request.
on-line optimization according to the gathered information. This level of control is executed on-line but An active clearance control (ACC) system can also potentially save engine on-wing life by tightly controlling not necessarily in real-time. The highest level is the the blade tip clearance. The tip clearance control can Supervisory Level, which is a discrete-event driven improve the life consumption by reducing rubbing and by process. The supervisory control modes are determined maintaining the engine operation close to the design according to the external commands and the status of temperature even when the engine starts to deteriorate.
engine health and performance conditions. The rationale The main obstacle to implementing a closed-loop ACC for this approach is that the life extending control is a system is the lack of proper tip clearance sensors and combination of continuous monitoring and discrete rapid response actuators. decision-making processes.
Controller A4"..chitecture External Commands/Events The basic concept of life extending control is to adjust the operational settings to minimize damage to critical components while achieving acceptable dynamic Engine Control 1..,_ Supervisor p performance of the plant. It should be emphasized that a High Level Objectives1 Supe..rvisoryLevel fundamental tradeoff exists between the level of ........... v- ...........
achievable performance and the ability to extend the life " Optimization L_ di" . [_ Performance Con tlons of system components generating that performance.
Coordinator _ Performance and In designing a life extending control system the following constraints must be considered: The life usage of a component cannot be measured. It has to be calculated using a life model along with I _r ! I En "he ! i F,.x¢cution Level measurable engine variables such as temperature, pressure, and time. The life model is usually obtained using experimental data and the prediction is probabilistic in nature.
Figure 2. Multi-level Life Extending Control The acceleration and deceleration performance requirements of an engine are usually established under a very strict guideline set by the FAA.
There are many life-limited parts in an engine and 3. Technology Road Map for LEC each typically has several critical failure modes.
The technology areas that can have impacts on the Furthermore, the damage states of all parts involved successful implementation of aircraft engine life need to be monitored and taken into account in the extending control are summarized below. These optimization process. Also, engine maintenance technologies are classified into the following three time practices, such as when to overhaul an engine and frames: under what conditions the used parts are salvaged, play an important role in defining what is optimum in Near term (3-5 years): Technologies that do not require a the control algorithm.
hardware change of current engine systems are generally Different stages of engine operation have different easier to implement and gain acceptance. Improvement in effects on life consumption. For example, during the computational power can also be a factor in LEC cruise condition rupture/creep is the main concern and implementation.
the absolute metal temperature will be used to Improved engine life model and tracking: An ideal calculate the life usage of an engine part at cruise.
engine life model will include models of all critical Low cycle fatigue (LCF) and thermo-mechanicai components. Each component life model can be fatigue (TMF) are determined mainly by the engine calculated using measurable engine variables. All life models are verified experimentally, and their an engine, life of a component will not only be stochastic characteristics are established.
calculated, but it will also be analyzed for possible damage reduction and to ensure that adequate control Modified accel and decel schedules: Accel and decel schedules can be modified to minimize the accommodation authority is provided.
temperature gradients and thermal stresses of life Current NASA activities on LEC include a short-term limited components while meeting the defined study program led by Scientific Monitoring, Inc. (SMI) performance requirements.
On-line engine model: An on-line engine model that and Honeywell that is concentrating on the near term implementation of LEC for the current engine can track engine performance and degradation is especially important for both LEC and health configuration [6]. This program is implementing smart accel/decel schedules to minimize the peak temperature monitoring purposes. It can provide better calibrations to the life calculation for degraded engines, which are and temperature gradient to minimize damages. Under another NASA contract, GE Aircraft Engine (GEAE) has usually at the end stage of their usable life. It can also provide more accurate optimization settings for the completed a trade study to evaluate potential LEC cruise conditions for LEC.
schemes for long-term objectives [7]. Controller Integrated control system for engine health architecture and control algorithm studies are being done management: To implement the proposed multi-level through in-house research as well as university grants, on- life extending control system under the given site research fellowships, and summer fellowships.
constraints is a challenge. A detailed design of how to 4. Conclusions carry out the defined functions and how to resolve the inherent multi-objective optimization problem is The concept of Aircraft Engine Life Extending Control is described. Elements of LEC are discussed in required at this stage.
detail. Technology research areas that may have major Intermediate term (5-10. years): There are several impacts on LEC are also classified according to their technology areas that are on the verge of a breakthrough readiness level for implementation. In summary, the and are very promising for LEC applications. This group Aircraft Engine Life Extending Control is a high pay-off requires small modifications to the engine. technology area that is in its early stage. Inter- Sensors for life prediction: Although a direct life disciplinary research is required to make this technology measurement is not feasible, it is possible to develop applicable to aircraft engines.
sensors that will bring the life calculation closer to the References feedback loop. These sensors include crack detection and measurement sensors, stress measurement sensors, 1. Lorenzo, C. F., et al. "Life Extending Control for and tip clearance sensors. All these sensors must be Rocket Engines," NASA TM 105789, 1992.
2. Lorenzo, C. F., et al. "An Intelligent Control System capable of operating under harsh engine conditions.
- Improved active clearance controls: An improved for Rocket Engines: Need, Vision, and Issues," IEEE active clearance control (ACC) system has the Control Systems Magazine, Vol. 11, Jan. 1991, pp.42- 46.
potential of reducing wear, improving specific fuel 3. Lorenzo, C. F., et al. "Design of Life Extending consumption (SFC), reducing temperatures, and extending life by well controlled transient and steady Control Using Nonlinear Parameter Optimization."
NASA TM 3700, 1998.
state clearances during operation. In addition to the tip 4. Holmes, M., et al. "Life Extending Control of clearance sensors, a new actuator system for a rapid response tip clearance control is also needed. Reusable Rocket Engines" AIAA J'. Guid. Control and Dyn., May/June 1997, pp.621-623.
Long term (10 or more years): Some prospective LEC 5. Ray, A., et al. "Damage-Mitigating Control of technologies require major design changes of engines. Mechanical Systems," ASME J. Dyn. Syst. Meas.
- Flexible engine cooling heating: Engine component Control, Vol. 116, No. 3, Sept. 1994, pp. 437-455.
life can benefit from extra control authority such as 6. Jaw, L., et al. "Design of an Intelligent Life-extending actuators for extra cooling/heating modulation, and Control for Turbine Engines," 2001 American Control extra engine bleed capability to shift operating conference, Washington, DC, June 2001.
conditions.
7. Wiseman, M., et al. "An Investigation of Life Designing for LEC control: It is possible to design an Extending Control Techniques for Gas Turbine engine with the consideration of overall cost including Engines." 2001 American Control Conference, control and maintenance. During the design phase of Washington, DC, June 2001.