FADEC advances allow better engine performance
Next generation of full authority digital engine controls lets pilots select more precise power and raise fuel efficiency.
Conceptual SISO feedback control system.
The basic performance requirements of any control system are absolute and relative stability, steady-state accuracy, and dynamic response (temporal and frequency) characteristics.
Since thrust cannot be measured directly in flight, the thrust rating parameter (TRP) is used instead (EPR or N1). For example, the upper diagram on p 80 shows a simple SISO digital control of the fan speed (LP-spool).
A pilot commands desired N1% by positioning throttles at certain power/throttle lever angles (PLAs). The position sensors/resolvers within the throttle internal mechanism translate the analog angular position signal into digital using analog-to-digital converters (ADCs). This is then compared to a sampled actual N1 speed in a comparator.
Typically, the speed sensors are analog (current/ voltage), which for digital control must be translated into digital. The control error is then forwarded to a digital controller (CPU-plus-memory) and the control decision is made based on the preprogrammed control laws and strategies.
The control decision made by FADEC is then forwarded to an actuator (say, FMV) for corrective action. Control commands may be simple discrete on/off logic signals or amplified continuous analog signals.
Current FADECs have centralized architecture. A single FADEC is engine-mounted and many harnesses provide connections to a multitude of sensors, actuators, communication buses (say, ARINC 429) with airframe (eg, digital air data computers, thrust control computers, throttle resolvers, etc) and power sources, as shown in the lower diagram on this page.
For reasons of weight and other savings, future engine control systems will incorporate distributed (decentralized) architecture. Next-generation ultrahigh-bypass turbofans with smaller lightweight cores will necessitate distributed FADEC systems.
However, distributed engine control will have to rely on emerging high-temperature electronics, sensor and actuation technologies. A further problem will be designing wired and wireless communication protocols and buses connecting all the control subsystems and FADEC.
The diagram on p 81 shows a schematic of a possible future distributed control. In this architecture, supervising FADEC can be installed in a more benign environment, but many sensors and local control nodes will be exposed to harsh operating conditions. In addition, power distribution to distributed onboard electronics will be more complex.
Basic digital SISO negative feedback control of the LP-spool or fan N1%
Before one can implement optimal control strategies, it is necessary to understand the natural dynamic behavior of the jet engine. Two fundamental approaches are available. One is to model mechanical and thermodynamic processes in various engine components (compressors, burner, turbines, etc) using differential equations.
The other scientific method is to perform experimental system identification of an engine during testing and operation. This is basically treating an engine as a "black box" and measuring its response based on known excitations. Both approaches are theoretically very complicated and not appropriate for any discussion here.
Many high-fidelity component-based jet engine mathematical models exist today. Jet engine dynamics truly is inherently nonlinear, with acceleration times and other parameters changing significantly throughout the working envelope (from idle to max thrust).
However, most control designs today are based on a very beautiful and elegant linear control theory. Therefore, we often resort to linearized engine dynamics around a given SP and design controller(s) around it.
For example, in existing modern engine control systems there may be 10–20 piecewise-linear SP controllers which are then "glued" together in one "supercontroller" covering the entire working envelope. Each linear SP regulator has control authority only in a small neighborhood of its set-point (say, N1 of 80% ± 5%).
The transient controller will take over when several SP regulators need to be bypassed and may subsequently delegate authority to a particular SP controller, but it will never allow minimum or maximum fuel flow rates to be exceeded during transient operation. This is often achieved today by simple fuel gain-scheduling, which is the simplest and most basic implementation of nonlinear adaptive control.
One of the popular high-fidelity civilian jet engine simulators is NASA's generic, 2-spool, high-bypass turbofan engine family of commercial modular aero-propulsion system simulation (C-MAPSS) programs.
A graphical user interface (GUI) exists to facilitate the user in designing set-point, transient and limit-protection controllers. C-MAPSS also allows users to simulate the performance of a damaged/ deteriorated engine by varying available health parameters.
Engine sensors and actuators
The job of sensors in modern jet aero engines is to capture dominant engine dynamics, and to do it accurately. What we cannot measure frequently we cannot control. A number of redundant sensors that measure air/gas and fuel/oil flows are installed in modern engines.
In addition, spool dynamics is measured using tachometers and magnetic probes. Jet pilots are familiar with standard cockpit indications such as N1/N2/N3, EPR, EGT/ITT/ TIT, fuel flow (Wf) and vibration gauges. However, there are many installed sensors that communicate with engine controller(s) only.
For example, there is no cockpit indication of the very important combustor static pressure (Pb or Ps3), but that information is crucial in calculating and scheduling proper fuel flow. Moreover, some measured/sampled values are only available to maintenance personnel and are for engine health condition monitoring, forecasting and management.
Many engine sensors operate in harsh environments. Sensors, which are the "eyes" and "ears" of the control systems, are required to operate reliably in high-temperature, high-pressure and corrosive environments, while simultaneously being exposed to significant acceleration/inertial loads.