FADEC advances allow better engine performance
Next generation of full authority digital engine controls lets pilots select more precise power and raise fuel efficiency.
Current centralized FADEC architecture.
Temperatures are typically sensed using thermocouples and RTDs, while pressures are measured using various total and piezoresistive probes. Clearly, a sensor response must be very fast compared to relevant engine dynamics—otherwise, "blurring" of the physical reality will occur, resulting in oscillatory or even unstable engine operation.
Sensors are also used to measure/verify the actual position of some continuously-variable or on/off valves, clearance between casing and the blade tips, etc. Sensors are often part of actuating servomechanisms.
In the future distributed-control architecture, many sensors, actuators and local controllers will be positioned closer to engine components and will thus be exposed to higher temperatures.
Current COTS electronics works fine with peak junction temperatures not exceeding 125°C, but future applications will require electronics working with temperatures on the order of 300–500°C. Among promising high-temperature sensor technologies are silicon-on-insulator (SOI) and silicon carbide.
Actuators are the "muscles" of a control system. An actuator takes the weak signal coming from the controller and converts or amplifies it into, say, force or torque. Typically, electric, hydraulic, pneumatic, mechanical actuators and hybrids of these are used for actuation purposes.
Actuators are by nature analog devices, so the digital control command needs to be converted into electric (voltage/current) signals using digital-to-analog converters (DACs) and then amplified and/or converted into another energy form. Engine actuators work in very harsh environments and need to satisfy multiple requirements regarding operational bandwidth, time response, force/torque capabilities, accuracy, etc.
The electrohydraulic servo valve is a typical example of an actuator used in fuel metering.
Engine controllers and control laws
Schematic of possible future decentralized (distributed) engine control architecture.
A typical FADEC controller consists of a central processor unit (CPU) with associated memories (RAM, ROM, EEPROM) and analog/ digital input/output interfaces. A digital microcontroller is a computer on a single integrated circuit (IC) that can often be found in embedded applications.
A digital controller will have a control program uploaded. A control program that executes all arithmetic and logic operations and implements control laws can be written in a low-level CPU-oriented assembler language or a number of high-level programming languages.
An engine control system with a sampling frequency of 50 Hz (cycles/second) will repeat the control program every 20 milliseconds. All sensor values and set-point PLAs will be sampled and digitized using signal conditioners, multiplexers, sample-and-hold circuits and ADCs. For example, it may take 1–2 milliseconds to read all sensor information and digitize it.
A numerical "error" between the SP and sensed/measured value (say of EPR) will be evaluated and sent to the microcontroller's CPU for arithmetic and logic operations. Different control laws will be implemented and the control action command calculated. The primary feedback loop in EPR/N1 control uses the fuel ratio unit (RU=Wf/Pb) resulting in the required fuel flow being achieved by actuation of the FMV's electrohydraulic servo valve.
Simultaneously, regulation of VSVs, VBVs, VIGVs, etc will be made, based on their own feedback loops or, more frequently, by scheduling it based on the main control variable (say, N1 or EPR).
Traditional linear control laws are proportional (P), PI and PID. The problem with the simple P-control is that it generates steady-state errors which must be corrected by using integral (I) control. Modern jet engine digital controllers see the implementation of various sophisticated control strategies, such as feedforward and cascade loops, lag-lead compensators, optimal linear/nonlinear control strategies, robust control, adaptive control, fuzzy logic and neural networks, artificial intelligence (AI), etc.
Future of engine control systems
SNECMA is in charge of the core, control system (FADEC) and power transmission for the SaM146 engine for the Sukhoi Superjet 100, as well as engine integration and flight testing.
What is the future likely to bring for digital jet engine controls? The current trend is to move toward distributed engine control, where control nodes take "intelligent" independent action in controlling particular local dynamics and are connected to a supervisory FADEC located in a benign environment via communication links (eg, ARINC 629, CAN, TTP).
Micro-electromechanical systems (MEMS) technology provides technical solutions for future miniature localized control nodes where everything can be integrated on a small silicon chip. One can, for example, imagine a fan or compressor in which each blade will have its own feedback microcontrol loops to exercise local control of the boundary layer and avoid stall/surge. Such active compressor stall and surge control (ASSC) promises increased efficiency, lower TSFC and an increase in blade pressure ratio which could result in shorter and lighter engines.
Also, active control of lean-combustion thermoacoustic instabilities would require fast-modulating fuel injector flow control. By rapid modulation of the fuel flow on the order of several hundred hertz, active control of structurally unstable combustion could be achieved. In a way, this, and active stall-surge control, would be similar to how modern braking systems work, where anti-skid valves achieve fine modulation of hydraulic pressure with frequencies of several hundred hertz.
Integration of the health monitoring and management system with the engine control system, and further integration with the flight control systems, will lead to aircraft intelligent integrated supercontrol. It is expected that such design philosophy will lead to increased maneuvering and aircraft control capabilities, lower emissions and noise, better dispatch reliability and higher overall efficiencies. Of course, there will be a price to pay for such high system complexities.
Nihad Daidzic is president of AAR Aerospace Consulting, located in Saint Peter MN. He has worked for many years on the US and European space programs. Daidzic is also a university professor of aviation and mechanical engineering.