Study and Implementation of Algorithms for in flight

performance analysis of the PW4000-100 Turbofan engine

for the purpose of Engine Condition Monitoring

João Pedro Rego Freitas

Thesis to obtain the Master of Science Degree in

Aerospace Engineering

Supervisor: Prof. António José Nobre Martins Aguiar

Examination Committee
Chairperson: Prof. Fernando José Parracho Lau
Supervisor: Prof. António José Nobre Martins Aguiar
Members of the Committee: Prof. Pedro da Graça Tavares Alvares Serrão

July 2014
“Every flyer who ventures across oceans to distant lands is a potential
explorer; in his or her breast burns the same fire that urged adventurers of old
to set forth in their sailing-ships for foreign lands.”
Jean Batten, New Zealand Aviatrix

Créditos da Imagem de Capa: Miguel Nóbrega

i
Acknowledgements

This dissertation marks the end of a journey to become an aerospace engineer that started back in
2008. During the last five and a half years, I lived a great variety of experiences that caused me to
meet and make friendships with a great diversity of people. These people made me grow as a person
and contributed to that I could overcome all obstacles which have arisen. The work presented here
was performed at TAP Portugal, a company that I have always admired and followed with particular
interest. I hope it can continue to carry our flag higher and farther for a long time.
First and foremost, I would like to thank Engr. António Aguiar for giving me the opportunity to develop
this work and for his guidance, shared scientific knowledge and availability during the time I was at
TAP. I would like to offer my special thanks to Engr. Hélder Mendes, the co-supervisor for this work,
for his patience, persistence, technical knowledge and for the time and efforts dedicated to it.
I would like to express my sincere thanks to my everyday colleagues, André Simões, Tiago Duarte
and Luís Vargas for the friendship, motivation and for their contribution to the continuous progress of
this work. I would also like to thank Mr. Manuel Ribeiro, Engr. Manuel Tereno, Engr. Rogério São
Pedro, Engr. Alexandre Garrinhas and all the other people at TAP who have contributed with their
experience and with valuable information to the conclusion of this work.
My deepest thanks to my family for their unconditional support over the years and for being my biggest
source of emotional strength. This work is dedicated to them. Last but not least, to my closest friends,
for all the good times that we have been sharing.

ii
Abstract

The first part of this dissertation describes the implementation of a Flight Data Monitoring (FDM)
database in the Analysis Ground Station (AGS) software for decoding recorded data on an Airbus
A310. The results for a selection of recorded parameters are analysed and the procedure that
computes the flight phase is described. The second part of the dissertation describes the study and
implementation of algorithms for performance trend monitoring of PW4000-100 engines. Two tools
were developed using the R programming language. The first tool is responsible for the acquisition of
stability points in cruise from recorded flight data using specific conditions and criteria and, after
validation, it was implemented in AGS. The second one uses the stability points from the flights
processed in AGS to derive engine baseline models and calculate the performance trends with
respect to these models. The engine parameter data is corrected using the inlet temperature and
pressure and the performance characteristics of the engines are displayed for standard day
conditions. The engine baselines are composed by linear regression models adjusted to the data
obtained from the flights after the engines return from shop visits. The trend monitoring results were
compared to those obtained with the manufacturer’s ECM software and confirmed the potential of the
methodology developed to monitor the gradual deterioration in the engine’s performance and to be
used in the detection of shifts that might be representative of engine faults.

Keywords: Flight Data, Condition-Based Maintenance, Engine Condition Monitoring, Performance

Trends, Corrected Gas Turbine Parameters, Stability Points

iii
Resumo

A primeira parte desta dissertação descreve a implementação de uma base de dados de

Monitorização de Dados de Voo no software AGS (Analysis Ground Station), para descodificação de
dados gravados num Airbus A310. Os resultados para uma seleção de parâmetros gravados são
analisados e o procedimento que calcula a fase de vôo é descrito. A segunda parte da dissertação
descreve o estudo e implementação de algoritmos para monitorização do desempenho de motores
PW4000-100. Dois algoritmos foram desenvolvidos utilizando a linguagem de programação R. O
primeiro algoritmo é responsável pela aquisição de pontos de estabilidade em cruzeiro a partir de
dados de voo gravados usando condições e critérios específicos e, após validação, foi implementado
no AGS. O segundo utiliza os pontos de estabilidade dos vôos processados no AGS para derivar
modelos base do motor e calcular as tendências de desempenho em relação a estes modelos. Os
parâmetros do motor são corrigidos com a temperatura e pressão da tomada de ar e as
características de desempenho dos motores são apresentadas para condições de atmosfera padrão.
Os modelos base do motor são compostos por modelos de regressão linear ajustados aos dados
obtidos dos vôos realizados após os motores voltarem de shop visits. Os resultados da monitorização
nas tendências do motor foram comparados com aqueles obtidos com o software ECM do fabricante
e confirmaram o potencial da metodologia desenvolvida para monitorizar a degradação gradual no
desempenho do motor e para ser usada na detecção de mudanças que podem ser representativas de
falhas do motor.

Palavras-chave: Dados de Voo, Manutenção Baseada na Condição, Monitorização da Condição do

Motor, Tendências de Desempenho, Parâmetros de Turbinas a Gás Corrigidos, Pontos de
Estabilidade

iv
Contents

Acknowledgements ............................................................................................................................. ii
Abstract ............................................................................................................................................. iii
Resumo ............................................................................................................................................. iv
List of Figures .................................................................................................................................. viii
List of Tables ...................................................................................................................................... x
Abbreviations .................................................................................................................................... xii
Nomenclature ................................................................................................................................... xv
1. Introduction .................................................................................................................................1
1.1 Motivation .............................................................................................................................1
1.2 Objectives .............................................................................................................................3
1.3 Thesis Outline .......................................................................................................................3
2. The Airbus A310-325 Aircraft.......................................................................................................5
2.1 General Characteristics .........................................................................................................5
2.2 On-board Transmission and Recording Protocols ..................................................................5
2.2.1 ARINC 429 Protocol ..........................................................................................................5
2.2.2 ARINC 717/573 Protocol ................................................................................................. 10
2.3 ATA Chapters ..................................................................................................................... 12
2.4 ATA 31 Recording Systems: Aircraft Integrated Data System (AIDS) ................................... 13
3. Dataframe Programming ........................................................................................................... 15
3.1 Flight Data Department at TAP ............................................................................................ 15
3.1.1 AGS: A tool for flight data analysis ................................................................................... 15
3.1.2 Database version for the A310-325 aircraft ...................................................................... 17
3.2 Aircraft Definition ................................................................................................................. 18
3.3 Equipment Definition ........................................................................................................... 18
3.4 Parameter Implementation .................................................................................................. 18
3.4.1 Parameter Definition ........................................................................................................ 18
3.4.2 Examples ........................................................................................................................ 20
3.5 Dataframe Construction....................................................................................................... 21
3.5.1 Regular Parameters ........................................................................................................ 21
3.5.2 Superframe Parameters .................................................................................................. 22
3.6 Database Validation ............................................................................................................ 23
4. Decoding Results and Flight Phase Computation Procedure ..................................................... 24
4.1 Results for Recorded Parameters ........................................................................................ 24
4.1.1 Airframe Systems ............................................................................................................ 25
4.1.2 Powerplant ...................................................................................................................... 31
4.2 Additional Procedures: Flight Phase Computation ............................................................... 32
4.2.1 Flight Phases Definition ................................................................................................... 33
4.2.2 Implementation ................................................................................................................ 34
v
5. Engine Condition Monitoring...................................................................................................... 39
5.1 The Condition-Based Maintenance Philosophy .................................................................... 39
5.2 Systems for Data Collection and Analysis ............................................................................ 41
5.3 Engine Parameter Measurements ....................................................................................... 42
5.3.1 The Turbofan Engine ....................................................................................................... 42
5.3.2 Performance and Mechanical Parameters ....................................................................... 43
5.4 Parameter Trend Monitoring ................................................................................................ 44
5.4.1 Cruise Performance Trends ............................................................................................. 44
5.4.2 Take-off Considerations................................................................................................... 47
6. Study of Algorithms for ECM Trend Monitoring .......................................................................... 50
6.1 Case Study: PW4168A Engine ............................................................................................ 50
6.1.1 Description of the Engine................................................................................................. 50
Engine Cruise Report <01> ....................................................................................................... 51
6.2 Stability Points .................................................................................................................... 52
6.2.1 Observation Window and Stable Frame ........................................................................... 52
6.2.2 Quality Number ............................................................................................................... 53
6.2.3 Algorithm for Extraction of Stability Points ........................................................................ 54
6.2.4 Results ............................................................................................................................ 58
6.3 Engine Parameter Corrections............................................................................................. 60
6.3.1 General Formulation ........................................................................................................ 60
6.3.2 Application to the Case Study Engine .............................................................................. 62
7. Baseline Model Definition and Trend Monitoring Results ............................................................ 64
7.1 Implementation of Procedures in AGS ................................................................................. 64
7.2 Engine Baseline Model ........................................................................................................ 65
7.2.1 Engine Selection for Baseline .......................................................................................... 66
7.2.2 Bleed and Pack Selection ................................................................................................ 67
7.2.3 Results ............................................................................................................................ 68
7.3 Trend Monitoring ................................................................................................................. 71
7.3.1 Calculation and Plotting of the Cruise Trends................................................................... 71
7.3.2 Results ............................................................................................................................ 73
8. Conclusions .............................................................................................................................. 79
8.1 Achievements ..................................................................................................................... 79
8.2 Future Work ........................................................................................................................ 80
9. References................................................................................................................................ 81
Appendix A. ARINC 429 Data Sources .......................................................................................... 84
Appendix B. List of Recorded Parameters ...................................................................................... 85
Analog Input Parameters............................................................................................................... 85
Discrete Input Parameters ............................................................................................................. 85
Digital Data Inputs (DITS) ARINC 429 ........................................................................................... 85
Appendix C. Superframe Words ..................................................................................................... 90

vi
 Superframe #1 .............................................................................................................................. 90
Superframe #2 .............................................................................................................................. 90
Superframe #3 .............................................................................................................................. 91
Appendix D. A310: Attitude Angles and Flight Control Surfaces...................................................... 92
Flight Path Vector and Attitude Angles .......................................................................................... 92
ATA 27: Flight Control Surfaces .................................................................................................... 92
Appendix E. A310: Mean Aerodynamic Chord and Center of Gravity Calculation............................ 93
Appendix F. Flight Phase Profiles .................................................................................................. 94
Airbus A310 .................................................................................................................................. 94
SAGEM......................................................................................................................................... 94
Appendix G. A310-325: ECM Instrument Reading Example ........................................................ 95
Appendix H. Codes from the Algorithms developed ........................................................................ 96
Code for the extraction of Stability Points ...................................................................................... 96
Code for the definition of the Engine Baseline Models and the plotting of results for Trend
Monitoring ........................................................................................................................... 96
Appendix I. Baseline Model Results: CS-PWB aircraft .................................................................. 97

vii
List of Figures

Figure 1.1 World Annual Traffic 1967-2012 (RPKs) [2] ........................................................................1

Figure 1.2 Engine Maintenance Costs distribution for 2012 and 2009 comparison [8] ..........................2
Figure 2.1 ARINC 429 Data Transmission Example [12] ......................................................................6
Figure 2.2 ARINC 429 Electrical Characteristics [12] ...........................................................................7
Figure 2.3 ARINC 429 Word Format [12] .............................................................................................7
Figure 2.4 General BCD Word Format [12] ..........................................................................................9
Figure 2.5 General BNR Word Format [12] ..........................................................................................9
Figure 2.6 Frame, Subframe and Word ............................................................................................. 11
Figure 2.7 AIDS Architecture ............................................................................................................. 13
Figure 2.8 Location of AIDS equipment in the Airbus A310 ................................................................ 14
Figure 3.1 AGS activities [23] ............................................................................................................ 17
Figure 3.2 Dialogue window for New Aircraft ..................................................................................... 18
Figure 3.3 Dialogue window for Parameter Implementation ............................................................... 19
Figure 3.4 Flight Number Implementation .......................................................................................... 20
Figure 3.5 AFCS Longitudinal Modes Implementation ....................................................................... 20
Figure 3.6 Flaps position calibration law: (a) table format [21]; (b) graphic display. ............................ 21
Figure 3.7 Regular parameter: Pitch Angle (PITCH) .......................................................................... 22
Figure 3.8 Superframe parameter: Gross Weight (GW) ..................................................................... 23
Figure 4.1 Pressure Altitude results ................................................................................................... 25
Figure 4.2 Flight Speed results .......................................................................................................... 26
Figure 4.3 Pitch and Angle of Attack results ...................................................................................... 26
Figure 4.4 (a) Axes of the aircraft (b) Positive roll angle [27] .............................................................. 27
Figure 4.5 Results for the Longitudinal and Normal accelerations ...................................................... 27
Figure 4.6 Lateral acceleration and Roll angle results ........................................................................ 28
Figure 4.7 Results for the Auto Throttle and Manual Throttle discretes .............................................. 28
Figure 4.8 Results for the Auto Throttle mode discretes..................................................................... 29
Figure 4.9 Flaps and Slats position during approach and landing....................................................... 30
Figure 4.10 Results for the Ailerons and Slats position and Roll angle ............................................... 30
Figure 4.11 Trim Tank Fuel Quantity and CG results ......................................................................... 31
Figure 4.12 Gross Weight results ...................................................................................................... 31
Figure 4.13 Powerplant parameters results: (a) N2; (b) EGT; (c) EPR; (d) FF. ................................... 32
Figure 4.14 N2 and EPR results during the take-off stages ................................................................ 34
Figure 4.15 Flight phase authorized sequences [24] .......................................................................... 36
Figure 4.16 Configuration results during (a) Take-Off and (b) Landing ............................................... 37
Figure 4.17 Flight Phase Results....................................................................................................... 38
Figure 5.1 Engine data acquisition processes at TAP ........................................................................ 41
Figure 5.2 Two-Spool Turbofan Engine [38] ...................................................................................... 43
Figure 5.3 P&W EHM Short Term Trend Plot Report ......................................................................... 45
viii
Figure 5.4 Slow Drifts vs. Step Shifts [30] .......................................................................................... 46
Figure 5.5 Fingerprints for the CFM-56 family of engines [30] ............................................................ 47
Figure 5.6 EGT Considerations [30] .................................................................................................. 48
Figure 5.7 Deterioration effect on the EGT Margin and OATL [42] ..................................................... 49
Figure 6.1 PW4168A: Gas Path configuration areas and Engine Stations .......................................... 51
Figure 6.2 Format of the A330’s Engine Cruise Report <01> [46] ...................................................... 51
Figure 6.3 Stable and non-stable variation of a parameter inside an Observation Window ................. 52
Figure 6.4 Individual and Gliding Window methods............................................................................ 53
Figure 6.5 Flowchart of the Algorithm for Extraction of Stability Points ............................................... 57
Figure 6.6 Representation of the Waiting and Search times............................................................... 58
Figure 6.7 Flight Results: Search Periods, Waiting Times and Stability Point ..................................... 59
Figure 6.8 Twin spool mixed flow turbofan [47] .................................................................................. 60
Figure 7.1 Flowchart: Implementation of Procedures for ECM in AGS ............................................... 65
Figure 7.2 Cruise Performance Characteristics of the JT3C-6 engine [39] ......................................... 66
Figure 7.3 Baseline Results: (a) EGTK vs. EPR (b) FFK vs. EPR (c) N2K vs. EPR (d) N1K vs. EPR.. 68
Figure 7.4 Baseline Results for EGTK vs EPR at (a) 37000ft and (b) MN=0.80 .................................. 69
Figure 7.5 CS-PWA Baseline Model Results: (a) Aircraft Stability Criteria (b) Converged Stability
Criteria. ............................................................................................................................................. 70
Figure 7.6 Distribution of Baseline Points by PKS Code .................................................................... 71
Figure 7.7 Distribution of Baseline Points by (a) Mach Number and (b) Altitude and Search Period ... 71
Figure 7.8 Delta Parameter Calculation ............................................................................................. 72
Figure 7.9 Process of Plotting the Cruise Performance Trends .......................................................... 73
Figure 7.10 CS-PWA: Timeline for Engine Removals/Installations ..................................................... 73
Figure 7.11 Cruise Performance Trends for the CS-PWA aircraft: (a) Delta EGT; (b) Delta FF; (c) Delta
N2..................................................................................................................................................... 74
Figure 7.12 CS-PWB: Timeline for Engine Removals/Installations ..................................................... 75
Figure 7.13 Cruise Performance Trends for the CS-PWB aircraft: (a) Delta EGT; (b) Delta FF; (c) Delta
N2..................................................................................................................................................... 76
Figure 7.14 P&W EHM: CS-PWB Engine #1 Long Term Trend Plot Report ....................................... 77
Figure 7.15 P&W EHM: CS-PWB Engine #2 Long Term Trend Plot Report ....................................... 78
Figure D.1 Representation of the Flight Path angle, Pitch angle and Angle of Attack [11]................... 92
Figure D.2 Representation of the Primary and Secondary Flight Control surfaces on the A310 aircraft
[11] ................................................................................................................................................... 92
Figure E.1 A310: Airplane Dimensions [52] ....................................................................................... 93
Figure F.1 Flight Profile with Airbus Flight Phases [21] ...................................................................... 94
Figure F.2 Flight Profile with the SAGEM Flight Phases [24] .............................................................. 94
Figure I.1 CS-PWB Baseline Model Results: (a) Aircraft Stability Criteria (b) Converged Stability
Criteria. ............................................................................................................................................. 97
Figure I.2 Distribution of Baseline Points by PKS Code ..................................................................... 98
Figure I.3 Distribution of Baseline Points by (a) Mach Number and (b) Altitude and Search Period .... 98

ix
List of Tables

Table 2.1 Airbus A310-325 Specifications [10], [11] .............................................................................5

Table 2.2 Receiver Voltage Levels ......................................................................................................6
Table 2.3 ARINC 429 Bit Characteristics .............................................................................................7
Table 2.4 SSM codes for BCD, BNR and Discrete Data ......................................................................8
Table 2.5 Label 256 computation ........................................................................................................8
Table 2.6 BCD Word encoding example: Distance to DME ..................................................................9
Table 2.7 BNR Word encoding example: True Airspeed .................................................................... 10
Table 2.8 Computation of the True Airspeed ..................................................................................... 10
Table 2.9 Examples of Regular Parameters ...................................................................................... 12
Table 3.1 Recorded Superframes...................................................................................................... 22
Table 4.1 Recorded Parameters........................................................................................................ 25
Table 4.2 Flap and Slat Control Lever Positions ................................................................................ 29
Table 4.3 Airbus A310-325 Flight Phases Definition .......................................................................... 33
Table 4.4 SAGEM Flight Phases ....................................................................................................... 34
Table 4.5 Conditions for the normal transition between the flight phases (Database version 10079) .. 35
Table 4.6 List of parameters used in the Flight phase computation procedure (*Same information as
LDG_ON_1 and LDG_ON_2 (0-AIR, 1-GROUND); **Array) .............................................................. 36
Table 4.7 Flaps and Slats Configuration ............................................................................................ 37
Table 6.1 Specifications of the PW4168A Engine [44] ....................................................................... 50
Table 6.2 Discrete Parameters used in the Basic Conditions ............................................................. 54
Table 6.3 Binary Parameters used in the Basic Conditions ................................................................ 54
Table 6.4 Logic for True Basic Conditions and Parameters used ....................................................... 55
Table 6.5 Stable Frame Parameters .................................................................................................. 55
Table 6.6 Initial Stability Criteria ........................................................................................................ 56
Table 6.7 List of Filtered Parameters ................................................................................................. 56
Table 6.8 Aircraft and Converged Stability Criteria............................................................................. 59
Table 6.9 Common Gas Turbine Parameter Corrections ................................................................... 61
Table 7.1 List of Procedures for ECM implemented in the A330-223 fleet database version............... 64
Table 7.2 Engine Installations after Shop Visits with Performance Restoration .................................. 67
Table 7.3 Logic for ABC and PKS codes ........................................................................................... 67
Table A.1 ARINC 429 Data Sources in database version 10079 ........................................................ 84
Table B.1 List of Analog input parameters ......................................................................................... 85
Table B.2 List of Discrete input parameters ....................................................................................... 85
Table B.3 List of Digital input parameters: Flight Control Computer (FCC) ......................................... 85
Table B.4 List of Digital input parameters: Thrust Control Computer (TCC) ........................................ 86
Table B.5 List of Digital input parameters: Flight Control Unit (FCU) .................................................. 86
Table B.6 List of Digital input parameters: GMT Clock ....................................................................... 86
Table B.7 List of Digital input parameters: System Data Analog Converter (SDAC)............................ 86
x
Table B.8 List of Digital input parameters: Flight Warning Computer (FWC)....................................... 87
Table B.9 List of Digital input parameters: Ground Proximity Warning System (GPWS) ..................... 87
Table B.10 List of Digital input parameters: Instrument Landing System Receiver (ILS) ..................... 87
Table B.11 List of Digital input parameters: Radio Altimeter ............................................................... 87
Table B.12 List of Digital input parameters: VHF Omnidirectional Range Receiver (VOR) .................. 87
Table B.13 List of Digital input parameters: Flight Management Computer (FMC).............................. 87
Table B.14 List of Digital input parameters: Flight Navigation Symbol Generator (FNSG) .................. 88
Table B.15 List of Digital input parameters: Distance Measuring Equipment (DME) ........................... 88
Table B.16 List of Digital input parameters: Traffic Alert and Collision Avoidance System (TCAS)...... 88
Table B.17 List of Digital input parameters: Flight Navigation Control Panel (FNCP) .......................... 88
Table B.18 List of Digital input parameters: Air Data Computer (ADC) ............................................... 88
Table B.19 List of Digital input parameters: Inertial Reference System (IRS) ..................................... 88
Table B.20 List of Digital input parameters: Electronic Engine Control (EEC) ..................................... 89
Table B.21 List of Digital input parameters: Fuel Quantity Indicator (FQI) .......................................... 89
Table B.22 List of Digital input parameters: Center of Gravity Control Computer (CGCC) .................. 89
Table C.1 Superframe #1 .................................................................................................................. 90
Table C.2 Superframe #2 .................................................................................................................. 90
Table C.3 Superframe #3 .................................................................................................................. 91
Table D.1 List of the Flight Control surfaces on the Airbus A310 ........................................................ 92

xi
Abbreviations

A/C Aircraft
A/THR Auto Thrust
ABC Aircraft Bleed Code
ACARS Aircraft Communication Addressing and Reporting System
ACMS Aircraft Condition and Monitoring System
AGL Above Ground Level
AGS Analysis Ground Station
AIDS Aircraft Integrated Data System
AMM Aircraft Maintenance Manual
APU Auxiliary Power Unit
ARINC Aeronautical Radio INCorporated
ATA Air Transport Association of America
BCD Binary Coded Decimal
BNR Binary
BPRZ Bipolar return-to-zero
C.P. Control Panel
CG Center of Gravity
CGCC Center of Gravity Control Computer
CM Condition Monitoring
CPT Corner Point Temperature
CSV Comma Separated Values
DAR Digital AIDS (Airbus A310)/ACMS (Airbus A330) Recorder
DFDAU Digital Flight Data Acquisition Unit
DFDR Digital Flight Data Recorder
DME Distance Measuring Equipment
DMU Data Management Unit
ECAM Electronic Centralized Aircraft Monitoring
ECC Engine Electronic Control
ECM Engine Condition Monitoring
EGT Exhaust Gas Temperature
EGTHD EGT Hot Day
EHM Engine Health Monitoring
EPR Engine Pressure Ratio
FAA Federal Aviation Administration
FD Flight Data
FDA Flight Data Analysis
FDIU Flight Data Interface Unit
FDM Flight Data Monitoring
xii
FDR Flight Data Recorder
FDRPL Flight Data Recording Parameter Library
FF Fuel Flow
FL Flight Level
FOQA Flight Operations Quality Assurance
FRT Flat Rate Temperature
GDP Gross Domestic Product
GE General Electric
GMT Greenwich Mean Time
HP High Pressure
HPT High Pressure Turbine
HT Hard Time
IATA International Air Transport Association
ICAO International Civil Aviation Organization
ISA International Standard Atmosphere
L.A. Linear Accelerometer
LCC Low Cost Carrier
LP Low Pressure
LPT Low Pressure Turbine
LSB Least Significant Bit
MAC Mean Aerodynamic Chord
MOQA Maintenance Operations Quality Assurance
MSB Most Significant Bit
N1 Low-pressure rotor/Fan speed
N2 High-pressure rotor/Core speed
ND Navigation Display
OAT Outside Air Temperature
OATL Outside Air Temperature Limit
OC On-Condition
OEM Original Equipment Manufacturer
OQAR Optical Quick Access Recorder
PFD Primary Flight Display
PKS Air Conditioning Pack Code
PW/P&W Pratt & Whitney
QAR Quick Access Recorder
RPK Revenue Passenger Kilometers
RPM Rotations per minute
SDI Source/Destination Identifier
SF Subframe
SSM Sign/Status Matrix

xiii
SV Shop Visit
TAP Transportes Aéreos Portugueses
TAS True Airspeed
TAT Total Air Temperature
TCC Turbine Case Cooling
TO Take Off
UTC Universal Time Coordinated
WF Fuel Flow

xiv
Nomenclature

Latin Symbols
Theta ( ) correction exponent
Delta ( ) correction exponent
Raw delta EGT
Raw delta FF
Raw delta N2
EGT value from the stability point
Corrected EGT
Corrected EGT (from the baseline model)
Corrected EGT (computed from the value of the stability point)
FF value from the stability point
Corrected FF
Corrected FF (from the baseline model)
Corrected FF (computed from the value of the stability point)
Tip Mach number
Number of samples of the parameter in the observation window
Rotational speed (RPM)
New Filtered Value
New Raw Value
N1 value from the stability point
Corrected N1
N2 value from the stability point
Corrected N2
Corrected N2 (from the baseline model)
Corrected N2 (computed from the value of the stability point)
Old Filtered Value
Pressure at ISA Sea Level conditions (1013.25hPa)
Total pressure at the inlet
Generic gas path parameter
Corrected gas path parameter
Inlet Pressure (Station 2)
Engine Quality Number
Radius of the fan blades
Samples per second (DMU)
Time Constant (DMU)
Temperature at ISA Sea Level conditions (288.15K)
Total temperature at the inlet

xv
 Inlet Temperature (Station 2)
Maximum variation of the parameter allowed in the observation window
Inlet velocity of the air
Exhaust velocity
Variance of parameter I in the observation window
Tangential velocity
Weight factor of parameter I
̅ Mean of the parameter in the observation window

Greek Symbols
Angle of Attack
Smoothing coefficient
Flight Path Angle
Pressure correction factor
Pressure correction factor computed with the total pressure
Propulsive Efficiency
Pitch Angle
Temperature correction factor
Temperature correction factor computed with the total temperature

Constants
Gas constant, 287 Jdeg-1kg-1
Ratio of specific heats/Adiabatic index, 1.4

xvi
1. Introduction

1.1 Motivation
The aviation industry plays an important role in the global economic activities. Over the last decades,
air transportation evolved from a luxury good to an everyday good and has been characterized by an
almost continuous increase in passenger traffic numbers. In 2013, airline scheduled operations
handled 3.1 billion passengers, up from 2.9 billion passengers from 2012 [1]. Airbus [2] predicts that
air traffic will double in the next 15 years at an annual growth rate of 4.7% in RPK (Revenue
Passenger Kilometers), based on current projections. Despite this growth, the airline industry has
consistently reported losses that ascend to billions of dollars only in the last decade, which resulted in
many premature bankruptcies. The poor performance of the air transport industry in terms of
profitability can be explained by the apparent inability of the airlines to adjust to fast variations in the
demand, caused by variations in the macroeconomic climate - represented, for example, by the Gross
Domestic Product (GDP) [3]. External demand shocks, such as the 9/11 terrorist attacks or the
financial crisis of 2007/2008, produce gaps between demand and capacity, where large reductions in
revenue are not compensated quickly enough by the corresponding reduction in the airline costs. In
addition to these factors, the constant fluctuations in the oil prices and the increasing competition from
Low-Cost Carriers (LCC), also affect the stability of the airlines and have forced more established
carriers to search for ways of reducing costs and streamlining the different areas of operation, in order
to become more efficient than the competition and to guarantee their economic viability.

Figure 1.1 World Annual Traffic 1967-2012 (RPKs) [2]

Despite the excellent safety record of air travel, aircraft accidents always have major economical and
emotional consequences. The Flight Data Analysis (FDA), also called Flight Data Monitoring (FDM), is
a program used in the prevention of aeronautical accidents that consists in the decoding and analysis
of the data from the aircraft recorders on a routine basis. TAP Portugal was one of the first airlines in
the world to use its flight data in a preventive way through the implementation of an FDM program
containing both FOQA (Flight Operations Quality Assurance) and MOQA (Maintenance Operations
Quality Assurance) procedures, starting in the early sixties.
1
The objective of FOQA procedures is to identify divergent practices amongst crew members or
difficulties to adhere to the established standard procedures. The information is then compiled into
reports and corrective actions may be taken, leading to improvements in the safety performance of the
airline. References [4] and [5] deal with some practical cases of FOQA procedures and describe their
implementation using specific tools available at TAP for processing recorded flight data. The MOQA
procedures are used for maintenance purposes and are mainly executed as a diagnosis tool, to help
the various airline maintenance departments to identify and correct on-going problems and failures in
the aircraft equipment or to conduct studies for improved fuel consumption, for instance.
Flight data can also be used to continuously monitor the performance of the aircraft and the condition
of its systems. In addition to the diagnostic analysis, monitoring the aircraft/aircraft systems greatly
relies on a prognostic analysis, where engineering personnel predict the time left until component
failure based on its deterioration levels. This type of analysis is known to reduce the risk of potential
failures during operation and allows planning for maintenance in anticipation. The work performed in
[6] focuses on the development and validation of a method for on-ground data acquisition of data
during cruise for aircraft performance monitoring. The objective was to have data leading to more
reliable assessments of the degradation in each aircraft and to overcome the lack of flexibility from the
on-board systems currently used for these functions.
Engine maintenance is often the highest maintenance cost and can have a major impact on the
profitability of airlines [7]. Engine Condition Monitoring (ECM) is the process used by the operators to
monitor the health of their engines on a daily basis. The performance of each engine is checked
against a reference model using sophisticated computational tools, where the deterioration levels can
be visualized as a function of time. This information is used for planning engine removals, alert about
poor performance and to provide indications on precursors to actual engine failures that could lead to
extensive repairs. Generally, the sources of data used in this process are the on-board computers.

Figure 1.2 Engine Maintenance Costs distribution for 2012 and 2009 comparison [8]

The increasing recording capacity of the flight data recorders, the development of newer aircraft fitted
with dedicated sensors for a more comprehensive and accurate prognostic of the aircraft/aircraft
systems and the prospect of transmitting real-time information, give FDM programs almost unlimited
capabilities to enhance safety levels and reduce airline operating costs, allowing airlines to respond to
the harsh economic requirements.

2
1.2 Objectives
The current dissertation describes the work that was performed within the Flight Data department at
TAP Portugal. The first objective of the work was the implementation of an FDM database in the AGS
(Analysis Ground Station) software capable of decoding the flight data from the recorders of an Airbus
A310 aircraft. This required the knowledge of the on-board transmission and recording protocols.
Additionally, the database includes some procedures for an improved flight analysis. These
procedures allow the correct separation of the different phases during each flight and the correct
identification and storage of the analysed flights.
The second and main objective of the work was the development of algorithms for engine
performance analysis, for the purpose of Engine Condition Monitoring. Two separate tools were
created using R programming language with direct application to the PW4168A1 engine. The first tool
is responsible for the acquisition of stability points from the recorded flight data. The stability points
contain information about aircraft and engine parameters averaged over a period during cruise where
a set of conditions and stability criteria were met. The logic used by the on-board computer systems
was initially used for validation and, afterwards, a new set of more restrictive criteria was studied and
led to a reduction in the number of points encountered in each flight. Only one stability point is
captured every four hours during cruise and its selection is based on the quality number that is
computed from statistical variables. The tool was then implemented in AGS. The second tool
establishes a baseline model with the expected performance characteristics of the engine and
calculates the performance deterioration levels of each engine. The inputs are the stability points
acquired from the flights processed in AGS. The performance of the engine is modelled from a series
of flights after the engine returns from shop visits with performance restoration. The engine parameter
data is corrected taking into account the temperature and pressure at the engine’s inlet.
The potential of the developed method to monitor the gradual deterioration of the engines will be
evaluated and the capability to detect failures or malfunctions in the engine will also be studied by
crossing the data obtained with information from the engine manufacturer’s software.

1.3 Thesis Outline

After this introductory chapter, the thesis is organized in seven chapters:
 Chapter 2: introduces the general characteristics of the Airbus A310-325, the aircraft from
which the data will be decoded. It describes the ARINC 429 protocol, used in the transmission
of data between the avionics equipment of the aircraft, and the ARINC 717 protocol, which
describes the format of the data stored in its recorders. The architecture of the recording
system in the A310-325 is shown and the function of each recorder is described.
 Chapter 3: contains the different steps that were followed in order to implement a new
database version in AGS, containing the necessary information to decode the recorded

1
PW stands for Pratt & Whitney
3
 parameters. The recordings are organized in a structure known as frame or dataframe. The
knowledge of the protocols described in Chapter 2 is fundamental during this procedure.
 Chapter 4: this Chapter presents the results for a selection of recorded parameters, which are
displayed as a function of time and grouped according to their ATA chapter. The Flight Phase
computation procedure and its implementation are described afterwards, mainly due to its
importance for the flight data analysis;
 Chapter 5: introduces the topic of Engine Condition Monitoring. The role of ECM in the engine
maintenance program is discussed, as well as its contribution to the condition-based
maintenance philosophy. The most common systems used for data collection and analysis,
the type of parameters measured and the parameter trend monitoring method used to analyze
the engine performance and evaluate its deterioration are also presented.
 Chapter 6: this Chapter starts by introducing the case study engine, the PW4168A. The
definition of observation window, stability point and the necessary equations for the calculation
of the quality number are presented. The algorithm for the extraction of stability points is
described and the relevant results presented. Finally, the correction procedure used for
performance trend monitoring of the PW4168A engine is also presented and uses equations
that were found in the literature.
 Chapter 7: goes through the process of definition of engine baseline models, which are based
on the stability points acquired from the algorithm developed in the previous chapter. The
performance trend monitoring results containing the evolution of the deterioration levels with
respect to the derived baseline models are shown and analyzed for two aircraft.
 Chapter 8: presents the final conclusions about what has been achieved with the current work,
together with relevant topics for future work that could be followed in order to continue the
work developed so far.

4
2. The Airbus A310-325 Aircraft

The first part of this work deals with the decoding of Airbus A310 recorded data. In this Chapter, the
general characteristics of the aircraft are presented as well as the on-board protocols used for the
transmission and recording of the data. Afterwards, the A310 recording system is described.

2.1 General Characteristics

The Airbus A310 is a wide body transport jet airliner powered by two turbofan engines. The A310
made its maiden flight in 1982 and during its production run 255 units have been delivered to different
customers, with a large majority of them still in operation worldwide [9]. Depending on the seating
layout the aircraft is certified to carry up to 275 passengers. Table 2.1 lists the specifications for the
A310-325 variant, powered by PW4156A Engines.
General Characteristics
Overall length 46.66 m
Height 15.80 m
Wing span 43.90 m
Fuselage width 5.64 m
Cabin length 33.25 m
Maximum cabin width 5.28 m
Mean Aerodynamic Chord (M.A.C.) 5.8287 m
Wheelbase 15.22 m
Crew 2 (minimum)
Seating Capacity 220 (2-class configuration)
Performance
Powerplant 2 Pratt & Whitney PW4156A
Thrust 56 000 lbf (x2)
Range 8050 km
Maximum Operating Mach number (MMO) 0.84
Maximum taxi weight (MTW) 164 900 kg
Maximum take-off weight (MTOW) 164 000 kg
Maximum landing weight (MLW) 124 000 kg
Maximum zero fuel weight (MZFW) 114 000 kg
Maximum fuel capacity 61 070 litres
Table 2.1 Airbus A310-325 Specifications [10], [11]

2.2 On-board Transmission and Recording Protocols

The communication between the main systems in the Airbus A310 is made through the ARINC 429
protocol. The format in which the data is stored in the different recording units on-board the aircraft is
specified by the ARINC 717/573 protocol. ARINC stands for Aeronautical Radio INCorporated, a
company established in 1929 by four major airlines that develops and operates systems and services
to ensure the efficiency and performance in the communications in the aviation industry [12]. It has
evolved to become the global industry leader in the development of specifications and standards for
avionics equipment. In 2013, the company was acquired by Rockwell Collins [13].

2.2.1 ARINC 429 Protocol

The continuous increase in the flux of data from the aircraft systems dictated the need to standardize
the transmission between the aircraft electrical data buses. The ARINC 429 is a protocol that defines

5
how the different avionics equipment and aircraft systems should communicate with each other. They
are interconnected by wires in twisted pairs and the communication is done with a unidirectional data
bus standard, known as Mark 33 Digital Information Transfer System (DITS). The transmission and
reception is made on separate ports.
The ARINC 429 protocol is one of the most commonly used protocols in the aviation industry, being
installed on many commercial aircraft, including the Airbus A310/A320 and A330/A340 and Boeing
727, 737, 747, 757 and 767. However, alternative systems are being developed on most recent
aircraft, such as the Boeing 777, in order to reduce the weight of the wires needed and to transmit
data at higher rates [12].

2.2.1.1 ARINC 429 Electrical Characteristics

Each ARINC 429 bus uses two signal wires to transmit the messages, which consist of 32 bit words.
The messages are transmitted at a bit rate of 12.5 or 100 kilobits per second, depending on the
specification of the bus. The transmission is asynchronous, with consecutive words separated by at
least 4 bits with null value (zero voltage), eliminating the need for an external clock signal.
The bits are transmitted using a bipolar return-to-zero (BPRZ) modulation consisting of three states:
HI, LO and NULL. The information is retrieved from the voltage difference between the two wires, also
called the differential voltage. The nominal transmission voltage is 10±1 volts, with either a positive or
negative polarity. Therefore, to each wire corresponds a signal leg that ranges between -5V and +5V.
When one leg is +5V, the other is -5V and vice-versa. One wire is called the “A” (or “+” or “HI”) and the
other is the “B” (or “-” or “LO”). The differential voltage at the receiver terminals depends on the line
length and the number of receivers connected to each bus, which is limited to 20 by the protocol. In
addition, the voltage levels can be perturbed by noise and pulse distortion [14]. Depending on the
differential voltage, the receiver terminals associate the following voltage ranges with the three signal
states in table 2.2.
Voltage Differential (A to B) Signal State
+6.5 V to +13 V HI
-2.5 V to +2.5 V NULL
-13 V to -6.5 V LO
Table 2.2 Receiver Voltage Levels

Figure 2.1 shows a transmission example for a 32 bit data word, with the representation of the signal
legs from wires A and B.

Figure 2.1 ARINC 429 Data Transmission Example [12]

6
The ARINC 429 bit characteristics [12] are summarized in table 2.3 for both transmission rates and a
graphical representation is shown in figure 2.2. The Time Y is the time between two successive flanks
and the Time X is the time the signal stays at a HI or LO level.
Parameter High Speed Low Speed
Bit Rate 100 kbit/s 12.5 kbit/s
Time Y (one bit) 10 µsec ± 2.5% 1/(bit rate) µsec ± 2.5%
Time X 5 µsec ± 5% Y/2 µsec ± 5%
Pulse Rise time 1.5 ± 0.5 µsec 10 ± 5 µsec
Pulse Fall time 1.5 ± 0.5 µsec 10 ± 5 µsec
Table 2.3 ARINC 429 Bit Characteristics

Figure 2.2 ARINC 429 Electrical Characteristics [12]

2.2.1.2 Format of the Data Word

The ARINC 429 data words have a length of 32 bits and typically use the format shown in figure 2.3,
which includes five primary fields: parity bit (P), Sign/Status Matrix (SSM), data, Source/Destination
Identifiers (SDI) and the Label. The bits from 1 to 32 are specified by the protocol.

Figure 2.3 ARINC 429 Word Format [12]

Parity
The Most Significant Bit (MSB) in the ARINC 429 word format is the parity bit (P). The ARINC 429
uses an odd parity, which means that the total number of “1” bits in each 32-bit word should be odd.
This is ensured by setting the parity bit to ”0” when there is an odd number of “1” bits from bits 1 to 31
and to “1” when it is even. The parity bit allows the receiver to identify if transmission errors have
occurred. However, this method of identifying errors is only effective when the number of incorrect bits
is odd. When this is not the case, the errors can remain undetectable.

Sign/Status Matrix (SSM)

Bits 30 and 31 contain the Sign/Status Matrix (SSM). This field contains information about the
conditions of the hardware equipment, operational mode or validity of the data content, in the form of a
code. The applicable codes and their meanings, which differ depending on the type of encoded data,
are indicated in table 2.4 [14].

7
 Meaning
Bit Number
BCD BNR Discrete
31 30
0 0 Plus, North, East, Right, To, Failure Warning Verified Data, Normal
Above operation
0 1 No Computed Data No Computed Data No Computed Data
1 0 Functional Test Functional Test Functional Test
1 1 Minus, South, West, Left, Normal Operation Failure Warning
From, Below
Table 2.4 SSM codes for BCD, BNR and Discrete Data

Data
Bits 29 to 11 contain the data from the parameter(s) in various formats. There are some standard
types of data, which will be presented in the next section together with some examples, and non-
standard types specifically implemented by the manufacturers. Usually, not all the bits are used, only
those necessary to cover the range and resolution of the information transmitted.

SDI
Bits 10 and 9 form a two-bit code that gives the Source/Destination Identifier (or SDI). This is used to
identify the receiver to which the information is destined when there are multiple receivers. It is also
used to identify the source of the transmission or, in some cases, for transmission of data. When this
occurs the SDI field is not used and this field is overlapped by the parameter Data field [12].

Label
Bits 8 to 1 contain a label identifying the data type of the 32-bit word and the parameters associated
with it. A label is assigned to each parameter that is transmitted through the aircraft’s avionic
equipment. The labels are listed in the ARINC specification [14] and are represented as octal numbers
with three digits, from 0 to 377. Table 2.5 exemplifies how label 256 is computed from the transmitted
bits.
Word bits 1 2 3 4 5 6 7 8
Binary 1 0 1 0 1 1 1 0
Octal 2 5 6
Table 2.5 Label 256 computation

The most significant digit occupies two bits and the least significant digit and middle digit occupy 3 bits
each. Although the maximum label number corresponds to 377, there are only 255 labels available:
. In the label, the most significant bit (bit 8) is transmitted first. With the
other bits from the ARINC word, the reverse order is used. The transmission order is as follows [12]:
8, 7, 6, 5, 4, 3, 2, 1, 9, 10, 11, 12, 13, …, 32

2.2.1.3 ARINC 429 Data Types

The data transmitted with the ARINC 429 protocol can be of several types. For each type, there is a
specific format of encoding the data in the 32 bit words. The following types of data and their formats
are addressed in this section: BCD (Binary Coded Decimal), BNR (Binary) and Discretes.

8
BCD (Binary Coded Decimal)
This is a very common format that is generally used to represent parameters limited to specific values.
Some examples include radio frequencies or distances to radio-navigation equipment. On the other
hand, the BNR type is more suitable for parameters such as temperatures, speeds, accelerations, etc.
and will be explored next. Figure 2.4 displays the general format of a BCD word. The data field in a
BCD word contains up to five sub-fields, which represent different decimal digits. The SSM field gives
the sign, according to the logic in table 2.4.

Figure 2.4 General BCD Word Format [12]

Four bits are allocated to each digit, except for the most significant sub-field, which encodes the digit
with 3 bits only. Therefore, the maximum decimal value represented by the first digit is 7. In the event
that this value is greater than 7, bits 29, 28 and 27 are set to zero (padded with zeros), and the second
sub-field becomes the most significant. The most significant digit is now represented by four bits,
which allow reaching a maximum value of 15: .
An example of a BCD word is in table 2.6. The data indicates the distance to the DME (Distance
Measuring Equipment), a radio-navigation equipment: 257.86 NM.
32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1
P SSM CHAR 1 CHAR 2 CHAR 3 CHAR 4 CHAR 5 SDI LABEL
0 0 0 0 1 0 0 1 0 1 0 1 1 1 1 0 0 0 0 1 1 0 0 0 1 0 0 0 0 0 0 1
+ 2 5 7 8 6 201
Table 2.6 BCD Word encoding example: Distance to DME

BNR (Binary)
The BNR type is also a very common ARINC 429 data format, in which the data is stored as a binary
number. The general format and fields are represented in figure 2.5. Bit 29 is the sign bit and bit 28 is
the most significant bit of the data field. Negative numbers are encoded in complement of two of the
positive values: when bit 29 is “1”, the number is negative and when it is “0”, the number is positive. Bit
th
28 represents half of the maximum value (range) of the parameter, bit 27 represents 1/4 of the range,
etc.

Figure 2.5 General BNR Word Format [12]

The numerical value of the parameter is obtained by multiplying the bit values by their respective
fraction of the range. In addition to the range, which constitutes an interval of the values the parameter
can have, the resolution is an important parameter to consider in the computation. It defines the
precision of the least significant bit of the parameter. These two properties, range and resolution, are
related to each other according to the equation below [5]:

9
 (2.1)

Where is the number of bits used by the parameter. For many parameters it is not necessary to fill
all the available bits with information. takes values up to 20 (bits 28 to 9), in which case the SDI is
not recorded. Table 2.7 gives an example of BNR encoded data for the True Airspeed (TAS). The “P”
denotes pad “zero”. A possible use for these pad bits are discrete parameters.
32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1
P SSM DATA FIELD SDI LABEL
0 1 1 0 0 1 0 0 0 1 1 0 1 0 1 0 0 0 0 P P P 0 0 0 0 0 1 0 0 0 1
0 0 + 565 Knots 210
Table 2.7 BNR Word encoding example: True Airspeed

The parameter characteristics – range, resolution, number of data bits, etc. - are available from the
ARINC specification and/or the Aircraft Maintenance Manual (AMM). The range for the example
15
parameter is 2048 and the number of significant bits is 15. The resolution is equal to 2048/2 =0.0625.
The TAS is always positive, so bit 29 is always set to zero. Table 2.8 shows how the value of the
parameter is computed.
Bit Nr. Bit Value
Binary Decimal [Knots]
28 0 (2048/2)*0=0
27 1 (2048/22)*1=512
26 0 (2048/23)*0=0
25 0 (2048/24)*0=0
24 0 (2048/25)*0=0
6
23 1 (2048/2 )*1=32
22 1 (2048/27)*1=16
21 0 (2048/28)*0=0
20 1 (2048/29)*1=4
19 0 (2048/210)*0=0
18 1 (2048/211)*1=1
17 0 (2048/212)*0=0
16 0 (2048/213)*0=0
15 0 (2048/214)*0=0
14 0 (2048/215)*0=0
Parameter Value ( ) 565
Table 2.8 Computation of the True Airspeed

Discrete
The Discrete type of data is used whenever the parameter can be encoded with only one bit. The
discrete data is accommodated in the unused pad bits of data words or in dedicated words, which can
store up to 19 different parameters. For instance, these parameters can describe the status of engine
valves, flight surfaces, navigation modes, etc. Some examples of the two states indicated by the bit
are: ARMED/NOT ARMED, AIR/GROUND, ENGAGED/NOT ENGAGED, ON/OFF, SELECTED/NOT
SELECTED.

2.2.2 ARINC 717/573 Protocol

The ARINC 717 protocol was created to describe the equipment and installation standards capable of
meeting mandatory flight data recording requirements prescribed by the FAA (Federal Aviation
Administration) and other aviation authorities. This protocol supersedes the ARINC 573 specification
that defined the output format for use in Flight Data Recorders (FDR) from a continuous data stream
of Harvard Bi-Phase 12-bit words, which were then formatted in frames [12]. The ARINC 717 protocol

10
performs the same functions but includes a number of different bit rates and frame sizes, to reflect the
evolution in the aircraft recording systems [15]. It also encodes the data stream in BPRZ format, the
same as ARINC 429.

2.2.2.1 Data Frame Description

The format in which the data is stored in the different recording units on-board the A310-325 aircraft is
specified by the ARINC 717 protocol. The stream of data transmitted to these units is organized in a
frame (or dataframe). Each frame corresponds to four seconds of recorded flight data and represents
information about the parameters being sampled with time.
A frame is divided into four subframes and each takes one second to be recorded [16]. The subframes
n+6
are composed by 64, 128, 256, 512 or 1024 12-bit words (2 , with n from 0 to 4). The words in
ARINC 429 format are converted to the ARINC 717 12-bit format by aircraft data acquisition units. If
the number of recorded bits from a parameter exceeds the size of the word, more words are assigned.
Figure 2.6 illustrates the relationship between the structures of a frame.

4 seconds
1 second

SUBFRAME 1 SUBFRAME 2 SUBFRAME 3 SUBFRAME 4

WORD 1 WORD 1 WORD 1 … WORD n+6

1 2 3 4 5 6 7 8 9 10 11 12 12 bits

Figure 2.6 Frame, Subframe and Word

The bits from each word are numbered from one to twelve, as seen in figure above. Bit one
corresponds to the Least Significant Bit (LSB) and is transmitted first, while bit 12 is the Most
Significant Bit (MSB) and the last to be transmitted. By having information about the subframe(s),
word(s) and bit(s) it is possible to locate the parameter recordings in the frame. This information,
together with the parameter ARINC 429 characteristics and the recorded bits, provides the necessary
tools to compute its value from the recorded flight data.
Generally, there is a recorded parameter that consists of a frame counter, which is incremented every
four seconds. It is used by ground processing tools to detect lost data or recognize synchronization
problems. The first word of each subframe has a constant value and is used for synchronization.

2.2.2.2 Regular Frame

Most parameters are recorded at least once every four seconds, i.e., in every frame. These
parameters compose the regular frame. The recording rate of a regular frame parameter is directly
related with the number of times that parameter appears in the dataframe. If the parameter is recorded

11
once per second (1 Hz), it appears once per subframe and a total of four times per frame. The
parameters can be output at lower rates of 1/4 Hz or 1/2 Hz, and appear in one or two subframes,
respectively. When a parameter is recorded at rates higher than 1 Hz, there are various samples per
subframe and each one is called an instance of the parameter. Table 2.9 contains some examples of
recorded parameters in the A310-325 aircraft that help to clarify the previous concepts.
Parameter Name Recording Rate Samples per Frame Subframes
Radio Height #1 1 Hz 4 1,2,3,4
Vertical Acceleration 16 Hz 64 1,2,3,4
Flaps Position 1/2 Hz 2 2,4
Selected Vertical Speed 1/4 Hz 1 4
Pitch Angle 4 Hz 16 1,2,3,4
Table 2.9 Examples of Regular Parameters

The number of samples per frame is simply obtained from multiplying the recording rate by the
number of subframes. Higher recording rates are used for parameters that tend to vary more rapidly,
such as the Pitch Angle or the Normal acceleration, while parameters that experience slow variations
are recorded less frequently, such as the Selected Vertical Speed.

2.2.2.3 Superframe
The superframe parameters have a recording rate of 1/64 Hz and hence they are recorded once every
64 seconds and appear once every 16 frames. Superframe parameters are recorded on dedicated
words within a subframe that are named as “superframes”. The four most significant bits (12 to 9) of
the superframe contain a counter that is incremented at the end of each frame, i.e., every four
seconds, and ranges from 0 to 15. This counter indicates the parameter or parameters that are being
recorded.
Generally, the superframe parameters contain information that doesn’t change throughout a flight or
that doesn’t change as frequently as the accelerations, speeds, etc. By using superframes, one can
put up to 16 of these parameters together in a word slot and free vital space for the recording of other
parameters if needed. Some of the superframe parameters recorded in the frame of the A310 include
information about the Hour, Day, Month, Flight Number, Gross Weight, etc.

2.3 ATA Chapters

The Air Transport Association of America (ATA) is America’s oldest and largest airline trade
association. It was founded in 1936 and has since been renamed as Airlines for America (A4A). This
organization has played a major role in many decisions regarding aviation, including the Airline
Deregulation Act in the United States or the creation of the air traffic control system [17]. ATA was
responsible for the creation and release of important standards, developed with the purpose of
providing a specification for developing technical airline documentation. These standards have
enabled both airlines and suppliers to increase their levels of efficiency [18].
ATA Spec 100 is one of the most widely used standards for commercial aircraft documentation and
the latest version was released in 1999. The Spec 100 contains format and content guidelines for
technical manuals written by aviation manufacturers and suppliers and is used by airlines and other
segments of the industry for the maintenance of their products. The Spec 100 provides the aviation

12
industry with a standard for aircraft systems numbering, also referred as ATA system or chapter
numbers [19]. This standard system facilitates the learning and understanding for pilots, aircraft
maintenance technicians and engineers alike [20].

2.4 ATA 31 Recording Systems: Aircraft Integrated Data System (AIDS)

The aircraft recording systems shall be configured to comply with the mandatory flight data recording
requirements. They should acquire and record the mandatory parameters as received from the
aircraft’s systems. The list of mandatory parameters is prepared by the civil aviation authorities, on a
country basis [16].
The main function of the Aircraft Integrated Data System (AIDS) of the Airbus A310 is to convert the
various critical (including the mandatory) parameters into a recordable format and to record them on a
Flight Data Recorder (FDR) [21]. The system can be expanded to include data processing units to
monitor the condition of the connected aircraft systems. The diagram in figure 2.7 displays the
architecture of the A310’s AIDS. The Basic AIDS fulfills the mandatory recording requirements and is
composed by the following equipment:
 A Digital Flight Data Acquisition Unit (DFDAU);
 A Digital Flight Data Recorder (DFDR);
 A three-axis linear accelerometer (L.A.);
 A Control Panel (C.P.).
The Expanded AIDS is obtained by adding the following equipment:
 A Data Management Unit (DMU);
 A Digital AIDS Recorder (DAR);
 An Interactive Display Unit (IDU);
 A Printer.

ARINC 717 (573)

ARINC 429 Buses DFDR
ARINC 717 (573)

Analog
Parameters ARINC 717 (573) QAR
DFDAU
Discrete
Parameters
ARINC 717 (573) DAR
ARINC 429 DMU
L.A.
Printer

C.P. IDU

: Optional : Basic AIDS : Additional (Expanded AIDS)

Figure 2.7 AIDS Architecture

13
The DFDAU acquires, conditions and processes the aircraft parameters coming from the several
sensors and on-board computers [22]. These parameters are of various types: analog, discrete and
DITS (ARINC 429). The analog inputs are converted to digital values. The DFDAU then transmits the
required parameters to the DFDR in the format specified by the ARINC 717/573 protocol at a rate of
64 12-bit words per second (768 bits/sec). The DFDR or the “black box”, as often called by the media,
is a device designed to record the mandatory flight parameters as defined by the applicable civil
aviation authorities and to withstand the conditions likely to be encountered in an aircraft crash. The
DFDR has the ability to store data collected during up to 25 hours of flight and when it reaches its
maximum capacity the oldest data is overwritten by the new data. Therefore, the DFDR is not the most
recommendable device to sustain an efficient and routine FDM process.
The DFDAU sends a copy of the DFDR data frame to the QAR using one pair of bi-polar data lines.
The Airbus A310-325 from which the data will be decoded and analyzed is fitted with a Basic AIDS
and an Optical Quick Access Recorder (OQAR), where the flight data is recorded on a magneto-
optical disk. Because the QAR is located in the avionics compartment and not in the rear of the aircraft
close to the tail cone like the DFDR, it is primarily used by maintenance personnel to acquire the
mandatory parameter recordings. Figure 2.8 shows the physical location of the different AIDS
equipment in the aircraft, including the DFDR and QAR recorders. The avionics compartment is
located in the lowerfuselage forward of the nose landing gear.

DFDAU,DMU L.A.

QAR/DAR

DFDR

Avionics Compartment

Figure 2.8 Location of AIDS equipment in the Airbus A310

The DMU is part of the Expanded AIDS. It receives digital parameters from the DFDAU through a pair
of ARINC 429 data lines and processes the data for a variety of functions, including Condition
Monitoring (CM). In later chapters, we will cover how the DMU assists the operator with this type of
maintenance and how the processed information can be retrieved. When fitted in the aircraft, the DMU
also produces a stream of data for recording in a DAR. Most QAR units can be used as DARs, in
which case the recorded parameters are not limited to the mandatory parameters.

14
3. Dataframe Programming

This Chapter presents the different stages involved in the process of programming a database in AGS
with the information required for a proper analysis of the data recorded in the A310’s OQAR. The
results of this implementation for different recorded parameters will be subject to a further scrutiny in
the next Chapter.

3.1 Flight Data Department at TAP

The Flight Data (FD) Department at TAP Portugal is part of the Safety Management Department. The
FD department is responsible for the creation and continuous execution of the airline’s FDM
processes. The department is provided with the necessary tools for analyzing the flight data that is
stored on the recording units from the airline’s fleet. Instead of relying solely on the “snapshot”
information that is sent by the data processing computers, the complete flights are processed in order
to generate useful information to monitor safety and risk of the whole operation as part of a Safety
Management System (SMS) program. Having this into account, its members work closely with several
airline’s departments – Maintenance and Engineering, Pilots and Fleets, Technical Support, etc. – in
order to enable the airline to increase its safety levels and reduce its maintenance and operating costs
[4]. The flight data reading and processing tasks are carried at computer ground stations by a flight
data analysis tool: the Analysis Ground Station (AGS) from SAGEM.

3.1.1 AGS: A tool for flight data analysis

The AGS is a computer system for flight data analysis whose primary function involves processing all
the available data coming from the aircraft recorders. The AGS is capable of decoding the data frames
recorded in the ARINC 717/573 format containing the flight parameters. In order to do this, this tool
needs to be programmed by inserting information regarding the location of the parameters in the data
frame and their ARINC 429 characteristics.
The data frames are typically different according to the A/C type, its registration and, of course, the
recording device used. At TAP, there has been a continuous effort towards the standardization of the
FDM processes across the different fleets of aircraft. This includes the information recorded in the
data frames and also the type of devices used. The flight data from TAP’s aircraft comes from DAR
type recorders, which have high recording capacities (typically around 512 12-bit words per second).
Because the DAR receives data from a DMU, a customized data frame can be built by configuring the
latter. These two features allow a vast number of parameters to be recorded, which are not limited to
the mandatory parameters. The information is then used to support several FDM activities.

3.1.1.1 Database Version

AGS works on a Database basis, which is the configuration used to perform the flight data analysis for
a particular aircraft or fleet of aircraft. The programming of a Database is divided in two parts [23]:
 The Decoding Frame for parameter conversion in Engineering values, containing the
following components:
o the Equipment List;

15
 o the Parameter List;
o the Dataframe;
o the Procedures for Additional Parameters computation - these procedures
compute additional parameters (called computed on ground parameters) from the
recorded parameters for a more consistent analysis, performing parameter filtering
and corrections or other parameter computations to improve the flight analysis;
 and the Procedures for Automatic Flight Analysis containing the following components:
o the Procedures for Flight Operations: they are within the scope of the FOQA
program and include studies to identify and correct deficiencies during flight
operation with the aim of reducing safety risks;
o the Procedures for Maintenance: these procedures are used in the MOQA
program to perform computations for maintenance analysis and for monitoring
functions as required by the airline’s Maintenance departments.
After programming the Decoding Frame the AGS has enough information to convert the recorded
parameter data in binary format to its corresponding engineering values. The procedures are a set of
instructions, test conditions and computations programmed in AGS by SAGEM [24]. These
procedures use a dedicated simple programming language, yet powerful, composed of instructions
and that can be mixed with the standard “C” language. For instance, the procedures for additional
parameters (or additional procedures) ensure that each individual flight is separated and correctly
identified by its flight number, etc. These procedures compute a lot of information that is important for
the correct storage of the flights and used for an accurate representation of the flight parameters
throughout the flight, including the several flight phases. The procedures for Flight
Operation/Maintenance are executed after the procedures for additional parameters.

3.1.1.2 Automatic Analysis Process and the AGS cycle

When the Database version is fully configured, it is used in the Automatic analysis. This is the process
of decoding all flight data recordings contained in the data storage devices as defined by the operator.
The treatments done during the Automatic analysis include [23]:
 the Flight Identification in terms of date, flight number, origin and destination, according to the
additional procedures programmed in the database;
 the conversion of recorded parameters into engineering values;
 the computation of additional parameters;
 the storage in the database of the analyzed flights with the raw data (binary data) and
associated results, for statistical analysis or later replay of the flights;
 and the generation of reports and storage of data in external files.
The Automatic Analysis is one the major tasks performed by the FD Department and is done on a
daily basis. A representation of its location within the AGS activities is represented in figure 3.1. The
figure shows the different ways of acquiring the flight data. In most of TAP’s aircraft, the flight data is
directly recorded to DAR hard-drives and then transmitted to the airline’s servers via wireless networks
when the aircraft is on the ground. In the Airbus A310-325 described in Chapter 2, the data is

16
recorded on optical disks and the process is not so efficient because it requires trained personnel to
remove and substitute the disks from the QAR. In addition, the data is not as readily available as when
it is transmitted wirelessly.

Figure 3.1 AGS activities [23]

3.1.2 Database version for the A310-325 aircraft

A new database version was implemented in AGS containing the necessary information to decode the
flight data recorded in the QAR of the Airbus A310-325. This database version is identified with the
number 100791 and contains a data frame in the format specified by the ARINC 717/573 protocol.
Each subframe is composed of 64 words with 12 bits each.
A total of 219 parameters from various signal types were implemented: 172 coming from the digital
buses of the aircraft in the ARINC 429 format, 37 discrete and 10 analog parameters. The complete
list of parameters is in Appendix B. All these parameters are mandatory, since the QAR stores a copy
of the data recorded in the DFDR. Therefore, the DFDR Word and Frame Assignment [21] was used
to allocate the positions of the parameters in the dataframe. It consists of one data frame with the
description of the parameters in each word slot and information about the ARINC 429 input data bits.

1
This is an internal TAP numbering and there is no relation with similar numberings that may exist on
other AGS systems.
17
3.2 Aircraft Definition
The addition of the aircraft to the A/C list is mandatory since all the flights analyzed are linked to the
A/C registration (marked on the tail of the aircraft). AGS will use this definition to know how it will
access the recording support device and to know the analysis configuration based on the Database
version. The information required for the A/C definition is displayed in figure 3.2. It includes the A/C
Tail registration, the airline and the A/C type.

Figure 3.2 Dialogue window for New Aircraft

The upper right area allows defining the characteristics of the recorders installed on-board the aircraft.
A QAR or DAR recorder is classified as a Maintenance type recorder, whereas the DFDR is a Crash
type recorder, due to its characteristics. The access to the A310-325’s QAR media (what actually
holds the data in the storage device) is direct because the AGS reads the data directly from the optical
disks. The serial number and installation date of the engines are not mandatory but may useful for an
ECM analysis.

3.3 Equipment Definition

The DFDAU is connected to various sensors and data sources to collect the parameters, so it
becomes fundamental that these data sources are defined in AGS. Appendix A provides a list of the
ARINC 429 data buses added to version 10079. The equipment ID is an hexadecimal value that
comes from the ARINC 429 specification [14]. For the analog and discrete type parameters the
information related with the data source equipment is defined as EQUIP.

3.4 Parameter Implementation

This section concerns the implementation of the parameters that are recorded in the dataframe. This
is required before the parameters are allocated to their respective word slots. The information to be
inserted requires the knowledge of the format of the data words from the ARINC 429 protocol, which
was described in the previous Chapter.

3.4.1 Parameter Definition

The parameter information that needs to be inserted in AGS comprises the following:

18
  Parameter Name: name used for data visualization after decoding and the name used in the
procedures programming;
 Description: description associated with the parameter. It can be filled with manufacturer
specification or the description available in AGS;
 Reference: the number in the manufacturer specification (AMM or other);
 Parameter Type: defines the data type of the parameter (BNR LINEAR (A*X), BCD,
DISCRET, etc.);
 ATA code: this code classifies the parameters according to the standardized groups defined
in ATA’s Spec 100 Code, allowing grouping of parameters from the same ATA chapter;
 Source Identification: it comprehends the Equipment, Label and SDI;
 Input Raw Data: for parameters in the ARINC 429 format, this field defines the Sign bit
position, the MSB bit position and the number of significant data bits. For an ARINC 429
discrete parameter, the definition of the input bit and the status of the Zero and One States
provides enough information;
 Resolution/Range: the Resolution of the LSB is the smallest possible value transmitted to the
acquisition unit. The Range is automatically calculated from the Resolution and the number
of data bits or the Resolution can be calculated if the Range is introduced first;
 Display Format: information about the Engineering Units and the format in which the data is
displayed after being decoded;
 Operational Range: defines lower and upper operating limits for the parameter that can be
used to signal abnormal situations when the data is visualized.
Figure 3.3 shows the dialogue window for the implementation of a BNR LINEAR parameter, in this
case the Computed Airspeed. The fields that were described above are indicated. Examples of the
implementation of different types of parameters are given next.
Description Reference

Parameter Name
Parameter Type

Source
Identification

Input Raw Data

Display Format
Resolution of LSB

Range

Operational Range

Figure 3.3 Dialogue window for Parameter Implementation

19
3.4.2 Examples

BCD: Flight Number

The flight number comes from the Flight Management Computer (FMC) and is encoded in BCD
format. It is composed by four digits, each decoded from a separate group of four data bits. For each
digit it is necessary to specify the MSB. The digits are put together via an additional procedure.

Figure 3.4 Flight Number Implementation

PACKED BITS: AFCS Longitudinal Modes

The PACKED BITS is a type of parameter in which a determined number of data bits have a specific
meaning when combined. It differs from the data types presented for the ARINC 429 protocol. Figure
3.5 shows the implementation of a parameter that indicates the longitudinal modes of the Automatic
Flight Control System (AFCS). Bits 21 to 18 are decoded to their corresponding decimal value (or raw
value in AGS) and each value in the 1 to 15 range indicates if a certain mode is active.

Figure 3.5 AFCS Longitudinal Modes Implementation

20
BNR COUPLE OF VALUES: Flaps Position
There are some parameters that are not acquired from the ARINC 429 data sources and need to be
converted to digital values. An example of such parameter is the Flaps Position, which is acquired in
the DFDAU as an analog parameter. It was implemented as a BNR COUPLE OF VALUES parameter.
This type of parameter is used when a calibration law is obtained from couples of values. These
values are in table format in figure 3.6a. A graphic representation of the flaps angle as a function of
the recorded raw value is in figure 3.6b.
Flaps Position [Deg]

DFDAU Raw Value

(a) (b)
Figure 3.6 Flaps position calibration law: (a) table format [21]; (b) graphic display.

3.5 Dataframe Construction

After implementing all the parameters in the database, it is necessary to create the dataframe
structure by specifying where each parameter is recorded. The majority of the parameters are
recorded in the regular frame. The dataframe from version 10079 also accomodates 4 superframes.

3.5.1 Regular Parameters

The regular frame parameters can be added by means of a dialogue window similar to the one
displayed in figure 3.7. The dataframe is represented in the bottom part of the window – word, bits and
subframes - and allows the user to visualize and select the location of the parameter in the word slots.
Each parameter can contain up to three parts, which need to be allocated separately. It is a common
practice to use one of these for the sign bit, if applicable. The three parts are only used for parameters
recorded with more than 12 bits. AGS will use the ARINC 429 data bits information for each part to
compute the correct parameter values.
The recording rate must be defined according to the number of times that the parameter/parameter
part appears in the dataframe. For parameters with several instances in each subframe, it is only
necessary to select the location of one of them. AGS will automatically assign the location of the
remaining instances from the fact that they are equally spaced in the dataframe. Different parts from a
parameter can have different recording rates. For instance, the sign bit and the most significant bits
(27-24) of the Standard Altitude (ALT_STD) are recorded at 1/4 Hz (subframe 4, Word 57) and the
least significant bits (24-14) are recorded at 1 Hz (subframes 1 to 4, Word 58).

21
 1st Part 2nd Part 3rd Part

Parameter Name
Description

Recording Rate Input Raw Data

Word Bits

Subframe

Figure 3.7 Regular parameter: Pitch Angle (PITCH)

3.5.2 Superframe Parameters

To decode the superframe parameters it is first necessary to define the location of the superframe
words. Table 3.1 specifies the location of the four superframes recorded in the frame. Superframes 1
to 3 are presented in Appendix C, containing the description and name of the recorded parameters -
Superframe #4 contains information about the operating status of the DFDAU. Bits 12-9 of each
superframe are occupied with a counter that is updated every four seconds.
Superframe Nr. Subframe Word
1 1 56
2 1 57
3 4 41
4 3 51
Table 3.1 Recorded Superframes

The dialogue window for the allocation of the superframe parameters in the memory is very similar to
that in figure 3.7, for a regular parameter. The superframe structure is displayed and the user can
select the position of the parameter in any of the 16 cycles. Because only bits 8-1 are available for
output, many of the superframe parameters occupy more than a cycle. In total, 31 of these parameters
were added to the dataframe and they all have a period of 1/64 Hz, which means that they are only
updated every 64 seconds. Some parameters, such as the latitude and longitude, have parts recorded
in both the regular frame and superframe.

22
 Parameter Name

Period

Subframe & Word

Input Raw Data Bits

Frame/Cycle

Figure 3.8 Superframe parameter: Gross Weight (GW)

3.6 Database Validation

Before all the tasks inherent to the automatic analysis can be conducted, the database version needs
to be validated. The programming of a database is a process subject to a variety of user-induced
errors because of all the new information that is inserted. AGS provides a validation function that is
able to detect some of these errors. To ensure the consistency of the database, this validation was run
several times during the implementation of the parameters and the construction of the dataframe.
The validation is a very useful procedure to detect inconsistencies between the ARINC 429 data bits
and the bits that are recorded. However, it covers all the components of a database version, including
the additional procedures and the procedures for automatic flight analysis and it is a good practice to
run a validation whenever changes are made or new procedures added.
To ensure that the new database version doesn’t actually contain errors, further validation has to be
conducted by analysing the flight data decoding results. This is the main subject of the next Chapter.

23
4. Decoding Results and Flight Phase Computation
Procedure

The first part of this Chapter presents the decoding results that were obtained after the flights were
processed with the newly programmed database. The implementation of the Flight Phase computation
procedure is described afterwards and relies on a collection of both recorded and computed on ground
parameters.

4.1 Results for Recorded Parameters

After programming the dataframe in database version 10079, the flight data from several flights
performed by the A310-325 aircraft was decoded and analyzed. This analysis consisted on the
observation and interpretation of data from the recorded parameters as a function of time. The time is
computed from the parameters that give the current Greenwich Mean Time (GMT) throughout the
flight.
There were several types of errors found during the analysis that were not detected by the internal
AGS validation procedures. The more typical were user induced, such as incorrect location of the
parameter in the frame or incorrect definition of the parameter range or resolution, and were promptly
solved. Other errors were attributed to improper information supplied in the documentation.
The objective of this section is to present the decoding results for several parameters that are
recorded in the dataframe. Each sub-section shows parameter results from a particular ATA system
that were selected so as to provide a general insight on how these systems are operated throughout
the flight. The complete set of parameters displayed is in table 4.1. The data from these parameters
was exported from AGS to .csv files and then handled by graphics software.
Parameter
Description Parameter Type Equip/Label/SDI ATA
Name
ALT_STD Altitude standard BNR LINEAR FNSG-1/203/XX 34
IAS Computed Airspeed BNR LINEAR FNSG-1/206/XX 34
GS Ground Speed BNR LINEAR IRS-1/312/XX 34
PITCH Pitch Angle BNR LINEAR FNSG-1/325/XX 34
AOAL Angle of Attack (ADC-1) BNR LINEAR FNSG-1/221/XX 34
LONG Longitudinal Acceleration BNR SEGMENTS EQUIP/331/00 31
VRTG Normal Acceleration BNR SEGMENTS EQUIP/333/00 31
LATG Lateral Acceleration BNR SEGMENTS EQUIP/332/00 31
Landing Gear Squat Switch 32
LDG_ON_1 DISCRETE EQUIP/270/00
LH
ROLL Roll angle BNR LINEAR FNSG-1/325/XX 34
ATS_EGD Autothrottle Mode Engaged DISCRETE TCC-1/274/XX 22
MTHR Manual Throttle Armed DISCRETE TCC-1/274/XX 22
Autothrottle Thrust EPR 22
ATS_EPR DISCRETE TCC-1/274/XX
Mode
ATS_MACH Mach Mode Engaged DISCRETE TCC-1/274/XX 22
ATS_SPD Speed Mode Engaged DISCRETE TCC-1/274/XX 22
ATS_RTD Retard Mode Engaged DISCRETE TCC-1/274/XX 22
BNR COUPLE OF 27
FLAP Flap position (calibrated) EQUIP/137/XX
VALUES
BNR COUPLE OF 27
SLAT Slats position (calibrated) EQUIP/127/XX
VALUES
All Speed Aileron LH 27
AIL_1 BNR LINEAR SDAC/310/10
position

24
 Parameter
Description Parameter Type Equip/Label/SDI ATA
Name
All Speed Aileron RH 27
AIL_2 BNR LINEAR SDAC/310/01
position
FQTT Trim Tank fuel quantity BNR LINEAR FQI/263/00 28
CG Center of Gravity BNR LINEAR CGCC/076/00 31
GW Gross weight BNR LINEAR FMC-1/075/XX 34
N21 N2 Engine #1 BNR LINEAR EEC-1/344/XX 77
N22 N2 Engine #2 BNR LINEAR EEC-1/344/XX 77
EGT1 EGT Engine #1 BNR LINEAR EEC-1/345/XX 78
EGT2 EGT Engine #2 BNR LINEAR EEC-1/345/XX 78
EPR1 EPR Actual Engine #1 BNR LINEAR EEC-1/340/00 77
EPR2 EPR Actual Engine #2 BNR LINEAR EEC-1/340/00 77
FF1 Fuel Flow Engine #1 BNR LINEAR EQUIP/347/10 73
FF2 Fuel Flow Engine #1 BNR LINEAR EQUIP/347/01 73
Table 4.1 Recorded Parameters

4.1.1 Airframe Systems

4.1.1.1 Navigation (ATA 34)

Figure 4.1 displays the evolution of the pressure altitude or Standard Altitude (ALT_STD). This is the
altitude in the International Standard Atmosphere (ISA) [25] with the same pressure as the part of the
atmosphere in which the aircraft is flying. The ISA pressure setting used is 1013.25 hPa,
corresponding to the pressure at the reference pressure surface: the Mean Sea Level (MSL). The
altitude in the figure shows the typical profile on a commercial flight. During the flight, the aircraft
transitions to higher flight levels because the optimum cruise altitude increases as the aircraft weight
decreases (as fuel is burned) [26]. The initial cruise altitude is 31000 ft. (FL3701) and before starting
the descent it is 36000 ft. (FL360).

Figure 4.1 Pressure Altitude results

The profile shows some spikes during the climb and descent phases of the flight. This situation is
corrected via an additional procedure that calculates the Standard Corrected Altitude (ALT_STDC),
also represented in figure 4.1. The ALT_STDC parameter will be used to represent the altitude
throughout the remainder of this section, whenever considered relevant for the analysis.

1
FL stands for Flight Level
25
Figure 4.2 below shows the variation of the different speeds recorded in the frame: the Computed
Airspeed (CAS) and the Ground Speed (GS). In this dataframe, the True Airspeed (TAS) and the
Mach number are not recorded but are computed on ground.

Figure 4.2 Flight Speed results

The Flight Path Angle ( ), Pitch Angle ( ) and Angle of Attack ( ) are related by the following
expression:
(4.1)

The three attitude angles are represented in the vertical plane in Appendix D.
The evolution of the Pitch Angle (PITCH) and the Angle of Attack (AOAL) is depicted in figure 4.3. The
flight path angle can be computed on ground using equation 4.1. During most of the climb the Pitch
Angle is kept at a higher angle than the Angle of Attack, resulting in a positive Flight Path Angle and
an ascending trajectory. The opposite occurs during the descent. In cruise, the Pitch Angle and the
Angle of Attack have approximately the same value and during this phase the trajectory is horizontal.
The biggest difference between these two parameters occurs in the final moments of the flight before
Landing, when the Angle of Attack is much higher than the Pitch Angle.

Figure 4.3 Pitch and Angle of Attack results

26
4.1.1.2 Indicating/Recording Systems (ATA 31)
Figure 4.5 shows the evolution of the Longitudinal (LONG) and Normal Accelerations (VRTG) during
take-off and climb until the cruise altitude is reached. The Left Main Landing Gear Squat Switch
(LDG_ON_1) is a discrete that assumes the value 1 on the ground and 0 in flight and is plotted
together with the two accelerations to help visualize the take-off moment. The axes of the aircraft used
for the accelerations and the sign convention for the Roll angle are shown in figure 4.4.

(a) (b)
Figure 4.4 (a) Axes of the aircraft (b) Positive roll angle [27]

As expected, the Normal Acceleration remains close to 1 during most of the time with the exception of
the moments after the aircraft takes off. The maximum Longitudinal Acceleration is registered during
the take-off roll and initial climb stages, corresponding to the use of take-off thrust, and then it keeps
decreasing as the aircraft is accelerated to its cruising speed. During cruise, both accelerations
approach the 1 g and 0 g values, respectively.

Figure 4.5 Results for the Longitudinal and Normal accelerations

Figure 4.6 displays the Lateral Acceleration (LATG) throughout the complete flight, together with the
Roll Angle (ROLL). The coupling between these two parameters is evident: the biggest variations in
the Lateral Acceleration occur at high Roll Angle values. A negative roll angle induces a positive
lateral acceleration and vice-versa. This is in accordance with the definitions in figures 4.4a and 4.4b.

27
 Figure 4.6 Lateral acceleration and Roll angle results

4.1.1.3 Auto Flight (ATA 22)

Figure 4.7 shows the status of the Auto Throttle Mode (ATS_EGD) and Manual Throttle (MTHR)
discretes for the entire flight. It is observed that during most of the time that the aircraft is airborne the
auto throttle is engaged. The manual throttle is armed on the ground when the thrust is controlled
manually.

Figure 4.7 Results for the Auto Throttle and Manual Throttle discretes

The Auto Thrust (A/THR) can be engaged in three modes with different functions [11]:
1. Thrust (THR): Acquires and maintains the selected thrust limit or thrust target;
2. Speed/Mach (SPD/MACH): Acquires and maintains the selected speed or Mach Number;
3. Retard (RETARD): Retard the throttles to idle during initial descent.
The engaging of the A/THR modes is shown in figure 4.8. During the climb, a combination between
the THR and the MACH modes are used. After the aircraft reaches the cruise altitude, the SPD mode
is engaged and maintained throughout this flight phase. The RETARD mode is engaged during initial
descent as expected. The SPD mode is again used during the approach and landing stages.

28
 Figure 4.8 Results for the Auto Throttle mode discretes

4.1.1.4 Flight Controls (ATA 27)

The control of the aircraft is achieved by the primary flight controls (e.g.: elevator, ailerons, rudder),
which provide pitch, roll and yaw control and the secondary flight controls (e.g.: flaps and slats) that
consist of lift and drag augmenting devices [11]. The primary flight control surfaces are mechanically
controlled, while the secondary flight controls are electrically controlled by dedicated computers. All
flight control surfaces are hydraulically actuated. Appendix D provides a table that lists both the
primary and secondary flight control surfaces and a picture representing these surfaces on the A310-
325 aircraft.
Table 4.2 shows the position of all the Slat and Flap controls [28], as selected by the Flap and Control
Lever in the cockpit, and the state of the Krueger Flap1.
POSITION SLATS SLATS FLAPS FLAPS KRUEGER
SEQUENCE (OUTER,CENTER) (INNER) (OUTER) (INNER) FLAP
1 (0/0) 0 0 0 0 RETRACTED
2 (15/0) 17 16 0 0 EXTENDED
3 (15/15) 17 16 8 15 EXTENDED
4 (20/20) 21 20 16 22.1 EXTENDED
5 (30/40) 25.4 24 31.5 41 EXTENDED
( ): NUMBER INDICATED ON THE SLAT AND FLAP POSITION INDICATOR
Table 4.2 Flap and Slat Control Lever Positions

The parameter SLAT records the (Outer and Center) Slats Position and the FLAP parameter the
(Inner) Flaps Position. The two recorded parameters experience sudden changes and therefore the
corrected parameters SLAT1C and FLAP1C are presented in figure 4.9. The corrections to the
recorded parameters were made via an additional procedure with a function that eliminates the
observed oscillations. The usage of these secondary flight control surfaces is depicted only during the
approach and landing.

1
Lift/Drag augmenting device located on the leading edge of the wing
29
 Figure 4.9 Flaps and Slats position during approach and landing

Figure 4.10 presents the actuation of both ailerons during the early stages of the flight until an altitude
of approximately 9000 ft. is reached. It is observed that the values are symmetrical between each
other, allowing for quicker maneuvers. The axis of symmetry is 0º when the aircraft is operating with a
clean configuration (i.e. no Flaps or Slats extended) and 7º when the Slats are extended to 15º (as
indicated in the Control Lever – table 4.2). This is because the aircraft is equipped with a system that
drops the ailerons to improve its aerodynamic characteristics [11].
When an aileron is deflected up it decreases the lift on the related wing and when deflected down it
increases its lift. In figure 4.10 at around 5000 ft the Left Aileron (AIL_1) goes up and the Right Aileron
(AIL_2) goes down, causing the left wing to go down and the right wing to go up and inducing a
negative roll angle.

Figure 4.10 Results for the Ailerons and Slats position and Roll angle

4.1.1.5 Fuel (ATA 28)

The Fuel System includes five wing tanks and a tail trim tank. The trim tank has a capacity of 10848
lbs. (~4931 kg) and is installed in order to increase the fuel tank capacity and decrease the fuel
consumption by reducing the drag during cruise, by maintaining the Center of Gravity (CG) of the
aircraft close to the certified aft limit [11].

30
The fuel quantity in the trim tank (FQTT) is displayed in figure 4.11. The aircraft CG (CG) is also
represented and is given in percentage of the Mean Aerodynamic Chord or %MAC – see Appendix E
for a representation of the MAC reference line and considerations about the calculation of %MAC.
Both parameters are recorded in the superframe and computed by the Center of Gravity Control
Computer (CGCC). During the cruise of the selected flight the fuel quantity in the trim tank increases
to almost 6500 lbs., before stabilizing around 6200 lbs. The CG is moved aft of the aircraft to almost
35%MAC as a result of this fuel transfer.

Figure 4.11 Trim Tank Fuel Quantity and CG results

The Gross Weight (GW) is also computed by the CGCC and figure 4.12 depicts its evolution. The
aircraft starts the selected flight with a weight of 117549 kg and ends with 108641 kg.

Figure 4.12 Gross Weight results

4.1.2 Powerplant
Figures 4.13a to 4.13b show the evolution of some powerplant related parameters recorded in the
dataframe. These include: the high-pressure rotor speed (N2), Engine Pressure Ratio (EPR), Exhaust
Gas Temperature (EGT) and Fuel Flow (FF). The low-pressure rotor speed (N1) is not recorded. A
description of these parameters is given in the next chapter.
The N2 speeds in figure 4.13a reach a maximum of about 95% of the defined nominal RPM values.
This maximum is achieved in the beginning of the climb. The speeds then diminish until reaching the

31
cruise altitude, where they stabilize at around 86-87% of the nominal value. Figure 4.13b shows the
evolution of the EGT. The profile is similar to the N2 parameters with a peak EGT just over 500ºC
during takeoff. Figure 4.13c shows the EPR for the same flight and a maximum of about 1.55 at the
top of climb. The EPR values for the take-off are situated at a lower value of around 1.35, which
indicates that probably a reduced thrust setting was used. The Fuel Flow (FF) for both engines is
displayed in figure 4.13d. The Fuel Flow reaches values of almost 8000 kg/hour during take-off and
the fuel consumption per engine during cruise is of about 2000 kg/hour.

(a) (b)

(c) (d)
Figure 4.13 Powerplant parameters results: (a) N2; (b) EGT; (c) EPR; (d) FF.

4.2 Additional Procedures: Flight Phase Computation

In the previous section, the results for some recorded parameters were presented and the accuracy
and quality of the data was checked. Following this, the Procedures for Additional Parameters have
been programmed in the current database version. The primary function of these procedures is to
process the recorded data while the frames are being decoded and calculate computed on ground
parameters that improve the flight analysis. The procedures are executed according to their number
and defined rate and can be used as subroutines to be called whenever needed. Inside each
procedure, it is possible to use the value the parameter held seconds before by using the suffix .

32
The Flight Phase Computation procedure is of extreme importance for the flight analysis. This
procedure computes the Flight Phase based on several recorded and computed on ground
parameters. The Flight Phase parameter (FLIGHT_PHASE) is used in a variety of maintenance and
flight operations related procedures to monitor the aircraft and systems operation during specific
stages of the flights. It is also mandatory for the correct separation of each flight contained in a media.

4.2.1 Flight Phases Definition

4.2.1.1 Airbus
The aircraft systems compute the flight phase according to an Airbus definition [21]. There are 12 flight
phases in the A310-325, which are used to define the flight-phase-related inhibition of the Electronic
Centralized Aircraft Monitoring (ECAM) [11]. The starting and ending conditions for each of the flight
phases are described in table 4.3 and the related flight profile is displayed in Appendix F.
Flight Phase Nr. Start Condition End Condition
1 Electrical Power ON 2nd Engine Start
2 2nd Engine Start 1st Engine TO Power Application
st
3 1 Engine TO Power Application Speed>70 Kt
4 Speed>70 Kt Lift Off
Lift Off+1 Minute or Height >400 Ft
5 Lift Off
(Whatever if first achieved)
Lift Off+1 Minute or Height>400 Ft
6 Height>1000 Ft
(Whatever if first achieved)
7 Height>1000 Ft Height<1000 Ft
8 Height<1000 Ft Height<400 Ft
9 Height<400 Ft Touch Down
10 Touch Down Speed<70 Kt
11 Speed<70 Kt 1st Engine Shut Down
12 1st Engine Shut Down 5 minutes after 1st Engine Shut Down
Table 4.3 Airbus A310-325 Flight Phases Definition

This convention for the flight phases is mainly intended for system’s synchronization within the aircraft
and does not provide sufficient separation of the flight phases to suit the intended flight data analysis.
For instance, flight phase 7 includes the cruise and most of the climb and descent phases.

4.2.1.2 SAGEM
As mentioned above, the Airbus definition for the flight phases used in the aircraft is very limited.
Having that in mind, the procedure was implemented based on the definition from SAGEM that
includes a total of 14 flight phases [24]. The phases are numbered from 1 to 14 and ordered according
to a logical sequence from the beginning of the flight. Two of the flight phases in table 4.4 are not
common to occur during normal operation: Rejected Take-Off and Go-Around. For example, the first
will occur in case an engine parameter exceedance is detected during take-off and the second can
occur when strong winds prevent the crew from landing the aircraft.
Nr. Phase FLIGHT_PHASE
1 Engine Stopped ENGINE_STOP
2 Taxi-Out TAXI_OUT
3 Take-Off TAKE_OFF
4 Rejected Take-Off REJECTED_TO
5 2nd Segment SND_SEGMENT
6 Initial Climb INITIAL_CLIMB
7 Climb CLIMB
8 Cruise CRUISE

33
 9 Descent DESCENT
10 Approach APPROACH
11 Final Approach FINAL_APPROACH
12 Landing LANDING
13 Go-Around GO_AROUND
14 Taxi-In TAXI_IN
Table 4.4 SAGEM Flight Phases

The Flight Profile from SAGEM is also available in Appendix F, together with the conditions that
characterize the transitions between the respective flight phases.

4.2.2 Implementation
The conditions used in the definition by SAGEM can be modified and adjusted according to the aircraft
type and the operator needs. The N2 is the engine thrust parameter used by SAGEM to detect the
transition to the Take-Off phase. It was first implemented in version 10079 but it was observed that in
many flights the aircraft transited directly from the TAXI_OUT to the CLIMB phases. This was because
the N2, recorded in the superframe, was not sampled during some take-off runs. Figure 4.14 contains
data from the N2 and EPR parameters the take-off run and initial climb stages, in order to illustrate this
situation. Again, the Left Main Landing Gear Squat Switch (LDG_ON_1) is used to indicate the
moment in which lift-off is detected. The EPR, recorded at a rate of 1 Hz, smoothly increases during
take-off to 1.35, while the N2 speeds jump from about 60% to 95% after the aircraft is airborne. This
jump occurs because the high rotor speeds are only updated every 64 seconds.

Figure 4.14 N2 and EPR results during the take-off stages

As a consequence, lift-off was not detected because it requires a flight phase equal to TAKE_OFF or
GO_AROUND. The decision was made to use the EPR, instead of the N2, to detect the Take-Off. The
TAKE_OFF and first stages of the climb were then correctly detected in all the flights analyzed. The
N2 recordings continued to be used in the remaining flight phases.
When the Flight Phase from the aircraft is recorded as a parameter, it can be used in conjunction with
the other parameters to enable the transitions between different flight phases. Although this parameter
is used by the airborne computer systems, it is not recorded in the QAR and therefore is not available
to compute the flight phase.
Table 4.5 lists the transitions between the flight phases and the associated conditions that were
implemented in the procedure from database version 10079. The transitions to the Rejected Take-Off
and Go-Around flight phases are categorized as abnormal and are not included in the table.
34
 Previous Phase Next Phase Conditions for Transition
During 10 seconds:
- Engines not stopped:
o N2 from Engines #1 and #2 > 55% or
ENGINE_STOP TAXI_OUT o Fuel Flow from Engines #1 and #2 > 50 Kg/h
for 2 seconds
- Left Main Landing Gear compressed
Previous Phase < 2nd Segment and Engines not at TO Power
EPR from Engine #1 or #2 > 1.2
Longitudinal Acceleration > 0.05 g
Ground Speed increase equal to or higher than 3 knots per
TAXI_OUT TAKE_OFF second during 2 seconds
Ground Speed > 25 knots
nd
Previous Phase < 2 Segment or Left Main Landing Gear
compressed
Height > 50 ft and Engines not stopped
Altitude < 8000 ft.
Configuration ≥ Take-Off Configuration
TAKE_OFF SND_SEGMENT
Pitch > 6º
EPR from Engine #1 or #2 > 1.2
Previous Phase ≠ than Final Approach and Landing
Height > 400 ft. or Configuration < Take-Off Configuration
SND_SEGMENT INITIAL_CLIMB
Previous Phase ≤ Initial Climb or > Landing
Clean Configuration
INITIAL_CLIMB CLIMB
Previous Phase < Cruise
Altitude > 10000 ft.
Clean Configuration
CLIMB CRUISE
Vertical Speed between -300 and 300 ft./second during at least 15
seconds
Clean Configuration
- Vertical Speed < -420 ft./second and Altitude > 8000 ft. during 5
CRUISE DESCENT
seconds
- Previous phase = Cruise or Descent
Height < 3000 ft.
DESCENT APPROACH Previous Phase ≥ and ≤ Final Approach or equal to Go Around
Configuration > 0 and < 4
Previous Phase ≥ Cruise
Configuration ≥ 4
APPROACH FINAL_APPROACH
Height < 1000 ft.
Pitch < 6º or Engine #1 or #2 not at TO Power
Height < 50 ft.
Ground Speed > 30 kt.
Engines Not Stopped
FINAL_APPROACH LANDING
Previous Phase > 2nd Segment
Configuration > 4
Engine #1 or #2 not at TO Power
Previous Phase ≥ Landing
Left and Right Main Landing Gear compressed
Engines Not Stopped
LANDING TAXI_IN
Touch Down detected (STEP_TD)
Ground Speed < 30 kt. or difference between current heading and
heading at Touch Down detection (DELTA_HEAD) > 10º
Height < 50 ft.
Engines Stopped:
- N2 from Engines #1 and #2 < 55% or
- Fuel Flow from Engines #1 and #2 < 50 Kg/h for 2
TAXI_IN ENGINE_STOP
seconds
Ground Speed < 3 kts
Left and Right Main Landing Gear compressed
nd
Previous Phase < 2 Segment or > Go Around
Table 4.5 Conditions for the normal transition between the flight phases (Database version 10079)

Most of the parameters used in the implementation are computed on ground parameters. The
following table lists all the parameters used and their corresponding procedures, if applicable.

35
 Parameter Name Description Procedure Number
LDGL* Landing gear switch left 80
LDGR* Landing gear switch right 80
ALT_STDC Altitude Standard corrected 81
GSC Ground speed corrected 105
EPR1C EPR Actual Engine #1 corrected 106
EPR2C EPR Actual Engine #1 corrected 107
N21C N2 Engine #1 corrected 108
N22C N2 Engine #2 corrected 108
FF1C Fuel Flow Engine #1 corrected 109
FF2C Fuel Flow Engine #2 corrected 109
CONF Landing & Take-off configuration 123
TLA1C Throttle resolver angle (TRA) Engine #1 124
TLA2C Throttle resolver angle (TRA) Engine #2 124
HEIGHT Height above runway 211
IVV Inertial vertical speed 200
LONGC Longitudinal acceleration 260
TTRANS** Flight phase authorized sequences 507
DELTA_HEAD Difference between touchdown heading and current heading 509
STEP_TD Touch down step 605
PITCH Pitch angle Recorded
Table 4.6 List of parameters used in the Flight phase computation procedure (*Same information as
LDG_ON_1 and LDG_ON_2 (0-AIR, 1-GROUND); **Array)

To be noticed that more than half of the listed parameters consist of corrected parameters which
contain the same information as the recorded parameters. The corrected parameters do not
experience large sudden changes that can occur during the recordings or when the data is decoded,
thereby preventing inadvertent transition between the flight phases.
The HEIGHT parameter computes the altitude above ground level (or AGL altitude). At low altitudes,
generally up to 2500 feet [5], it corresponds to the value given by the radio altimeter. Thus, the true
vertical distance to the ground is used. At higher altitudes the radio altimeter doesn’t give accurate
readings and, therefore, the HEIGHT parameter consists of an estimate that is computed from the
corrected standard altitude (ALT_STDC) and the altitude standard of the origin and destination
runways.
The inertial vertical speed (IVV) is not recorded and is thus computed from the variation in the altitude.
There are three types of transitions between the flight phases: Normal, Abnormal and Forbidden. The
TTRANS parameter uses the information in figure 4.15 to classify the different transitions and prevent
those that are forbidden from occurring.

Figure 4.15 Flight phase authorized sequences [24]

36
There are a total of five combinations between the flaps and slats positions that can be selected with
the Flap and Slat Control Lever (see table 4.2). These combinations are called configurations. The
SLAT1C and FLAP1C parameters, shown in section 4.1, are converted to the angular values assigned
to each configuration with a dedicated procedure that yields the SLATC and FLAPC parameters. The
CONF parameter is calculated afterwards from these two parameters using the logic in table 4.7.
POSITION SEQUENCE SLATC FLAPC CONF
1 (0/0) 0 0 0
2 (15/0) 15 0 1
3 (15/15) 15 15 2
4 (20/20) 20 20 3
5 (30/40) 30 40 4
Table 4.7 Flaps and Slats Configuration

Figure 4.16 shows the results obtained for the CONF parameter during the take-off and landing. It is
possible to see that, for this flight, configuration 2 was selected for Take-off and configuration 4 was
selected for Landing.

(a) (b)

Figure 4.16 Configuration results during (a) Take-Off and (b) Landing

Finally, figure 4.17 shows the evolution of the FLIGHT_PHASE parameter throughout a selected long-
haul flight. It is possible to observe that the transitions between the flight phases occur in the correct
order defined in table 4.4. The transitions to higher flights levels in cruise are indicated by the periods
where the FLIGHT_PHASE transitions from Cruise (8) to Climb (7) and then back to Cruise (8). The
aircraft transitions from Take-Off (3) to the 2nd Segment (5) and afterwards to the Initial Climb (6),
although this is not noticeable. The aircraft does not go through a Rejected Take-Off (4) or a Go-
Around (13), as expected in normal operation. Results similar to those in the figure below were
encountered for the other flights and demonstrate the correct implementation of the procedure and its
usefulness to the analysis.

37
Figure 4.17 Flight Phase Results

38
5. Engine Condition Monitoring

In this Chapter, the fundamentals of Engine Condition Monitoring (ECM) are introduced. The Chapter
starts by describing the role of ECM in the maintenance of aircraft engines and its contribution to the
condition-based maintenance philosophy, which allows the airlines to conduct maintenance work on
the engines based on their actual “health”, rather than following a more costly predetermined
schedule. Next, the systems used in the ECM process are presented. These have evolved from the
analysis and storage of data recorded manually by the flight crews to the monitoring and interpretation
of engine data in real-time by sophisticated software. After this, the nature of the parameters collected
for ECM purposes is discussed and the key performance parameters described. The Chapter finishes
with the description of the Parameter Trend Monitoring method, which is the most common technique
used to monitor the deterioration of the engine’s condition. Trend Monitoring provides assistance in
the diagnosis of engine failures and enables the determination of precursors to engine events. In
addition, it helps the operators to plan for engine removals more accurately. The removal intervals can
even be extended based on the rate of deterioration in the engine’s performance. All these activities
have the potential to contribute to lower maintenance costs.

5.1 The Condition-Based Maintenance Philosophy

Early airline maintenance programs were based on the philosophy that each part of the aircraft
required disassembly for inspection. Maintenance was carried out in the aircraft after a specified
number of hours in operation and the components/parts were removed and replaced by new or
overhauled components. However, the introduction of larger aircraft made this process almost
unbearable to airlines because of the rise in the costs associated with the replacement parts and the
time that was needed to perform all the maintenance tasks.
Over the last decades, aircraft maintenance has evolved according to the existing aviation safety
standards and the economic requirements from the air transport industry. This resulted in the
introduction of new maintenance processes with different philosophies. The Federal Aviation
Administration (FAA) recognizes three Primary Maintenance Processes [29], [30]:
 Hard-Time (HT): This is a preventive primary maintenance process. It requires that an
appliance or part be periodically overhauled in accordance with the carrier’s maintenance
manual or that it be removed from service.
 On-Condition Maintenance (OC): This is a preventive primary maintenance process. It
requires that an appliance or part be periodically inspected or checked against some
appropriate physical standard to determine whether it can continue in service. The purpose of
the standard is to remove the unit from service before failure during normal operation.
 Condition-Monitoring (CM): This is a maintenance process for items that have neither “hard-
time” nor “on-condition” maintenance as their primary maintenance process. CM is
accomplished by appropriate means available to an operator for finding and solving problem
areas. In effect, it obligates the user to apply knowledge gained by analysis of failures or other
indications of deterioration to consider action to improve performance.
39
The Hard-Time is the oldest maintenance process used in the aircraft industry and is still applied to
modern aircraft components. This type of maintenance defines a limit for the parts in terms of flight
hours, cycles or calendar time. The life of these parts is controlled and when the limit is reached they
are removed. The removed component can be repaired, overhauled or simply discarded. The On-
Condition maintenance is an evolution from the hard-time philosophy. For some parts, it was found
that it was possible to predict the probability of failure with great success by analysing some functional
parameters. When these start to present abnormal values, it is assumed that failure is imminent and
the part is removed from service. The on-condition maintenance tasks that are applied range from
visual inspections to workshop and laboratory tests. Condition Monitoring is not a predetermined and
preventative approach to aircraft maintenance like the first two processes, but a predictive approach
where the components/parts can actually fail in service. The target of condition monitoring is to define
improvements in the maintenance programs by increasing the aircraft availability and reducing
maintenance and operational costs. Nowadays, 70% to 80% of aircraft components have Condition
Monitoring as their primary maintenance process [31].
Historically, the engines were removed and overhauled after a fixed HT interval, even if they were
operating safely and in a satisfactory condition [7]. Condition-based Maintenance is a concept for
maintaining gas turbine engines that is gaining acceptance between airlines and engine
manufacturers [32]. One of the main objectives is to maximize the engine’s operational life and
overhaul only when the engine needs major maintenance work, thus saving in maintenance costs.
Depending on the engine model, technical condition, etc., each engine removal for a Shop Visit (SV)
may cost from less than 1 million to more than 10 million US dollars [3].
The Condition-based Maintenance is a predictive type of maintenance that determines the
maintenance needs based on the actual condition of the engine rather than on a preset schedule [33].
In order to have a regular overview of the proper functioning of the engines, Engine Condition
Monitoring is applied. It consists of a wide range of activities where the health of aircraft engines is
assessed and followed on a routine basis, from the moment the engine is put on-wing until its removal
[34]. Monitoring the condition of the engines involves the task of visualizing the evolution in time (or
trends) of engine operating parameters and searching for precursors of failure in its
modules/components. This is mainly a prognostic analysis that is based on the fact that most failures
do not occur instantaneously. The information is then used to ensure that preventive action is taken at
an early stage before safe operation is affected and that the root causes of the problem are identified.
When the engine gets to a state where its performance levels have deteriorated to the point where it
can no longer be operated within regulatory limits, ECM allows precise planning of its removal [34].
Under the Condition-based maintenance concept, the maintenance of aircraft engines is based on a
combination between actual engine health/condition and part-life limitations and the engines are
subject to control by the three Primary Maintenance Processes. General Electric (GE) [30] states that
these processes work hand in hand with one another and they carry equal weight in a maintenance
program.

40
5.2 Systems for Data Collection and Analysis
Several techniques and information can be used by an engineer to classify the condition of an aircraft
engine. The major source of information is the in-flight behavior of the aircraft and the process starts
with the recording of engine data during revenue service. The on-board systems used for acquiring
data for ECM tasks can present different capabilities and various levels of complexity. The most basic
system was also the most commonly used in older generation aircraft that didn’t have equipment for
automated data retrieval. The data is recorded manually by the pilots (or flight engineer) during the
take-off and cruise phases. The crew is generally requested to take into account some constraining
stability criteria in order to avoid irrelevant engine data. One of the disadvantages is that the
parameters measured are limited to those with cockpit instruments. Appendix G has an example of a
manual recording from the A310-325 aircraft. In addition to basic engine parameters, the flight
conditions (SAT, Mach, Altitude, etc.) and air bleed status are recorded during cruise. The criteria for
data acquisition are also described in this example.
Modern commercial aircraft are equipped with data acquisition systems that automatically record data
for monitoring purposes. They are capable of recording a much larger number of engine parameters,
coming from an increased number of sensors in the engines that can measure inter-stage
temperatures and pressures, for instance, as required. An important addition to these systems was the
DMU that allows performing a variety of additional functions, such as recording stable cruise data
using automated search criteria or recording takeoff data at the proper time using specific trigger
criteria. The quantity of data archived varies by application but will typically be composed of reports
that are taken at a limited number of engine operating points in each flight. The reports contain
information for engine vibration monitoring, engine limit exceedance data, event history recording and
engine performance monitoring. This topic will be addressed in more detail in Chapter 6.

Automatic Data Acquisition

Manual Data Acquisition

Manually
Recorded
Data

Powerplant Engineering
ARINC TAP Server
Department
Satellite G/S
Radio G/S

Figure 5.1 Engine data acquisition processes at TAP

There are several ways of transferring the data to the airline’s departments. Figure 5.1 represents
different engine data acquisition processes practiced at TAP. The most efficient method for automatic
data acquisition consists of transmitting the in-flight recorded data to ground stations via ACARS

41
(Aircraft Communications Addressing and Reporting System) messages, either by radio or satellite.
The messages are re-directed to the airline servers and then to the airline’s Powerplant Engineering
department. Although the cost of using ACARS is higher when compared to the other methods, there
is the added benefit of making the information immediately accessible to the engineers, who can
conduct real-time/near real-time assessments and plan in advance any required maintenance actions
to be undertaken on the aircraft, thereby optimizing the time on the ground. Manual data acquisition
methods involve fewer steps in the process, as it can be seen in figure 5.1, but they are less efficient.
The major disadvantage is that it can take a long time from the moment the recordings are executed to
the moment when they are analyzed. Moreover, the recorded data needs to be manually inserted by
trained operators into the specialized ECM tools. An alternative method to the crew manual recordings
consists of collecting the data stored in the aircraft’s DMU directly from printers in the cockpit.
Once the data arrives at the Powerplant Engineering department, it is interpreted by ECM analysis
software, typically provided by the Original Equipment Manufacturers (OEM) of the engines. ACARS
messages are typically transmitted in a format suitable for processing in the ECM software. When this
is not the case, the information needs to be converted. The primary function of the software tools is to
perform parameter trend monitoring. This is the primary process used by Powerplant engineers to
assess the engine’s condition and assess its deterioration, as mentioned before. From this
assessment, engineers communicate the results and recommend possible maintenance actions or, in
the most extreme cases, decide to remove the engines for a SV.
Some of the commercially available products from engine OEMs include the COMPASS (Rolls
Royce), the GE Diagnostics system (General Electric) or the P&W EHM (Pratt & Whitney). Engine
monitoring services are also included in the ADEM (Advanced Diagnostics and Engine Management),
the most advanced system provided by P&W.

5.3 Engine Parameter Measurements

5.3.1 The Turbofan Engine

Today, most of the airliners in operation are equipped with turbofan engines. They produce lower
noise levels than earlier generation jet engines and have considerably improved fuel economy.
The propulsive efficiency, , is a measure of the performance of a given propulsive system and can
be defined as the ratio of the thrust power (thrust times velocity) to the sum of that energy and the
unused kinetic energy of the jet engine [35]. The propulsive efficiency of a jet engine can be
expressed in terms of the inlet velocity of the air ( ) and the exhaust velocity ( ):

(5.1)

Reference [36] offers a derivation of this equation. An efficiency of 100% is attained if the exhaust
velocity is equal to the inlet velocity. The turbofan engine was originally conceived as a method of
improving the propulsion efficiency of the jet engine by reducing the mean jet velocity, particularly for
operation at high subsonic speeds [35].

42
In a turbofan engine, thrust is developed by a combination of two portions: a gas turbine engine uses
mechanical energy from the combustion and a ducted “fan”, which consists of a multi-bladed propeller
rotated by a low-pressure turbine, uses the mechanical energy from the gas turbine to accelerate the
air rearwards. The fan accelerates a large amount of air by a relatively small amount, which is more
efficient than accelerating a smaller volume of air by a large amount, according to equation 5.1.
The bypass ratio is the ratio of the air bypassing the engine core (secondary flow) compared to the
amount of air that goes through the engine core (primary flow). Early turbofan engines were “low-
bypass ratio” engines with approximately half of the thrust produced by the “fan” stage and the other
half by the primary flow. Current bypass ratios are around 5:1 (or higher) and the fan provides about
80 percent of the total thrust produced by the engine (e.g. CFM-56 engine [37]). The increase in the
bypass ratio brings an increase in the core thermal efficiency and improved fuel efficiency [7].
Figure 5.2 shows a turbofan engine with a two-spool configuration and its main assemblies, also
called modules. This type of engine is characterized by two independent stages of compression which
are mechanically linked by two separate shafts to two independent turbines.

Figure 5.2 Two-Spool Turbofan Engine [38]

5.3.2 Performance and Mechanical Parameters

There are two types of parameters used for monitoring gas turbine engine condition [39]: the
mechanical parameters and the performance parameters. The first group of parameters includes the
engine vibrations and oil temperature, pressure and consumption. These parameters are not
significantly influenced by the flight conditions and engine thrust setting. On the contrary, the
performance parameters are influenced by the flight conditions and thrust setting of the engine. The
following list provides a description of the key performance parameters in a turbofan engine:
 Engine Pressure Ratio (EPR): it is the ratio of the turbine discharge (exhaust) total pressure to
the total pressure at the front of the fan/compressor;
 Low-pressure rotor speed or Fan speed (N1): represents the actual RPM of the low-speed
rotor of the engine, and is usually expressed as a percentage of the reference rotation speed
of the fan;

43
  High-pressure rotor speed or Core speed (N2)1: represents the actual RPM of the hi-pressure
compressor and the high-pressure turbine, also as a percentage of some nominal value;
 Exhaust gas temperature (EGT): this is the temperature of the air that is exiting the low-
pressure turbine (engine exhaust);
 Fuel Flow (FF): represents the fuel needed by the engine in its actual performance state,
showing how much the engine is burning each hour.
The minimum gas path parameters that need to be recorded for engines that use the N1 as the thrust
reference parameter are the N1, N2, EGT and FF. For engines that use the EPR as the main thrust
reference parameter are the EPR, N1, N2, EGT and FF. The EPR is the primary thrust setting
parameter on Pratt & Whitney (e.g. 4168A) and Rolls-Royce engines, while General Electric engines
use the N1 (e.g. CF6-80E1).

5.4 Parameter Trend Monitoring

There is a significant variation in the flight conditions at the time that the engine recordings are taken.
Even during one flight the same flight parameters change continuously due to fuel consumption and
changes in the weather conditions. Therefore, a consistent procedure is necessary to obtain a relevant
time evolution of the engine parameters from the recorded data. Parameter trend monitoring is the
process in which the in-flight results are processed and then compared to a baseline model of how the
engine is expected to perform in the experienced conditions. The difference between the measured
data and the reference model is called the trend delta or parameter delta. The evolution of these delta
parameters is then used to assess the current state of the engine and estimate how its performance
has deteriorated with time.

5.4.1 Cruise Performance Trends

The most widely used method in the industry for cruise trend monitoring is to compare parameter data
from each engine to a “baseline” engine model presented in standard day conditions2 [40]. The
baseline models are generally developed by the engine manufacturers based upon flight-test data
and/or in-service experience and then fed to their ECM analysis software. In addition, a baseline
model can be derived from flight data of a fleet of aircraft that is then used to derive performance
characteristics for an average engine. For the comparison between the raw recorded data and the
“corrected” baseline model to be possible, it is necessary to correct the engine parameter data so that
the deviations are determined. The correction procedure and the expressions used are covered in the
next Chapter.
Nowadays, a vast quantity of engine parameters is recorded and is available for determining the
engine’s condition. However, ECM analysis is normally based on the evolution of parameters as the
engine temperatures and pressures, the rotor speeds, fuel flow, engine vibrations and oil monitoring.
The vibration data is not corrected to standard day units and trending is sometimes difficult [30]. Oil

1
Many Rolls-Royce engines have a three-spool design. For these engines, the N2 and the N3
designate the rotation rates of the medium-speed and high-speed rotors, respectively.
2
Temperature: 288.15 K, Pressure: 1013.25 hPa (Sea-level ISA atmospheric conditions)
44
pressure can be monitored and trended and in this case a comparison is also made with a reference
baseline. The engine parameters identified by manufacturers as being the main indicators of the
health of the engine and more sensitive to variations in its condition are the N1, N2, FF and EGT.
Figure 5.3 shows part of a short term Trend Plot report generated with the Pratt & Whitney’s EHM
software, which contains the trends of the performance parameters from a PW4156A engine
equipping the A310-325 aircraft. The input data consists of manual recordings in cruise like the one in
Appendix G. These plots, with the chronological trends of key engine parameters, are the most
common method used by engineering departments to monitor the engine’s condition.

5.4.1.1 Smoothed Parameters

It is a common practice to employ smoothing algorithms to the trend deltas that reduce the data
scatter and help to monitor the deterioration of the engine. Usually, the formula that is applied takes
the following shape:
(5.2)

Where is the smoothing coefficient. It quantifies by how much the smoothed trend line is influenced
by the variation between the last smoothed point and the new raw point - “raw” is not the recorded
value, but the recorded value after being corrected to standard day conditions.

Smoothed
Data

Raw Data

Figure 5.3 P&W EHM Short Term Trend Plot Report

The smoothed EGT, Fuel Flow, N2 and FF deltas are represented in figure 5.3 by upper case letters.
The most recent “raw” data can be visualized at the bottom of the graph represented by lower case
letters and they are included to allow immediate identification of rapid changes in the performance
levels of the engine that only become visible later in the smoothed data. Although this is not the case,
smoothed and “raw” data may be overlapped in the same plot. Powerplant engineers have to look at

45
both the smoothed and raw streams of data and search for shifts in the most recent trends in terms of
the evolution of the EGT, FF, N1 and N2 deltas.

5.4.1.2 Gradual vs. Rapid Performance Shifts

Figure 5.4 Slow Drifts vs. Step Shifts [30]

There are two major types of movements that can be identified in a trend [30]. Both are graphically
represented in figure 5.4. A Step Shift is a shift that occurs in the data in a short period of time and is
characterized by a sudden increase or decrease on the associated trend, whereas a Slow Drift is a
slow movement in the trend that occurs over a longer period of time.
As the engine accumulates flight cycles, the life of many parts is consumed and its performance
deteriorates [41]. This can be due to several reasons: dust/dirt ingestion and accumulation on the
turbine and compressor blades, increasing tip clearances on the compressor and turbine blades,
erosion of the airfoils, hot section oxidation, etc. Slow Drifts represent the gradual deterioration of the
performance of the engine that is caused by these factors. On the other hand, Step Shifts in the
observed delta parameters can indicate mechanical changes in the engines that may propagate to
failure and lead to the occurrence of incidents, such as in-flight shutdowns and aborted take-offs. The
effective monitoring of the cruise trends helps to minimize the risks associated with unexpected engine
failures, which result in unscheduled engine removals with additional maintenance costs to the airline,
and to avoid excessive degradation of the engine’s performance by checking the performance delta
parameters against potential problem limits.

Fingerprints
The detection of abnormal trends in the critical engine performance parameters is proving to be an
efficient technique for providing warning information of ongoing problems or impending failures prior to
serious malfunctions in the gas turbine engines [32]. Although it can be difficult to quantify the exact
parameter shifts for each degradation type [34], there are certain kinds of anomalies in the engine’s
operation that result in specific changes in the delta parameters being monitored. They are identified
by combined evolutions of key performance parameters and allow the problem to be localized within
the engine’s modules. The general patterns from a specific engine model are supplied by the
manufacturer in the form of the so-called fingerprints.
GE projects these shift patterns from a computer thermodynamic model containing the engine-cycle
model [30] operating at specific flight conditions.

46
 Figure 5.5 Fingerprints for the CFM-56 family of engines [30]

The impact of several types of damage and deterioration – loss of modular efficiency, bleed leakage,
etc. - on the overall performance parameters of the engine (EGT, FF and N2) is organized into
“Theoretical Parameter Shifts” tables, which are similar to the fingerprints. The operator uses these
tables as a starting point in the closer interpretation of the problems and relies on his skills built from
previous experience [34]. Figure 5.5 illustrates theoretical parameter shifts for a CFM-56 engine
operating at constant altitude, Mach, temperature and thrust setting (N1). In addition, fingerprints can
present causes for explaining similar trend shifts observed in more than one engine on the same
aircraft. Usually when this happens the problem is attributed to an instrumentation error.

5.4.1.3 Divergence
The former paragraphs have covered the method of monitoring the health of an engine based on a
comparison with a baseline model. On multi-engine aircraft there is another method, which consists of
comparing each individual engine parameter with the average value from all the engines in the aircraft.
The Divergence is a parameter used for these types of comparisons, which is mathematically obtained
from the following expression:

(5.3)

Which is applicable to a twin engine aircraft and can be any performance parameter. The main
advantage of using the divergence is that the measured parameters can be used and no corrections
are required. Any possible shifts caused by TAT, Altitude or Mach number, are eliminated [40].

5.4.2 Take-off Considerations

The appropriate indicator of the overall performance of the engine is based on the core flow
temperature and is measured at the turbine exit [42]. It is usually referred as EGT or TGT (Turbine
Gas Temperature). This temperature is measured on the gas path at the Low Pressure (LP) turbine
inlet or at the LP turbine exit. Higher temperatures for the same thrust indicate an increase in the fuel
burned and a decrease in the engine’s efficiency, meaning that deterioration is affecting the engine.

47
5.4.2.1 EGT Margin
To meet aircraft performance requirements, engines are designed to provide a constant Thrust up to a
designated Corner Point (CPT) or Flat Rate Temperature (FRT). Below this temperature the Thrust is
limited by software (pin programming), the same engine being able to provide different levels of thrust.
When the Outside Air Temperature (OAT) is lower than the FRT, the EGT is less than the limit. This
limit is called the EGT Red Line and is demonstrated during endurance tests required for engine
certification [42]. When OAT is higher than FRT, the fuel flow has to be reduced in order to keep the
EGT below the limit and protect the turbine hardware. As a consequence, the Thrust is also reduced
when the OAT increases.

Figure 5.6 EGT Considerations [30]

The EGT Margin is an estimate of the difference between the certified EGT Red Line and a projection
of the engine EGT to full (non-derated) take-off reference conditions. The observed/recorded peak
EGT during the take-off, which is acquired automatically in modern aircraft, is projected to the
reference condition of full take-off power, on a FRT day at sea level, using characteristics from the
Manufacturer’s thermodynamic model of the engine. This projected temperature represents the
expected EGT if the take-off actually occurred with the reference conditions. The projected EGT is
subtracted from the certified EGT redline to yield the EGT Margin or EGT Hot Day Margin (EGTHD
Margin). Therefore, the EGT Margin is not just the difference between the peak EGT and the EGT Red
Line, since it is very unlikely that a take-off occurs at the same reference condition and temperature.
The EGT Margin is routinely used to monitor the health of the engines using appropriate ECM tools,
together with the trends from the cruise performance parameter deltas. Trends in the takeoff EGT
Margin can also be used to detect shifts in the engine’s performance, which can dictate the need for
inspections and/or maintenance [42]. EGT Margin trends can also be used to forecast the remaining
time of the engine on the wing in terms of time or engine cycles, by predicting the point at which the
margin will be completely eroded. The EGT Margin for new engines is defined to meet most take-off
conditions and to allow airlines to operate the engine on-wing during an acceptable amount of time
[43]. The EGT Margin should not be used as the sole criterion to plan engine removals, although it is
one of the primary engine removal causes. Life Limited Parts constraints, cruise trends and the
number of EGT over limit occurrences and associated maintenance tasks should also be considered
prior to an engine removal decision.

5.4.2.2 Outside Air Temperature Limit

For the engines in which the take-off EGT is nearly constant with the OAT after the FRT, the Outside
Air Temperature Limit (OATL) is another indicator of the engine health. The OATL is a projection of

48
the highest ambient temperature at which an engine should be able to produce full thrust without
exceeding the certified EGT red line [42]. Both the OATL and EGT Margin are similar measures of
performance based on take-off data. They are not independent assessments of temperature
limitations in the engine. If the OATL is equal to the FRT of the engine, the EGT Margin is zero.

Figure 5.7 Deterioration effect on the EGT Margin and OATL [42]

49
6. Study of Algorithms for ECM Trend Monitoring

The main objective of this work is the development of algorithms for performance trend analysis of
engines. The methodology that will be presented in this Chapter and the next relies only on recorded
flight data to derive the engine baseline models and determine its deterioration levels. The objective is
to use this approach in complement to the trend monitoring provided by the OEM’s software.
The PW4168A engine was selected as the case study engine and will be presented in the first place.
The algorithm responsible for the extraction of stability points from the flight data is described next.
These points contain the fundamental aircraft and engine data for the required analysis. In order to
understand how they are acquired, some definitions need to be introduced and understood. Finally,
the correction procedure and the formulas used in the current work are presented. The correction
procedure employs some dimensionless parameters that take into account the inlet temperature and
ambient pressure and calculates the engine parameter values in Standard Day conditions.

6.1 Case Study: PW4168A Engine

The quantity of flight data available from the A310-325’s QAR is limited to the parameters that are
programmed in the DFDR frame. Some key engine performance parameters from the PW4156A are
not recorded, such as the N1, and others are not recorded with the desired frequency, like the N2. In
addition, there is not much historical flight data available from the aircraft. These reasons make the
implementation of good quality ECM processes based on recorded flight data to the PW4156A engine
very difficult. The following studies were then conducted on the PW4168A engine, which powers
TAP’s A330-223 aircraft. These aircraft are equipped with a Digital ACMS Recorder (DAR), with a
recording capacity of 512 words per second. The DAR collects data from the DMU, which offers the
possibility of programming which parameters are to be recorded and inserting information about their
range, accuracy and recording rate.

6.1.1 Description of the Engine

The PW4000 100-inch engine was developed specifically for the A330 wide-bodied twinjet and
entered into service in 1994. It is a dual rotor (two-spool configuration), axial flow turbofan engine with
separate primary and fan duct exhaust systems. TAP’s long-haul fleet currently operates 7 A330-223
aircraft, which are equipped with the PW4168A engine version. This version delivers a maximum static
take-off thrust of 68,600 lbs. at sea level conditions and has a bypass ratio of 4.9.
General Characteristics Nominal Performance (Sea Level, Static)
 Diameter, Fan tip: 2.54 m (100 inches)  Take-off thrust: 68,600 lbs
 Length: 4.14 m (163.1 inches)  Flat rated temperature: 30ºC
 Bypass ratio: 4.9
 Overall pressure ratio: 32.4
 Fan pressure ratio: 1.75
Table 6.1 Specifications of the PW4168A Engine [44]

Figure 6.1 identifies the different Gas Path configuration areas on the PW4168A engine and the
Engine Stations. The latter correspond to location points in the engine gas path that are used to
describe the engine’s operation. For instance, the EPR is the ratio of the turbine discharge total

50
pressure (Station 4.95) to the compressor inlet total pressure (Station 2) and the EGT is the total
temperature of the low pressure discharge gas flow (Station 4.95).
LOW PRESSURE
COMPRESSOR

HIGH PRESSURE

HIGH PRESSURE

LOW PRESSURE
COMPRESSOR

COMBUSTION

EXAHUST
TURBINE

TURBINE

TURBINE
SECTION

CASE
Figure 6.1 PW4168A: Gas Path configuration areas and Engine Stations

Engine Cruise Report <01>

The main component of the Aircraft Condition and Monitoring System (ACMS) of the Airbus A330-223
aircraft is the DMU, which acquires and processes data coming from the ARINC 429 data buses of the
aircraft. One of the tasks performed in real time by this unit, besides sending data for recording in the
DAR, is the generation of reports – the “Airbus Standard Reports” – according to specific trigger
conditions, which have been defined and validated by the manufacturer [45]. These conditions,
associated limits, new specific trigger criteria and the information displayed on the reports can be
reprogrammed by the operator using external tools. Within the scope of this work, the interest is on the
Engine Cruise Report (01), used for engine trend monitoring of TAP’s A330-223 fleet. It consists of a
collection of both aircraft and engine information where the majority of the values displayed are
averaged values from the stable frame, i.e., the period of time in which the trigger conditions and
criteria were met. Other parameter values correspond to the value at the moment of trigger.

Figure 6.2 Format of the A330’s Engine Cruise Report <01> [46]
51
6.2 Stability Points
To monitor the performance of an engine and to evaluate its deterioration, it is important to collect data
that is representative of the engine’s behavior. As mentioned in the previous sections, one of the
means of achieving this consists of using in-flight data processed by the aircraft’s DMU that is then
compiled into reports. The Engine Cruise Report, in particular, contains information about the aircraft
and the engine’s operation relative to a point during the Cruise flight phase, where a set of conditions
and stability criteria were respected during a certain period. This period is typically 100 or 120 seconds
and a point in these conditions is also called a Stability Point.
Depending on the flight conditions and the logic contained in the DMU, there can be several stability
points found for a particular flight. The method used to select which one is the ideal to represent the
engine’s performance depends on the recording system. In the A330-223 aircraft, the DMU selects the
best stability point based on the quality number parameter, which is a measure of the parameter
dispersion in the period where the conditions and criteria are met.
This section presents the development of an algorithm for extracting stability points from flight data
recorded in the DAR. The flights are exported from AGS with engineering values and then they are
processed by the algorithm. First, the stability criteria and trigger conditions employed by the DMU for
the Engine Cruise Report were used. Then, a new set of more restrictive stability criteria was applied,
leading to a reduction in the number of stability points for each flight. The objective of this study is to
acquire stability points with better quality than those acquired by the aircraft’s DMU and to use these
points for a trend monitoring analysis.

6.2.1 Observation Window and Stable Frame

Most of the parameters that make part of a Stability Point are averaged over a period of time that was
identified as meeting all the conditions and stability criteria. A period of time identified as a stability
point is called in the A330-223 aircraft a Stable Frame and has a duration of 100 seconds.
Before searching for stable frames, there are some conditions that need to be verified first. They are
used to avoid report triggering in flight phases where the parameters are of no interest and are called
the Basic Conditions. They include, for instance, a range for the Altitude, Mach number and fan
speeds where stable frames can be searched and impose some conditions on the status of the anti-
ice and bleed discretes, for instance. When the Basic Conditions are met, the DMU searches for
stable frames. It looks at the parameter data in each 100 seconds of flight and then computes the
difference between the maximum and minimum values. If the difference is not above the parameter’s
stability criterion and if this is verified for every parameter then the period is a stable frame. Stability
Criteria are defined for parameters such as the Ground Speed, Total Air Temperature, Fuel Flow,
Vertical Acceleration, etc.

Maximum allowed
variation of the parameter

Non-stable
100 Seconds Stable

Figure 6.3 Stable and non-stable variation of a parameter inside an Observation Window

52
Individual vs. Gliding Observation Windows
There are two distinct methods of searching stable frames. This first method consists of performing
computations for each individual 100 seconds observation window in the flight. Individual means that
each window is independent from the others, covering data from different periods of time. The second
method uses a gliding window, where the observation window advances seconds. The advantages
of this method are obvious: if is small (e.g. 1 second) the number of observation windows considered
for the stable frame search increases significantly, with the next window containing all the data of the
previous window except the first seconds. Thus, a lot of observation windows discarded by the first
method can be included in the second, leading to stability points with better quality.
100 Seconds 100 Seconds

Individual Windows: 1st Window 2nd Window

Gliding Windows:

1st Window 2nd Window

20 Seconds

Figure 6.4 Individual and Gliding Window methods

The A330-223’s DMU searches for the best stable frame using a gliding window method with an
advancing front of 20 seconds ( ). In the algorithm developed, the advance in the gliding window
is equal to the period with which the flights were exported from AGS.

6.2.2 Quality Number

The number of stability points encountered during a flight depends mainly on the flight conditions
experienced, the duration of that flight and the stability criteria. If several stability points are identified it
is necessary to have a consistent procedure that selects the best of them, which will then be used for
generating trend monitoring results. Stability points are selected based on the quality number that is
computed for the corresponding stable frame period [45]. The Engine Quality Number ( ), is the sum
of each of the individual quality numbers of the stable frame parameters and is computed using the
following formula:

(6.1)

Where is the weight factor, is the tolerance or the maximum variation of the parameter
allowed in the observation window and is the variance of the parameter defined as:
∑ ̅ (6.2)

With equal to the number of samples of parameter in the observation window. The mean ̅ is
calculated using the formula below:
∑ (6.3)
̅

Notice that the variance measures how close the parameter values in the observation window are to
the mean. The stability point is better when the overall variance of its parameters is lower. For a stable
frame parameter, the calculated mean is the value that will be recorded in the stability point. In our

53
computations the weighting factor is equal to 1 for all parameters, similar to what is currently used in
the aircraft [46].

6.2.3 Algorithm for Extraction of Stability Points

The algorithm to extract Stability Points from the recorded flight data was developed using the R
programming language. To acquire these points, the logic from the Engine Cruise Report was initially
implemented in the algorithm and consists of two main parts:
- Basic Conditions
- Stable Frame Criteria.
After validating the algorithm, this was used to study the implementation of a new set of stability
criteria. This is one of the advantages of using recorded data rather than the processed reports from
the DMU: the ability to refine existing calculation routines and study their effect on the results, from
flights that have already occurred. Over the next sub-sections the Basic Conditions, Stable Frame
Criteria and the parameters needed for the computations will be presented and the algorithm
described.

6.2.3.1 Basic Conditions

The Basic Conditions are used to enable or disable the search for stable frames and to avoid that
Stability Points are generated whenever the parameters are of no use. The discrete and binary type
parameters used for these conditions are displayed in tables 6.2 and 6.3, respectively.

Discrete Type Parameters

Recording
Parameter Description Equipment/Label/SDI
Frequency
FNAI1 ENG 1 NACELLE ANTI-ICE VALVE POS (OFF) ZC/076/11 1 Hz
FNAI2 ENG 2 NACELLE ANTI-ICE VALVE POS (OFF) ZC/076/11 1 Hz
WING ANTI-ICE OUTER POS - LEFT
WAO1 DMU-1/263/11 1 Hz
FILTERED
WING ANTI-ICE INNER POS - LEFT DMU-1/263/01
WAI1 1 Hz
FILTERED
WING ANTI-ICE INNER POS - RIGHT
WAI2 DMU-1/263/10 1 Hz
FILTERED
WING ANTI-ICE OUTER POS - RIGHT
WAO2 DMU-1/263/00 1 Hz
FILTERED
XBV_POS2 CROSS BLEED VALVE POSITION BMC-1/055/01 1 Hz
APUBV_O APU BLEED VALVE OPEN APU-ECB/037/01 1 Hz
EG_PRV_POS ENG PRV POSITION BMC-1/066/01 1 Hz
O_EGPRV_POS OPPOSITE ENG PRV POSITION BMC-1/067/01 1 Hz
EG_HPV_POS ENG HPV POSITION BMC-1/066/01 1 Hz
O_EGHPV_POS OPPOSITE ENG HPV POSITION BMC-1/067/01 1 Hz
Table 6.2 Discrete Parameters used in the Basic Conditions

Binary (BNR) Type Parameters

Recording
Parameter Description Equipment/Label/SDI
Frequency
ALT_STD ALTITUDE STANDARD (1013.25MB) ADC-1/203/01 1 Hz
MACHR1 MACH NUMBER ADC1 [TAP] ADC-1/205/01 1 Hz
N11 N1 ACTUAL ENGINE 1 DMC_ECAM/346/01 1 Hz
N12 N1 ACTUAL ENGINE 2 DMC_ECAM/346/10 1 Hz
TAT TOTAL AIR TEMPERATURE ADC-1/211/01 1 Hz
Table 6.3 Binary Parameters used in the Basic Conditions

54
The following table describes the logic for true Basic Conditions with the corresponding parameters.
For the discrete parameters, the bit’s value is also specified.
Logic for True Condition Parameter(s)
Nacelle Anti Ice Status Off FNAI1=1, FNAI2=1
Wing Anti Ice Status Closed WAO1=1, WAI1=1, WAI2=1, WAO2=1
20000ft.<Altitude<41100ft. ALT_STD
0.6<Mach Number<0.86 MACHR1
70<N1K<120 N11, N12, TAT
N1K=N1/(?2^0.5)
?2=(TAT+273.15)/288.15
Symmetrical Engine Bleed Configuration EG_PRV_POS=O_ENGPRV_POS,
EG_HPV_POS=O_ENGHPV_POS
Cross Feed Valve Closed XBV_POS2=0
APU Bleed Valve Closed APUBV_O=0
Table 6.4 Logic for True Basic Conditions and Parameters used

The conditions in table 6.4 must remain true during the observation window before stable frame
calculations for the maximum and minimum values are performed and stability points encountered.
When available, the parameters that are corrected in AGS using additional procedures - ALT_STDC,
N11C, N12C - were used. This holds true for the stable frame parameters.

6.2.3.2 Stable Frame Criteria

When the Basic Conditions for the observation window are checked, the differences between the
maximum and minimum values of the parameters in table 6.5 are computed. These parameters are
the Stable Frame Parameters.

Stable Frame Parameters

Recording
Parameter Description Eq/Label/SDI
Frequency
IALT INERTIAL ALTITUDE IRS-1/361/ 1 Hz
GS GROUND SPEED DMC_EFIS/312/01 1 Hz
ROLL ROLL ANGLE IRS-1/325/01 2 Hz
TAT TOTAL AIR TEMPERATURE ADC-1/211/01 1 Hz
N21 N2 ACTUAL ENGINE 1 DMC_ECAM/344/01 1 Hz
N22 N2 ACTUAL ENGINE 2 DMC_ECAM/344/10 1 Hz
EGT1 EGT ENGINE 1 (T495) DMC_ECAM/345/01 1 Hz
EGT2 EGT ENGINE 2 (T495) DMC_ECAM/345/10 1 Hz
GVRTI VERTICAL ACCELERATION IRS-1/364/01 8 Hz
MACHR1 MACH NUMBER ADC1 [TAP] ADC-1/205/01 1 Hz
N11 N1 ACTUAL ENGINE 1 DMC_ECAM/346/01 1 Hz
N12 N1 ACTUAL ENGINE 2 DMC_ECAM/346/10 1 Hz
PT21 LOCAL PT2 ENGINE 1 EIVMU-1/131/01 1 Hz
PT22 LOCAL PT2 ENGINE 2 EIVMU-1/131/10 1 Hz
FF1 FUEL FLOW ENGINE 1 (kg/hr) DMC_ECAM/244/01 1 Hz
FF2 FUEL FLOW ENGINE 1 (kg/hr) DMC_ECAM/244/10 1 Hz
EPR1 Actual EPR ENG #1 EIVMU-1/340/01 1 Hz
EPR2 Actual EPR ENG #2 EIVMU-1/340/10 1 Hz
HPT1 SELECTED HPTC POSITION ENGINE 1 EIVMU-1/330/01 1 Hz
HPT2 SELECTED HPTC POSITION ENGINE 2 EIVMU-2/330/10 1 Hz
LPT1 SELECTED LPTC POSITION ENGINE 1 EIVMU-1/331/01 1 Hz
LPT2 SELECTED LPTC POSITION ENGINE 1 EIVMU-1/331/01 1 Hz
Table 6.5 Stable Frame Parameters

The difference for each parameter is computed in order to be compared to a defined quantity, the
tolerance, which constitutes the stability criterion. Table 6.6 contains the stability criteria used in the
aircraft and initially implemented in the algorithm.

55
 Aircraft Stability Criteria
IALT 100 ft.
GS 5 Knots
ROLL 0.8º
TAT 1.1ºC
N21, N22 1.0%
EGT1, EGT2 22ºC
GVRTI 0.05 g
MACHR1 0.008
N11, N12 1.8%
PT21, PT22 0.05 PSIA
FF1, FF2 200 kg/hr
EPR1, EPR2 0.05
HPT1, HPT2 5%
LPT1, LPT2 5%
Table 6.6 Initial Stability Criteria

6.2.3.3 Filtering of Parameters

The on-board systems of the aircraft also filter the parameter data, in order to reduce the noise
typically encountered in the readings. The formula used to filter these readings is the following:
) (6.4)
Where:
NFV = New Filtered Value t = Samples per Second
OFV = Old Filtered Value T = Time Constant
NRV = New Raw Value
The division is constant for the filtered parameters and equal to 1/3. The result of implementing
the formula above is that instead of using directly the parameter reading (NRV), the value used is
equal to the OFV plus one third of the difference between the first and the latter. Table 6.7 lists the
parameters that are filtered in the algorithm developed.
Basic Parameters ALT_STDC, MACHR1
Stable Frame Parameters ROLL, TAT, N21C, N22C, EGT1C, EGT2C, GVRTI, N11C, N12C, FF1C, FF2C
Table 6.7 List of Filtered Parameters

6.2.3.4 Description
Figure 6.5 contains the flowchart for the algorithm developed with the R language, which is essentially
divided in two parts. The first part is responsible for reading and processing the files containing the
flight data and the second part writes the results to dedicated files. The flights were exported from
AGS in .csv format with a parameter rate of 8Hz to take into account all the recorded values from the
parameters listed in tables 6.2, 6.3 and 6.5. In addition to these parameters, others have been
exported containing the Flight Phase (FLIGHT_PHASE), the recorded date and time (DATE and
TIME_R, respectively) and the aircraft tail registration (AC_TAIL_23, AC_TAIL456 and AC_TAIL7). It
is also important to notice that the number of decimal places for each parameter in the exported flights
also needs to be defined and taken into account.

56
 Load Flight

Read Flight Data

YES Write file with Recorded Data and

End of Flight?
Results
NO
Filtering of the Parameters NO
Number of Stability
Points found > 0?
NO Flight Phase = Cruise YES
(FLIGHT_PHASE=8)?
YES
NO Write file with all the Stability Points
Waiting Time > 15
Minutes?
YES
NO Are the Basic Conditions Write file with the best Stability
met? Point(s)
YES

Search for Best Stability Point

To next Flight or End Analysis

NO Are the Stability Criteria
satisfied?
YES
Calculates Mean and Variance

Calculates Quality Number

Figure 6.5 Flowchart of the Algorithm for Extraction of Stability Points

The algorithm starts by loading the flight data in the file to a dataframe, a structure similar to a table,
which is then converted to a matrix for quicker computations. Afterwards, there is a cycle that reads
every line of data. Most parameters are recorded with frequencies lower than 8Hz so many entries are
missing data. R has some handy functions that allow dealing with the situation. For instance, auxiliary
variables were created with information from the last recorded value and new filtered values are only
calculated when the recorded parameter value is updated.
There are some conditions that need to be checked before searching for stability points. The first
condition is based on the Flight Phase. Instead of the FM_FWC parameter used in the aircraft, the
algorithm checks the FLIGHT_PHASE parameter computed from the Flight Phase Computation
procedure. Stability Points can only be generated in Cruise, when FLIGHT_PHASE=8. Next, a
condition for the Waiting Time was implemented. The Waiting Time corresponds to a period of 15
minutes when stability points cannot be searched or generated that occurs in two situations:
1) After Cruise is attained for the first time in the flight;
2) After a search period of 240 minutes (4 hours) for Stability Points.
Because long-haul flights performed by A330-223 aircraft have durations of 8, 9 and more hours, it is
possible to generate stability points from different periods of the flight with data acquired at different
altitudes, temperatures, engine settings, etc. Finally, the Basic Conditions in the observation window
are checked using the logic presented in table 6.4. If they are true, the algorithm will search stability
points for a maximum of 240 minutes. It computes the difference between the maximum and minimum

57
parameter values, using filtered values where applicable, and checks these against the tolerances
defined in table 6.6. When the differences are lower, a stability point is generated and the quality
number for the stable frame is calculated using equation 6.1.
Figure 6.6 uses a representative flight profile to better illustrate the conditions described above. Notice
that the counter implemented for the search period is only incremented when the Basic Conditions are
true. The figure also shows the point in which the Engine Cruise Report is generated by the DMU.
CRUISE attained for Basic Conditions Report Generated Start of new Search
the first time in the True (in the aircraft) Period
flight

Initial Waiting
Search Period Waiting Time
Time

15 Minutes 240 Minutes 15 Minutes

Figure 6.6 Representation of the Waiting and Search times

When the cycle reaches the end of the flight, a file with the recorded data and the results is written. If
stability points are encountered in the processed flight, two additional files are written: one containing
the data from all the stability points found and another containing the best point from each search
period.

6.2.4 Results
The algorithm was tested with a selection of flights from different aircraft of the A330-223 fleet in order
to make the results the most representative possible. The criteria from the aircraft in table 6.6 were
initially used for generating stability points. After this implementation, it was found that stability points
were not being generated from a high number of flights (approximately 50%). The reason why this
happened was because in many of these flights the anti-ice discretes maintain the ON status and
therefore the Basic Conditions were never met. However, this is an expected situation during normal
operation and the studies proceeded with the flights with stability points.

Flight Results
Figure 6.7 illustrates some of the results obtained from a flight processed by the algorithm. The
pressure altitude is represented by the ALT_STDC parameter and clearly illustrates the climb, cruise
and descent flight phases. The flag_cruise variable indicates when the aircraft is in Cruise. From the
moment the aircraft finishes its initial climb up to a stable cruise altitude until starting the descent for
landing, the aircraft is in cruise. The exception is made to some periods where the aircraft transitions
between flight levels. The basic_conditions variable indicates when the Basic Conditions are met. For
this flight, Basic Conditions exist during most of the cruise, except for the initial 15 minutes after cruise
is reached that are indicated with the black circle on the left of figure 6.7. The functioning of the
counter variables for the search and waiting times is illustrated here with red and black lines,
respectively. After the basic conditions are met for the first time, a search period of 4 hours (144000
eighths of second) for stability points begins. The search time is incremented whenever the basic
conditions are true and in cruise. After 4 hours, there is a waiting time of 15 minutes before starting a

58
new search period. For the flight pictured below, there is more than one search period. The
stability_point variable indicates when a stability point is found and it’s numbered after the
corresponding search period.

: Waiting Times

Figure 6.7 Flight Results: Search Periods, Waiting Times and Stability Point
Study of new Stability Criteria
The results in figure 6.7 show that a large number of stability points are encountered during the flights
using the aircraft stability criteria. For the flights studied there were thousands of stability points
encountered. This showed the potential to study a new set of more restrictive criteria. Table 6.8
displays the tolerances for the Converged Stability Criteria, together with the tolerances from the
aircraft used initially by the algorithm. The new criteria are the result of an iterative process: each time
the tolerances were reduced, the flights were processed by the algorithm and the number of stability
points checked. After this iterative process, only approximately 10% of the exported flights did not
generate any stability point and a significant reduction in the number of points in each flight was
obtained. The tolerances for the engine performance parameters (N2, EGT, N1 and FF) are
considerably lower in the new criteria. The tolerance for the EPR was not altered because the
recording resolution of the parameter is already lower than the value used in the aircraft. In practice,
this means that the value of the EPR is kept constant in every stable frame.
Parameter Aircraft Stability Criteria Converged Stability Criteria
IALT 100 ft. 35 ft.
GS 5 Knots 2 Knots
ROLL 0.8º 0.6º
TAT 1.1ºC 0.5ºC
N21, N22 1.0% 0.2%
EGT1, EGT2 22ºC 3ºC
GVRTI 0.05 g 0.04 g
MACHR1 0.008 0.0035
N11, N12 1.8% 0.3%
PT21, PT22 0.05 PSIA 0.04 PSIA
FF1, FF2 200 kg/hr 50 kg/hr
EPR1, EPR2 0.05 0.05
HPT1, HPT2 5% 5%
LPT1, LPT2 5% 5%
Table 6.8 Aircraft and Converged Stability Criteria

59
The comparison between both criteria is done in the next Chapter, where the results related with the
engine baseline models and trend monitoring are covered. Since the stability points acquired with the
Converged Stability Criteria correspond to periods of the flight with higher stability, it is necessary to
assess the potential of deriving more accurate engine baseline models from them.

6.3 Engine Parameter Corrections

Ambient conditions have a significant impact on the various parameters along the engine’s gas path,
such as flows, temperatures, pressures, speeds, etc. These not only vary with the power condition, but
also with the temperature and/or pressure at the engine’s inlet [47]. A change in these conditions
contributes to an attendant change in the gas parameter’s value, so it would be difficult to characterize
the aero-thermodynamic relationships between the different gas turbine engine parameters (even at a
constant engine operating point) unless the ambient conditions are accounted for. Generally, these
relationships are determined by using corrected engine parameters.
The purpose of correcting measured data from a gas turbine test is to make the results comparable to
those from other engines or with acceptance test criteria [48]. Data correction algorithms are also
applied to monitor the engine deterioration. The classical parameter corrections employ a number of
dimensionless parameters that are derived from dimensional analysis [47]. This has been the primary
technique for establishing the classical parameter corrections. References [47] and [36], describe the
principle of dimensional analysis using the work by Buckingham [49] and provide illustrative examples
applied to turbojet and turbofan engines, respectively. The equations presented in the next sections
use a different method that relies on basic calculus and some thermodynamic principles.

6.3.1 General Formulation

Consider the two-spool, mixed flow1, turbofan engine and its station numbers as illustrated in figure
6.8. The following discussion will use this type of engine as reference. The numerical demonstration
was taken from Reference [47].

Figure 6.8 Twin spool mixed flow turbofan [47]

In this discussion, a generic gas path parameter is denoted by and its equivalent corrected
parameter by . Generally, a change in the inlet temperature ( ) and pressure ( ) is accompanied
by an attendant change in any downstream gas path parameter , as mentioned initially. The

1. Mixed Flow in turbofan engines means that the cold air bypassing the core section of the
engine and the hot air exiting this section are exhausted through a single exit nozzle.

60
corrected parameter would be constant regardless of the change in the inlet condition ( , ) and
represents the value that the parameter would have at a fixed reference inlet condition. The
reference condition can be defined arbitrarily, but it is a common practice to select standard day
atmospheric conditions at sea-level for correction purposes ( =288.15 K and =1013.25 hPa). We
assume, without loss of generality, that . Thus, it follows that

( ) ( ) ( ) (6.5)

⁄ ⁄
( ) ( )
⁄ ⁄
(6.6)
⁄
( )
⁄

For the following expression, the simplifying assumption is made that the first two partials are constant
(the third is constant and equal to one), and denote them by and , respectively

(6.7)

Next, we define the dimensionless parameters (or correction factors) ⁄ and

⁄ and arrive at

(6.8)

Equation 6.8 is the most widely used parameter correction formulae. Notice that the corrected
parameter maintains the same units and is calculated with the theta ( ) and delta ( ) exponent
corrections. Table 6.9 summarizes some of the common gas turbine parameter corrections and the
standard classical values for the exponents and .
Parameter Corrected Parameter

Rotor Speed (N) 0.5 0

√
Fuel flow (FF) 0.5 (classical) 1
√
Temperature (T) 1 (classical) 0
Table 6.9 Common Gas Turbine Parameter Corrections

Notice the letter to indicate a corrected parameter. We will follow this nomenclature in the
forthcoming chapters. Reference [47] also offers a derivation of these (and additional) classical
corrections, from simple thermodynamic relationships and assumptions that do not require extensive
knowledge of the thermodynamics of the gas turbine operation.
Mach number similarity parameters are also applied in the correction procedure [48]. Consider the
Corrected Rotational Speed or Corrected Fan Speed. The tangential velocity ( ) is related to the
rotational speed (RPM) by the radius ( ) of the fan blades, and simultaneously to the tip Mach number

61
( ) and square root of the inlet temperature . This can be translated mathematically to the following
equation
√ (6.9)

Where and are the ratio of specific heats and the gas constant, respectively. Taking logs of the
expression and differentiating and assuming a constant Mach, we obtain

(6.10)

(6.11)
√
If we use an additional parameter to describe the axial Mach number, then the inlet flow angle is fixed
and the velocity triangles at the inlet in terms of the Mach numbers are the same as the baseline
reference condition. Because the Mach number describes the compressibility effects and the flow
angles dictate losses and work done in turbomachines, the efficiencies, temperature ratios and
pressure ratios are not altered during the correction procedure.

Approximations and Limitations

Generally, the exponents and will depend on the engine and cycle. The quality of the parameter
correction procedure can be improved when applying some corrections to the exponents a and b.
Reference [48], for instance, details how the exponents for theta and delta may be determined by
running a full thermodynamic computer model of the engine over a range of inlet temperatures and
pressures and exhaust conditions. Although this high level model may be available to the engine
manufacturers, operators usually do not possess detailed information to run engine simulation models.
There are, however, empirical methods that utilize actual engine data collected over a temperature
and/or pressure range and consist more on a statistical approach to the parameter correction
procedure. In [47], the author proposes an empirical formula that can be used for temperature and
pressure corrections in the same manner. This simple method was then tested against more complex
and iterative statistical methods and the exponents returned were essentially the same.
The impacts of changes in the viscosity with altitude (Reynolds number effects) are not considered
with equation 6.8. While the effects of the Reynolds number might not be important for some engine
models, there can be errors involved in the comparison of corrected parameters, at a reference
condition, if these effects are not accurately predicted. The ratio of specific heats and the gas constant
are assumed to be invariant in the atmosphere. This is a typical assumption when performing
dimensional analysis on gas turbine engines.
The ratio of specific heats and the gas constant are assumed to be invariant in the atmosphere. This is
a typical assumption when performing dimensional analysis on gas turbine engines.

6.3.2 Application to the Case Study Engine

6.3.2.1 Correction Factors

The computation of the theta and delta correction factors is done using the total temperature and the
total pressure at the PW4168A engine’s inlet, respectively. The computation is done with the average

62
values of the TAT, PT21 and PT22 parameters (see table 6.5) from the stability points. The correction
factors are calculated for standard day conditions using the equations below:

(6.12)

(6.13)

Where is the total temperature at the inlet and is the total pressure at the inlet. The subscript
indicates that the correction factors are computed with the total temperature and total pressure.
Because the inlet does no thermodynamic work [36], the free stream total temperature (TAT) is used
for the total temperature at both engine inlets.

6.3.2.2 Parameter Corrections

In the PW4168A, there are four key engine performance parameters being monitored: EGT, Fuel
Flow, N2 and N1. The recorded parameters from both engines are listed in table 6.5. They are filtered
with equation 6.4 and then averaged in the observation window of the stability point. The values from
each point are corrected for changes in the inlet condition using the equations below, which are similar
to others previously presented in this section.
 Corrected EGT:

(6.14)

 Corrected Fuel Flow:

(6.15)
√
 Corrected N2:

(6.16)
√
 Corrected N1:

(6.17)
√
The subscript denotes the parameter value from the stability point. The EGT is converted from
degrees Celsius to kelvins. The theta correction factor, , is equal in both engines because it is
computed from the TAT. The delta correction factor, , is used to correct the fuel flow and is
computed from the respective total pressure at the engine’s inlet.

63
7. Baseline Model Definition and Trend Monitoring Results

The engine baseline model and trend monitoring results are presented in this Chapter. The source of
engine parameter data are the stability points that are acquired with the algorithm described in the
previous Chapter that was adapted for implementation in AGS. This implementation is the first subject
of the Chapter. After that, the definition of the engine baseline models and the corresponding results
are presented. To derive these models, it was necessary to access the airline records for engine
removals and installations and to take into account the effects of bleed air in the performance
characteristics of the PW4168A engine. The trend monitoring results, showing the evolution in time of
the performance parameter deltas, for two selected aircraft are then presented and compared with the
results obtained with the engine manufacturer’s ECM software. The results shown in this Chapter are
generated with a tool developed with the R programming language that is responsible for establishing
the baseline models from the stability points generated in AGS and calculating the deterioration levels
of the engines for trend monitoring.

7.1 Implementation of Procedures in AGS

The algorithm for the extraction of stability points described in the previous chapter was adapted and
implemented in AGS using dedicated procedures. Table 7.1 lists the procedures that have been
implemented in the database version for TAP’s A330-223 fleet. The implementation of these
procedures is the subject of the current section.
Procedure Nr. Description Execution Rate
504 Filtering of Parameters – ECM 1Hz
505 Filtering and Delta parameters [GVRTI] - ECM 8Hz
506 Filtering and Delta parameters [ROLL] - ECM 2Hz
1000 Basic Condition for cruise stability detection 1Hz
1018 Stable condition for cruise stability detection - ECM 1Hz
8018 Interface for ECM trend monitoring (CRUISE) 1Hz
Table 7.1 List of Procedures for ECM implemented in the A330-223 fleet database version

The flowchart in figure 7.1 is very similar to the one used to describe the algorithm developed in R
(see figure 6.5) and provides visual information about how the procedures are related between them.
The processes are run over the recorded data according to the execution rate and in ascending order.
Procedures 504, 505, 506 filter the required parameters and have different execution rates because of
the different parameter recording rates. While procedure 504 has an execution rate of 1Hz, procedure
505 is executed at 8Hz to filter the GVRTI parameter and procedure 506 at 2Hz to filter the ROLL
parameter. Although the entire process in figure 7.1 only searches for new stability points from second
to second, these procedures ensure that all the recorded values are filtered and that the variation of
the parameters in the observation window is correctly calculated.
Procedure 1000 checks the Basic Conditions in the observation window, the cruise flight phase and if
the initial waiting time is superior to 15 minutes. When these conditions are true, stability points are
searched. Procedure 1018 calculates the variation of the stable frame parameters in the observation
window and checks if they exceed the stability criteria. When the criteria are satisfied, it calculates the

64
quality number (QE) and compares it to the lowest quality number found in the search period.
Whenever the new computed QE is lower, the best stability point for the period is updated.
At the end of each search period or at the end of the flight (if the last search period doesn’t reach the
240 minutes), procedure 8018 outputs a file with information from the best stability point of the last
period. These files are then put together to be used in the derivation of the engine baseline models,
which are described next, and to generate the trend monitoring results.
Filtering of the Parameters 504 505* 506**
* - Execution rate: 8Hz
NO Flight Phase = Cruise ** - Execution rate: 2Hz
1000
(FLIGHT_PHASE=8)?
YES
NO Waiting Time > 15
1000
minutes
YES
NO
Are the Basic Conditions met? 1000

YES

Search for Best Stability Point

NO Are the Stability Criteria

1018
satisfied?
YES

Calculate Quality Number (QE) 1018

QE is the lowest for the Search YES

1018 Update Best Stability Point
Period?

Output File for Engine Trend Monitoring

YES Write file with the Best Stability Point from
End of Search Period/Flight? 1018 8018
the Search Period

To next second

Figure 7.1 Flowchart: Implementation of Procedures for ECM in AGS

7.2 Engine Baseline Model

In Chapter 5, the process of parameter trend monitoring was described as the process by which the
recorded engine data is corrected and then compared to a baseline model of the engine. The
difference in the experienced and expected performance is calculated as a delta parameter and the
evolution of these parameters as a function of time is then used to assess the engine’s condition.
With earlier generation turbofan engines, the expected engine characteristics were made available to
the operators and trend monitoring was conducted by manually calculating the delta parameters.
Figure 7.2 shows the cruise characteristics of the JT3C-6, a two-spool turbojet engine. The EGT, FF,
N1 and N2 are corrected and plotted as a function of the EPR. The changes in the Reynolds number
are also illustrated with the different altitude lines. Nowadays, the recorded data is processed
automatically by ECM tools from the OEMs of the engines and delta parameter trends are immediately
available to the operator. Although these tools use an engine baseline model for these computations,
this information is generally not available to the operator.

65
 Figure 7.2 Cruise Performance Characteristics of the JT3C-6 engine [39]

This section presents the study that was undertaken with the objective of defining the cruise
performance characteristics of the PW4168A engine and compose baseline models from the data
acquired with the algorithms developed. The engine parameter data in the baseline models is
corrected with the dimensionless correction factors and the equations discussed in section 6.3. The
models will be used to perform a trend monitoring analysis totally based on flight data, instead of using
generic engineering models by the manufacturer, which are not known for the case-study engine.

7.2.1 Engine Selection for Baseline

As a starting point, it was necessary to access the records for engine removals and installations on
TAP’s A330-223 fleet. This was done in order to select a set of engines to be used for deriving the
baseline models. From these records, it was possible to see that some of the reasons for engine
removals include insufficient EGT Margin, “Service Convenience” and also damages in the LP and HP
turbines. The engine installations occurred after shop visits with Performance Restoration. In these
visits, the core section and/or other modules of the engine are disassembled and parts are inspected,
balanced and repaired or replaced as necessary. The objective is, as the name indicates, to restore
lost performance and return the engine to on-wing operation.

66
 REMOVED ENGINE DATA INSTALLED ENGINE DATA
A/C Tail Position EGT Date of EGT Date of
Margin Removal Margin Installation
CS-PWA 1 15ºC 09-09-2013 42ºC 10-09-2013
CS-PWA 2 34ºC 26-09-2013 30ºC 26-09-2013
CS-PWB 1 -6ºC 07-03-2013 23ºC 07-03-2013
CS-PWB 2 -8ºC 21-02-2013 34ºC 22-02-2013
Table 7.2 Engine Installations after Shop Visits with Performance Restoration

The two aircraft listed in table 7.2 have been selected for deriving engine baseline models for the fleet.
This selection was mainly based on the EGT Margin of the installed engines. The indicated EGT
Margin is a test cell margin and there are differences to on-wing EGT Margins. Due to TAP
confidentiality policies, the A/C tails correspond to fictitious registrations, which do not exist at the date
this work was done.

7.2.2 Bleed and Pack Selection

Bleed air can have a major effect on the performance curves of an engine and needs to be accounted
for in order to provide accurate estimates. Bleed air corresponds to high pressure air supplied by the
Pneumatic System of the aircraft that is used for several air conditioning, wing anti-icing and other
functions. The air is normally bled from the intermediate pressure stages of the HP compressor, in
order to minimize the fuel penalty. To acquire the stability points in Cruise it is necessary that the wing
and nacelle anti-ice valves are closed, in accordance with the Basic Conditions exposed before.
To take into account the bleed air effects in the ECM analysis, an Aircraft Bleed Code (ABC) and an
Air Conditioning Pack Code (PKS) are computed from the parameters recorded. These codes are
computed from the Engine Cruise Report using the Engine Control Word 1 (ECW1) and then input to
the P&W EHM software [50]. However, the ECW1 is not recorded in the A330-223 aircraft. As an
alternative, the ABC and PKS codes can be computed from the recorded data using the ENGINE
BLEED P/b switch and the PACK FLOW Selection switch discretes. An additional procedure was
implemented in AGS that computes the ABC and PKS codes. The codes are included in the file with
the information on the stability points that is output. Table 7.3 illustrates the logic used.

Cockpit Switch Parameter AGS Status Codes

ENGINE BLEED ABLD_PB_POS (Engine 1) OFF (=0) ABC1=0

Pushbutton Switch
ON (=1) ABC1=1

O_ABLD_PB_POS (Engine 2) OFF (=0) ABC2=0

ON (=1) ABC2=1

PACK FLOW Selection BLD_MIN_PACK (Aircraft) ON (=1) PKS=1

Switch
BLD_NOM_PACK (Aircraft) ON (=1) PKS=2

BLD_MAX_PACK (Aircraft) ON (=1) PKS=3

Table 7.3 Logic for ABC and PKS codes

67
7.2.3 Results
The baseline model for the PW4168A engine derived from the data of CS-PWA aircraft will be
presented next. The points used in the definition of each baseline model represented here were
collected over a period of three months after the respective engines were placed on the wing. This
period was found to be a good compromise between the number of points, the range of conditions
experienced and the gradual deterioration the engine experiences.

7.2.3.1 Baseline Model with EPR

The results in figure 7.3 represent the variation of the corrected cruise performance parameters, as a
function of the EPR, for the two engines on the CS-PWA aircraft. The data from each engine was
collected over the periods of time indicated, with bleed air being extracted from the engines (ABC=1)
and a normal pack flow selection (PKS=2).

(a) (b)

(c) (d)
Figure 7.3 Baseline Results: (a) EGTK vs. EPR (b) FFK vs. EPR (c) N2K vs. EPR (d) N1K vs. EPR

From the analysis of these results, it becomes clear that the EPR is restricted to certain values. This
occurs because the EPR parameters are recorded with a resolution of 0.0625. This results in a great
amount of points for each recorded EPR value. Figure 7.4a highlights the EGTK vs. EPR results for
Engine 1 at an altitude of 40000ft and figure 7.4b at a Mach number of 0.80. In both plots the number
of points is reduced but the scatter in the data is still high, meaning that the points were probably
acquired when the actual EPR was different from the recorded value.

68
 (a) (b)
Figure 7.4 Baseline Results for EGTK vs. EPR at (a) 37000ft and (b) MN=0.80

The results obtained with the EPR as the thrust reference parameter are limited to the recording
accuracy of this parameter, as demonstrated in the previous figures. The excessive scatter in the data
doesn’t allow the derivation of an accurate baseline model, even when the points are confined to a
certain altitude or Mach number. As a consequence, it was necessary to adapt the baseline models to
a new thrust reference parameter.

7.2.3.2 Baseline Model with N1

The N1 was selected as the thrust reference parameter instead. The disadvantage is that the
evolution of the delta N1 parameter is used to monitor the performance deterioration of the PW4168A
engine in the P&W EHM software. With this definition only three performance parameters will be
trended and monitored using the stability points: the EGT, FF and N2. The results in figure 7.5
represent the expected performance characteristics of the PW4168A engine for these three
performance parameters after being corrected and as a function of the N1K. The plots on the left
(figure 7.5a) contain data that was acquired using the Aircraft Stability Criteria and the plots on the
right (figure 7.5b) contain the data acquired with the Converged Stability Criteria (see table 6.8). The
data was collected from the CS-PWA aircraft in the same bleed conditions and over the same periods
of time as those presented in figure 7.3.
Aircraft Stability Criteria Converged Stability Criteria
(202 Stability Points) (187 Stability Points)

69
 (a) (b)
Figure 7.5 CS-PWA Baseline Model Results: (a) Aircraft Stability Criteria (b) Converged Stability Criteria.

The characteristics are much clearer in comparison with those in figure 7.3. The corrected EGT and
FF values in the left and right plots are higher for Engine 2 than for Engine 1. The latter was installed
in the aircraft with 42ºC of EGT Margin and the first with 30ºC, which allows to conclude that the
margin is in fact a good indicator of the relative performance of the engines. The N2 values from both
engines are very close.
Simple linear regression models were fitted to the data from Engine 1 using the least squares
approach. The equations for these models are displayed in the plots from figure 7.5, together with the
fitted model and the r-squared values. The displayed values show that the models obtained with both
criteria fit well the corrected engine data. As expected, the number of points represented in figure 7.5b
is lower due to the more restrictive criteria. Aside from the difference in the number of points
considered, the results obtained with the two criteria are very similar. Nevertheless, the results
presented in the remainder of this Chapter will use the points acquired with the Converged Stability
Criteria, mainly because they eliminate the possibility of acquiring data relative to periods of time when
there was excessive variation in the engine parameters. From the results above, it was also concluded
that it was not essential to consider different model equations for different altitudes to take into
account the changes in the Reynolds’ number.
Figure 7.6 shows the distribution of the stability points by Air Conditioning Pack Code (PKS), during
the period of time from which the baseline model in figure 7.5b was derived. From the 214 points, 187
were acquired with a normal pack flow selection (PKS=2), which corresponds to around 120 flights.

70
 Figure 7.6 Distribution of Baseline Points by PKS Code

Figures 7.7a and 7.7b show the distribution of the stability points from the baseline model by Mach
Number and by altitude and search period, respectively. The Mach numbers range from 0.78 to 0.83
and the most frequent are the 0.80 and the 0.81. The altitudes range from 35000ft to 41000ft and most
of the points from the 2nd and 3rd search periods are acquired at higher altitudes.

(a) (b)
Figure 7.7 Distribution of Baseline Points by (a) Mach Number and (b) Altitude and Search Period

7.3 Trend Monitoring

In this section, we describe the necessary procedures in order to calculate and plot the cruise
performance trends and then present the corresponding results from the two aircraft studied: CS-PWA
and CS-PWB. The trends shown correspond to the evolution in time of the delta parameters
calculated from the difference between the corrected measured data and the baseline models like the
one described in the previous section. The results will be used to test the methodology that was
developed in the previous section and to search for indications of deterioration that could be indicative
of engine failures.

7.3.1 Calculation and Plotting of the Cruise Trends

Figure 7.8 illustrates how the delta parameters are calculated. The measured point contains the
corrected value from the cruise performance parameter at a given corrected thrust (N1K). For this
thrust, a baseline value is calculated using the equation of the model that was derived for the same
71
bleed and pack flow conditions in which the point was acquired. The difference between the parameter
value from the measured and the baseline points yields the delta parameter. The delta is positive
when the point is above the baseline curve and negative when it is under.

Parameter
Cruise
Performance Measured Point
Parameter (Corrected Data)
Baseline Point
Corrected
Corrected Baseline
y=b0+b1x (Adjusted for Bleed and Pack Selection)

N1K

Figure 7.8 Delta Parameter Calculation

Depending on the parameter monitored, the delta parameter can be expressed as an absolute or a
relative difference. The equations used in the results of this section are written below.
 Delta EGT:
(7.1)

 Delta FF:

(7.2)

 Delta N2:

(7.3)

The delta EGT is expressed in degrees Celsius and the delta FF and N2 as percentages.
Smoothed delta parameters are calculated from the raw delta parameters above using equation 5.2
with a smoothing coefficient of 0.2:
(7.4)

The smoothed deltas help to provide more distinct trends of the gradual deterioration an engine is
experiencing.
The calculation of the cruise performance trends is one of the main tasks included the process
represented in figure 7.9, which is executed by a dedicated program written in R. The program
receives the stability points from AGS as input for the analysis – an example of the information
contained in these points is displayed. Firstly, the EGT, FF (WF), N1 and N2 values from each of
these points are corrected with the equations in section 6.3. Then, the Engine Baseline Model is
derived from the corrected engine data of the stability points for the specified time interval (three
months after engine installation, as mentioned previously) and engine. The raw and smoothed cruise
trends are calculated next with equations 7.1 to 7.4 and output in the form of plots. The results are
also output in table format to .csv files for a more detailed analysis.

72
 PRD_STBPT,QE,DATE,TIME,ACREG,TAT,PALT,MN,PKS,ENG_POSITION,ABC,PT2,EPR,N1,N2,EGT,WF,FILE_NO
Stability Points 1,0.220009,02122012,0250,CS-TXX,-39.0,39002,0.805,1,1,1,4.38,1.313, 81.0, 83.1,380.0,2584.9,682389
1,0.220009,02122012,0250,CS-TXX,-39.0,39002,0.805,1,2,1,4.37,1.313, 80.9, 83.0,382.1,2573.7,682389

Calculate the Correction Factors

and Corrected Parameters

Definition of the
Engine Baseline Model EGTK,FFK,N2K=f(N1K,ABC,PKS)

Calculate the Cruise Trends

Plot the Delta Performance

Raw Deltas Smoothed Deltas
Parameters

Figure 7.9 Process of Plotting the Cruise Performance Trends

7.3.2 Results

7.3.2.1 CS-PWA
Figure 7.10 is a timeline with the engines’ removals and installations on the CS-PWA aircraft that are
relevant for the analysis. It indicates that both engines were substituted during September 2013.
10-09-2013 Engine installed on
left wing (EGT Margin: 42ºC)
09-09-2013 Engine removed from
left wing (EGT Margin: 15ºC)

Sep Oct
2013 2013
26-09-2013 Engine removed from right
wing (EGT Margin: 34ºC)
26-09-2013 New Engine installed on
right wing (EGT Margin: 30ºC)

Figure 7.10 CS-PWA: Timeline for Engine Removals/Installations

Both the raw and smoothed delta parameters are represented in each one of the plots in figure 7.11,
for the period from July 2013 to March 2014. The baseline model used for the comparison of the data
is specified in the title of these plots. In this case, it is the same model from engine nr.1 whose results
were presented in figure 7.5b.
Figure 7.11a contains the plot for the delta EGT. The engine substitutions are clearly illustrated by the
th
step shifts in the deltas during September 2013. The smoothing always starts with the 10 stability
point acquired after a new engine installation. The aircraft was out of service during December, which
resulted in some breaks in the data from both engines. Because the baseline model was derived from
engine nr.1 in the period from 2013-09-10 to 2014-01-10, the delta EGT values in this period stay
close to zero, as expected. The engines with higher EGT Margins show better performance levels, i.e.,
73
lower deltas. This is true for the four engines represented. It is interesting to notice the increase in the
delta EGT after engine nr.2 was substituted. In this case, the engine was replaced for “Service
Convenience” and not because there were problems with its operation.

Breaks in the data

Smoothing starts after

the 10th point

(a)

(b)

(c)
Figure 7.11 Cruise Performance Trends for the CS-PWA aircraft: (a) Delta EGT; (b) Delta FF; (c) Delta N2.

Figure 7.11b contains the results for the delta FF. There is more scatter in the data than that from the
delta EGT plot. The engine substitutions are not as evident as in figure 7.11a and there is a small
74
reduction in the delta values after engine nr.1 was replaced. In figures 7.11a and 7.11b it is possible to
observe some gradual deterioration in the performance of the engine with the smoothed delta EGT
and smoothed delta FF parameters.
Finally, figure 7.11c contains the results for the delta N2. The values are almost all within the range of
-0.5% to 0.5%. However, the engine substitutions are much visible here than in figure 7.11b. After the
engine installations indicated in figure 7.10, the delta N2 values from engine nr.1 were higher than
those from engine nr.2. This is in contrast with what occurred with the delta EGT and delta FF
parameters.

7.3.2.2 CS-PWB
The timeline for the engines’ removals and installations on the CS-PWB aircraft is represented in
figure 7.12 and indicates that engine nr.1 was substituted once and engine nr.2 twice during the period
between February and June 2013.
07-03-2013 Engine removed from
left wing (EGT Margin: -6ºC)
07-03-2013 Engine installed on
left wing (EGT Margin: 23ºC)

Feb Mar Apr May June

2013 2013 2013 2013 2013
21-02-2013 Engine removed from 11-06-2013 Engine installed on
right wing (EGT Margin: -8ºC) right wing (EGT Margin: 20.4ºC)
22-02-2013 Engine installed on 25-05-2013 Engine removed from
right wing (EGT Margin: 34ºC) right wing (EGT Margin: 30ºC)
Figure 7.12 CS-PWB: Timeline for Engine Removals/Installations

The plots in figure 7.13 contain the raw and smoothed delta parameters from December 2012 to
February 2014. The flights from engine nr.2 after the installation in February 2013 were used for the
baseline model, which is represented in Appendix I.
Figure 7.13a contains the results for the delta EGT. The initial deltas are very high for both engines.
When engine nr.1 was replaced in February 2013 and engine nr.2 in March 2013, they presented
negative EGT Margins. A few months after new engines were installed the aircraft was out of service
for maintenance for about two months. During this period, engine nr.2 was replaced again by an
engine with a lower EGT Margin than the previous, which led to higher delta values.
There are two step shifts, which occur outside the dates for the engine removals/installations indicated
in figure 7.12. Both shifts are characterized by an increase in the delta EGT and delta FF and a
decrease in the delta N2K. The first occurs in the beginning of August for engine nr.2 and the second
in mid-October for engine nr.1. The two situations were detected with the P&W EHM software and
were related with faults in the Turbine Case Cooling (TCC) System. The TCC System controls and
distributes fan air to externally cool the high and low pressure turbines cases to [46]: increase case life
and reduce turbine blade tip clearance1 during take-off, climb and cruise operation for better fuel

1
The tip clearance is the distance between the blade-tip and the engine casing.
75
efficiency. The system is basically made up of three components: actuator, cables and valves. The
diagnosis of the two faults and the corresponding maintenance actions that were conducted in order to
restore the performance of the engine are described next. The data obtained with the P&W EHM is
also presented for comparison with the results in figure 7.13.

Engine Nr.2
Step Shift

Engine Nr.1
Out for Maintenance Step Shift

(a)

(b)

(c)
Figure 7.13 Cruise Performance Trends for the CS-PWB aircraft: (a) Delta EGT; (b) Delta FF; (c) Delta N2.

76
Engine #1 Shift
In addition to being detected with the P&W EHM software, the ECAM system of the aircraft also
detected a fault in the TCC system that was reported by the flight crew, which indicated a failure in the
HPTCC (High Pressure Turbine Case Cooling) actuator. The actuator is the only component of the
TCC system that provides feedback to the EEC (Engine Electronic Control). Thus, the failures in the
valves and cables can only be effectively monitored with the P&W EHM software.
The failure in the actuator was not confirmed by maintenance personnel. Even so, the replacement of
the actuator was scheduled and an inspection of the complete TCC system was ordered by the
Powerplant Engineering department. During this inspection, only a faulty HPTCC cable was found
and, as a result, it was decided to replace this cable instead of the actuator. The Long Term Plot
Report for this engine is displayed in figure 7.14. The evolution of the EGT, FF (WF), N1 and N2
smoothed deltas and EGT Margin is displayed together with an indication of the date in which the
cable was replaced. Notice that the FF data is more scattered than the remaining engine data. The
same thing happened in the results from figures 7.11b and 7.13b.

Figure 7.14 P&W EHM: CS-PWB Engine #1 Long Term Trend Plot Report

The shift is characterized by an increase in the delta EGT and delta FF, similarly to what is displayed
in figures 7.13a and 7.13b, respectively. There is a small increase in the delta N1 but the delta N2 is
practically unaffected. In figure 7.13b there is a decrease in the delta N2 levels. In this case, the ECM
analysis and the corresponding maintenance actions (inspections, repair, etc.) didn’t allow the engine
to fully restore the lost performance. The biggest improvement occurs in the EGT Margin, which
dropped by approximately 5ºC.

77
Engine #2 Shift
This shift also occurred due to a fault in the TCC system, as mentioned previously. In this case, the
variation of the delta parameters indicated a clear and distinct loss in the performance of the engine,
as illustrated by the Long Term Plot Report for engine #2 in figure 7.15. Having into account the
manufacturer fingerprints for the PW4168A engine and the skills built from previous experience,
powerplant engineers have ordered the inspection of the TCC system. From this inspection,
maintenance personnel have found a faulty HPT valve which was promptly substituted and the engine
returned to the previous performance. This situation is clearly illustrated in the figure below.

Figure 7.15 P&W EHM: CS-PWB Engine #2 Long Term Trend Plot Report

The shift in figure 7.15 is characterized by an increase in the delta EGT and delta FF, and a decrease
in the delta N2. The same pattern can be observed in the results from figure 7.13. The most
remarkable aspect is that the shifts in figures 7.13 and 7.15 are practically identical: +10ºC in the delta
EGT, +0.5% in the delta FF and approximately -0.25% in the delta N2. In addition, the deterioration
levels of the engine before the failure and after maintenance work was conducted, are similar in both
figures. This might indicate that the performance of the engine used to derive the baseline model from
the data is similar to the baseline performance of the engine used by the manufacturer software.

78
8. Conclusions

For an airline, tracking engine operations is of the utmost importance, both from a safety and an
economic point of view. The current dissertation is part of this everyday effort from people within TAP
Portugal. It makes the best use of the engine data available from on-board recorders of a fleet of
aircraft equipped with PW engines to produce tools for ECM performance trend monitoring.

8.1 Achievements
The first part of this dissertation dealt with the implementation of an FDM database for the decoding
and analysis of flight data that is recorded on-board an Airbus A310. The aircraft and its recording
systems were presented in Chapter 2, as well as the data transmission and recording protocols used
in the aircraft. The programming of the database in AGS was the subject of Chapter 3 and required
the knowledge of the previous protocols. In Chapter 4, the decoding results for a selection of
parameters were displayed and allowed to confirm the extreme quality of the recorded flight data. The
limitations in the recordings of the QAR became evident when the Flight Phase computation
procedure was described, since it was necessary to compute additional parameters and to adapt
some of the transitions according to the parameters available. All these topics were covered and the
results that were displayed show the correct implementation of these procedures in the database,
which can be expanded to include FOQA and MOQA procedures.
The second part of the dissertation (Chapter 5, 6 and 7) focused on the main objective of the work
which was the development of tools for performance trend monitoring, within the scope of Engine
Condition Monitoring. In Chapter 5, the role of ECM in the maintenance of aircraft engines was
presented and the importance of trend monitoring in the diagnosis and prognosis of engine faults, as
well as in the support of engine management decisions, was discussed. The necessary ECM concepts
and definitions to understand the upcoming results were also introduced in this Chapter.
In Chapter 6, the PW4168A engine was briefly presented. The algorithm that processes the recorded
flights and acquires the stability points was described and the results that were shown enabled to
understand this process. Afterwards, the rationale of the engine parameter corrections was exposed
and the corresponding equations applied to the case study engine were presented. The two
algorithms/procedures were then used to generate the results in Chapter 7.
The initial engine baseline model results in Chapter 7 underlined the necessity to improve the DAR
parameter recordings of the EPR, which in turn required the N1 to be used as the thrust reference
parameter. Afterwards, the results showed clear characteristics for the performance of the engine and
simple linear regressions were fitted to the data. The models obtained with two different stability
criteria were similar and it was not necessary to take additional variables - like altitude or speed – into
account. The trend monitoring results allowed evaluating the methodology of the work developed. In
the results from the first of two aircraft, it was possible to observe some gradual deterioration in both
engines, occurring over a period of several months. In addition, the deterioration levels of the engines
with higher EGT Margins were lower and showed that this parameter is a good indicator of the overall
efficiency of the engine. In the second aircraft, both engines evidenced a shift in the delta parameters

79
occurring at different times. The engines’ maintenance records and the results obtained with the P&W
EHM software were then collected and the relevant information presented. They allowed confirming
that the observed shifts were, in fact, indicative of engine faults. In both situations, the shift patterns in
the plots from the P&W EHM and those obtained from the flight data were similar. For the engine #2
fault, the variations in the EGT, FF and N2 deltas were practically identical.

8.2 Future Work

The methodology that was followed in this work relied exclusively on recorded flight data to create
engine baseline models and generate the performance trends by comparing engine data with these
models. The results that were presented have confirmed the potential of this approach, which can be
used as a complement to the engine’s manufacturer ECM software. Powerplant engineers can use the
tools that were developed for an improved insight of the engine’s operation.
One direct evolution from this work is to apply the same methodology to the CFM56 and CF6 engines
that equip the remaining of TAP fleet. The tools that were developed with the R programming
language may be adapted for this purpose. However, this requires a previous study of the aircraft and
its aircraft systems, namely the stability criteria used and the recorded parameters that are required.
During the course of this work it was identified the need to improve some of the on-board recorded
data. Once this is done, better data will be available that can be used to improve the whole process.
This work is part of a continuous improvement cycle, always aiming at the maximum safety. Future
actions may also include the use of statistical methods and data-mining techniques to improve the
detection of trends out of the data, based on manufacturer data or requirements from the airline’s
engineering departments.

80
9. References

[1] ICAO, "Safety Report - 2014 Edition," 2014.

[2] Airbus, "Future Journeys 2013 2032," Global Market Forecast, 2013.

[3] Euclides da Conceição Pereira Batalha, "Aircraft Engines Maintenance Costs and Reliability - An
Appraisal of the Decision Process to Remove an Engine for a Shop Visit Aiming at Minimum
Maintenance Unit Cost," Instituto Superior de Estatística e Gestão da Informação, 2012.
[4] Daniel Cardinha, "Flight Data Monitoring Process implementation within an airline's A330-223
fleet," Instituto Superior Técnico, MSc Thesis 2010.

[5] André Seabra de Oliveira Simões, "Programming/Decoding of flight data from an airline's A320-
200 fleet - Practical applications and optimization of flight analysis," Instituto Superior Técnico,
MSc Thesis 2010.

[6] Cátia Joana Lopes Palmeiro, "Estudo e desenvolvimento de um algoritmo de detecção de pontos
de estabilidade," Instituto Superior Técnico, MSc Thesis 2009.
[7] Shannon Ackert, "Engine Maintenance Concepts for Financiers," Macquarie AirFinance, 2011.
[8] IATA, "Airline Maintenance Cost Executive Commentary (FY2012 data)," November 2013.

[9] Airbus. (2014, March) Airbus A310. [Online]. http://www.airbus.com/aircraftfamilies/out-of-

production/a310/

[10] Airbus, A310 Flight Crew Operating Manual - Procedures and Performance, vol. II.

[11] Airbus, A310 Flight Crew Operating Manual - Systems Description, vol. I.

[12] Condor Engineering, "ARINC Protocol Tutorial," 2000.

[13] Rockwell Collins. The History of Rockwell Collins’ ARINC. [Online].

http://www.arinc.com/about/history.html

[14] ARINC, "ARINC Specification 429 Part 1-17: Mark 33 – Digital Information Transfer System
(DITS)," 2004.

[15] Howard Curtis et al., Aerospace Engineering Desk Reference. Oxford, United Kingdom: Elsevier,
2009.

[16] Bureau d'Enquêtes et d'Analyses (BEA), "Flight Data Recorder Read-Out - Technical and
Regulatory Aspects," Study 2005.

[17] Airlines for America. (2014, May) About A4A. [Online]. http://www.airlines.org/Pages/About.aspx
[18] Inmedius, "From Paper to Interactive Electronic Technical Publications (IETP) - The ATA iSpec
2200 Standard," 2012.

[19] Airlines for America. (2014, May) Spec 100: Manufacturers' Technical Data. [Online].
https://publications.airlines.org/commerceproductdetail.aspx?product=33
[20] Wikipedia. (2014, May) ATA 100. [Online]. http://en.wikipedia.org/wiki/ATA_100

[21] Airbus, A310 - Aircraft Maintenance Manual, Revision NO. 73, Jun 01/13.

81
[22] Teledyne Controls, "Functional Specification of Digital Flight Data Acquisition Unit (DFDAU) for
Airbus A300 and A310 Aircraft," Los Angeles, California, 1993.

[23] Sagem, Analysis Ground Station V14 - User Manual. Paris, France: SAFRAN, 2012.

[24] Sagem, Analysis Ground Station V12 - Method for Database Programming. Paris, France:
SAFRAN, 2008.

[25] Mustafa Cavcar, "The International Standard Atmosphere (ISA)," Anadolu University, Turkey,
2000.

[26] Jan Roskam and Chuan-Tau Edward Lan, Airplane Aerodynamics and Performance. Lawrence,
KS: DARcorporation, 1997.
[27] Airbus, Flight Data Recording Parameter Library.

[28] TAP Portugal, Comandos de Voo: A310 Vol. II Flaps, Spoilers, Speed Brakes, Slats, Krueger
Flap. Lisboa, Portugal: Formação Profissional.

[29] Frank H King, Aviation Maintenance Management. Carbondale, IL: SIU Press, 1986.

[30] General Electric, "Diagnostics Trend Interpretation Training," 2008.

[31] Aviation PLATFORM. (2014, June) AMC I: THE AIRCRAFT MAINTENANCE BASIC
PHILOSOPHIES, HARD TIME, ON CONDITION AND CONDITION MONITORING. [Online].
http://www.aviationplatform.com/index.php/amc/89-amc-i-the-aircraft-maintenance-basic-
philosophies-hard-time-on-condition-and-condition-monitoring

[32] Irem Y Tumer and Anupa Bajwa, "Learning about how aircraft engines work and fail," AIAA Paper
AIAA-99-2850, 1999.

[33] S K Sethiya, "Condition Based Maintenance (CBM)," Secy. to CME/WCR/JBP, 2006.

[34] R. Vingerhoeds, P. Janssens, B. Netten, and M. Aznar Fernandez-Montesinos, "Enhancing off-
line and on-line condition monitoring and analysis," Control Engineering Practice, vol. 3, no. 11,
pp. 1515-1528, 1995.
[35] Herbert Ian Howard Saravanamuttoo, Gordon Frederick Crichton Rogers, and Henry Cohen, Gas
Turbine Theory. Essex, England: Longman, 1996.

[36] João Teixeira Borges, "Acetatos de Propulsão," Instituto Superior Técnico, Class Lectures 2002.

[37] CFM, "CFM56 - Flight Operations Support," 2005.

[38] Wikipedia. Turbofan Operation. [Online].
http://en.wikipedia.org/wiki/Turbofan#mediaviewer/File:Turbofan_operation.svg
[39] Pratt & Whitney Aircraft, "Malfunction Analysis Through Gas Generator Comparisons and
Mechanical Trends," Aircraft Gas Turbine Operation Information Letter 1965.
[40] GE Aircraft Engines, CF-6 - Guidelines for Parameter Trend Monitoring, 1993, Transmittal Sheet.

[41] Donald L Simon, "An integrated architecture for on-board aircraft engine performance trend
monitoring and gas path fault diagnostics," NASA/TM—2010-216358, 2010.

[42] Arnaud Bonnet, "Avoiding high speed rejected takeoffs due to EGT limit exceedance," Safety First

82
 - The Airbus Safety Magazine, no. 4, pp. 8-13, 2007.

[43] Maumy O and Mungin B, "EGT Margin A300/CF6-50C2," FAST Magazine, no. 9, pp. 9-13, 1988.
[44] Pratt & Whitney, "PW4000 100-Inch Fan Propulsion System: The Advantage for the A330
Aircraft," Fact Sheet.

[45] Airbus, Getting to Grips with Aircraft Performance Monitoring, 2002.

[46] Airbus, A330 - Aircraft Maintenance Manual, 2014.

[47] Allan J Volponi, "Gas turbine parameter corrections," Journal of engineering for gas turbines and
power, vol. 121, no. 4, pp. 613-621, 1999.

[48] Joachim Kurzke, "Model based gas turbine parameter corrections," in ASME Turbo Expo 2003,
collocated with the 2003 International Joint Power Generation Conference, Atlanta, 2003, pp. 91-
99.

[49] Edgar Buckingham, "On physically similar systems; illustrations of the use of dimensional
equations," Physical Review, vol. 4, no. 4, pp. 345-376, 1914.

[50] Pratt & Whitney, P&W EHM Installation Manual, 2003.

[51] Airbus, Getting to Grips with Weight and Balance.

[52] Airbus, A310 - Airplane Characteristics For Airport Planning, Revision NO. 21, Dec 01/09.

83
Appendix A. ARINC 429 Data Sources

Equip.
System Data Source (Port [22]) Description
ID
01 FCC-1 (#11), 2 (#12) Flight Control Computer
Automatic Flight
03 TCC-1 (#4), 2 (#4) Thrust Control Computer
System
A1 FCU (#25) Fuel Control Unit
31 GMT-CLK (#7) GMT Clock
Independent
29 SDAC (#14) System Data Analog Converter
Instruments
26 FWC-1 (#13) Flight Warning Computer
System
23 GPWS (#23) Ground Proximity Warning System
Instrument Landing System
10 ILS-1 (#8), 2 (#35)
Receiver
07 RALT-1 (#9) Radio Altimeter
VHF Omnidirectional Range
11 VOR-1 (#10), 2 (#24)
Receiver
02 FMC-1 (#16), 2 (#36) Flight Management Computer
FNSG-1 (#15), 2 (#33), Flight Navigation Symbol
Navigation System 25
3 (#15 or #33) Generator
09 DME-1 (#29), 2 (#30) Distance Measuring Equipment
Traffic Alert and Collision
35 TCAS (#26)
Avoidance System
C5 FNCP-1 (#31), 2 (#34) Flight Navigation Control Panel
06 ADC-1 (#17) Air Data Computer
04 IRS-1 (#20), 3 (#20) Inertial Reference System
1A EEC-1 (#1) Electronic Engine Control
Engine Fuel and
05 FQI (#3) Fuel Quantity Indicator
Control Systems
3E CGCC (#5) Center of Gravity Control Computer
Table A.1 ARINC 429 Data Sources in database version 10079

84
Appendix B. List of Recorded Parameters

Analog Input Parameters

Name Description Name Description
BRAKE_LPA Brake Pedal Deflection LH SLAT Slat position (calibrated)
BRAKE_RPA Brake Pedal Deflection RH FLAP Flap Position (calibrated)
VRTG Normal (Vertical) Acceleration SPD_BRK_HDL Speed Brake Handle
LONG Longitudinal Acceleration FF2 Fuel Flow Engine #2
LATG Lateral Acceleration FF1 Fuel Flow Engine #1
Table B.1 List of Analog input parameters

Discrete Input Parameters

Name Description Name Description
HF HF keying AC_IDT A/C number**
VHF VHF keying FLIDENT Fleet identification***
REV_INPOS1 Reverser In Position Engine #1 ECAMDU_SYS_OFF ECAM DU System Off
REV_UNLOCK_1 Reverser Unlock Engine #1 ECAMDU_SYS_FLT ECAM DU System Fault
REV_INPOS2 Reverser In Position Engine #2 ECAMDU_WRN_OFF ECAM DU Warning Off
REV_UNLOCK_2 Reverser Unlock Engine #2 ECAMDU_WRN_FLT ECAM DU Warning Fault
Wing Anti-Ice Valve LH Alternate
HYD_L_PRES_Y Hydr System YELLOW Low Press AIW_ALTR1
UNCL
Wing Anti-Ice Valve LH Normal
HYD_L_PRES_G Hydr System GREEN Low Press AIW_NORM1
UNCL
Wing Anti-Ice Valve RH Normal
HYD_L_PRES_B Hydr System BLUE Low Press AIW_NORM2
UNCL
Wing Anti-Ice Valve RH Alternate
ECAMCP_4 ECAM CP Code 4 AIW_ALTR2
UNCL
ECAMCP_3 ECAM CP Code 3 EVENT Event (Depressed)
Inlet Anti-Ice Valve Engine #2
ECAMCP_2 ECAM CP Code 2 AIE2
OPEN
Inlet Anti-Ice Valve Engine #1
ECAMCP_1 ECAM CP Code 1 AIE1
OPEN
TA_ALERT Traffic Alert START_VLV_2 Start Valve Pos. Engine #2
SPOIL_GND_ARM Ground Spoiler Not Armed START_VLV_1 Start Valve Pos. Engine #1
AC1_BUS_OFF AC Bus 1 OFF LDG_ON_1 Landing Gear Squat Switch LH
LDG_SEL_NUP Landing Gear Selector Not Up LDG_ON Landing Gear Squat Switch NOSE
LDG_SEL_NDW Landing Gear Selector Not Down LDG_ON_2 Landing Gear Squat Switch RH
AC_TYP A/C type*
*Discretes 82 to 87; **Discretes 92 to 99; Discretes 88 to 91
Table B.2 List of Discrete input parameters

Digital Data Inputs (DITS) ARINC 429

 Automatic Flight System:
Flight Control Computer (FCC)
Name Description Name Description
FD_1 FD #1 Engaged YAW_TO Yaw Takeoff Mode
AP1_CWS A/P CWS #1 LNDTRA1 Landing Track
AP_EGD1 A/P CMD #1 FD_2 FD #2 Engaged
AFCS Longitudinal Modes (Matrix
AFCS_LONG_MOD AP2_CWS A/P CWS #2
#1)
AFCS_LAT_MOD AFCS Lateral Modes (Matrix #2) AP_EGD2 A/P CMD #2
PITCH_GA Pitch Go Around (GA) Mode
Table B.3 List of Digital input parameters: Flight Control Computer (FCC)

85
 Thrust Control Computer (TCC)
Name Description Name Description
ATS_EGD Autothrottle Mode Engaged ATS_MACH Mach Mode Engaged
MTHR Manual Throttle Armed ATS_EPR Autothrottle Thrust EPR Mode
ATS_ALVR ATS Arming Lever Off ATS_ALPHA_FLR Alpha Floor Mode Engaged
VNAV_MODE VNAV Mode – Profile Engaged ATS_RETARD Retard Mode Engaged
ATS_CL ATS Clutches Off (Takeoff) ATS_SPD Speed Mode Engaged
Table B.4 List of Digital input parameters: Thrust Control Computer (TCC)

Flight Control Unit (FCU)

Name Description
ALT_SEL Selected Altitude (Manual)
SPD_SEL Selected Airspeed (Manual)
SPEEDV_SEL Selected Vertical Speed
HEAD_SEL Selected Heading
MACH_SEL Selected Mach
Table B.5 List of Digital input parameters: Flight Control Unit (FCU)

 Independent Instruments System:

GMT Clock
Name Description
UTC_HOUR GMT/UTC (Hour)
DAT_DAY Date (Day)
DAT_MONTH Date (Month)
UTC_MIN_SEC GMT/UTC (Minutes, Seconds)
Table B.6 List of Digital input parameters: GMT Clock

System Data Analog Converter (SDAC)

Name Description Name Description
ELEV Elevator Position SPOIL_L3_R Spoiler 3 LH Ret
SPOIL_R2_R Spoiler 2 RH Ret SPOIL_L2_R Spoiler 2 LH Ret
SPOIL_R1_R Spoiler 1 RH Ret SPOIL_L1_R Spoiler 1 LH Ret
SPOIL_L7_R Spoiler 7 LH Ret RUDD Rudder Position
SPOIL_L6_R Spoiler 6 LH Ret AIL_2 All Speed Aileron RH Position
SPOIL_L5_R Spoiler 5 LH Ret XFEED X Feed Valve Closed
SPOIL_L4_R Spoiler 4 LH Ret AC2_BUS_OFF AC Bus 2 OFF
SPOIL_R5_R Spoiler 5 RH Ret PCKM_MODSEL2 Pack Manual Mode Select #2
SPOIL_R4_R Spoiler 4 RH Ret PCKM_MODSEL1 Pack Manual Mode Select #1
SPOIL_R3_R Spoiler 3 RH Ret OIQ_1 Oil Quantity Engine #1
SPOIL_R7_R Spoiler 7 RH Ret OIP_1 Oil Pressure Engine #1
SPOIL_R6_R Spoiler 6 RH Ret OIT_1 Oil Temperature Engine #1
AIL_1 All Speed Aileron LH Position OIQ_2 Oil Quantity Engine #2
EG2FV Fire Valve Eng 2 Not Fully Closed OIP_2 Oil Pressure Engine #2
EG1FV Fire Valve Eng 1 Not Fully Closed OIT_2 Oil Temperature Engine #2
BLD_VLV2 Bleed Valve Open Eng 2 VIB_N1FNT1 N1 Vibration Engine #1 (EVM)
BLD_VLV1 Bleed Valve Open Eng 1 VIB_N2FNT1 N2 Vibration Engine #1 (EVM)
APUBV APU Bleed Valve Not Fully Open VIB_N1FNT2 N1 Vibration Engine #2 (EVM)
STAB Stabilizer Position VIB_N1FNT2 N2 Vibration Engine #2 (EVM)
Table B.7 List of Digital input parameters: System Data Analog Converter (SDAC)

86
 Flight Warning Computer (FWC)
Name Description Name Description
LND_ASAP Land ASAP SMOKE Smoke Warning
AFT_CG_WARN Aft CG Warning GEAR_N_DW_LOK Gear Not Locked Down
FLVR2_C Fuel Lever Engine #2 Closed STALW Stall Warning
FLVR1_C Fuel Lever Engine #1 Closed CBAW Cabin Altitude (Warning)
FIRE_APU APU Fire FLAP_FAULT Flaps 1+2 Fault
DC_BUS_OFF DC Normal Bus OFF SLAT_FAULT Slats 1+2 Fault
SP5_FAULT Spoiler 5 Fault/Off YAW_FAUL2 Yaw Damper 2 Fault
SP32_FAULT Spoiler 3+2 Fault/Off YAW_FAUL1 Yaw Damper 1 Fault
SP14_FAULT Spoiler 1+4 Fault/Off PT2F Pitch Trim 2 Fault
SP7_FAULT Spoiler 7 Fault/Off PT1F Pitch Trim 1 Fault
SP6_FAULT Spoiler 6 Fault/Off OVS Overspeed (VFE, VLE, VMO, MMO)
FIRE2 Engine #2 Fire CONFW Configuration (Warning)
FIRE1 Engine #1 Fire LOW_PRESS_2 Engine #2 Low Oil Pressure
VMO_MMO_OVS VMO/MMO Overspeed LOW_PRESS_1 Engine #1 Low Oil Pressure
Table B.8 List of Digital input parameters: Flight Warning Computer (FWC)

Ground Proximity Warning System (GPWS)

Name Description
GPWS GPWS Modes Warning
Table B.9 List of Digital input parameters: Ground Proximity Warning System (GPWS)

 Navigation System:
Instrument Landing System Receiver (ILS)
Name Description
ILS_FRQ1 ILS-1 Frequency
ILS_FRQ2 ILS-2 Frequency
LOC_DEV1 ILS-1 Localizer Deviation
LOC_DEV2 ILS-2 Localizer Deviation
GLID_DEV1 ILS1 glideslope deviation
GLID_DEV2 ILS2 glideslope deviation
Table B.10 List of Digital input parameters: Instrument Landing System Receiver (ILS)

Radio Altimeter
Name Description
RALT1 Radio Height #1
Table B.11 List of Digital input parameters: Radio Altimeter

VHF Omnidirectional Range Receiver (VOR)

Name Description
Marker Beacon Passage (Outer OR
BEA_MK
Middle OR Inner)
VOR_FRQ1 VOR-1 Frequency
VOR_FRQ2 VOR-2 Frequency
SEL_CRS1 Selected Course #1
SEL_CRS2 Selected Course #2
Table B.12 List of Digital input parameters: VHF Omnidirectional Range Receiver (VOR)

Flight Management Computer (FMC)

Name Description Name Description
SPD_TGT Airspeed target EPR_TARGET EPR Target
FLTNUM Flight Number (LSH) BTW Bearing to Waypoint
FLTNUM Flight Number (MSH) LATP Pres. Pos Latitude (COARSE)
GW Gross Weight (LSH) LATP Pres. Pos Latitude (FINE)
GW Gross Weight (MSH) LONP Pres. Pos Longitude (COARSE)
FMC Database Update (Day-
NDBUPD LONP Pres. Pos Longitude (FINE)
Month-Cycle)*
FMC Database Update (Day-
NDBUPD MACH_TGT Mach Target
Month-Cycle)**
*Day; **Month
Table B.13 List of Digital input parameters: Flight Management Computer (FMC)

87
 Flight Navigation Symbol Generator (FNSG)
Name Description Name Description
AOAL Angle of attack (ADC 1) ROLL Roll angle
PFD_CPT_FAULT Pilot PFD Fault RALT2 Radio Height #2
ND_CPT_FAULT Pilot ND Fault AOAR Angle of attack (ADC 2)
PFD_FO_FAULT F/O PFD Fault GPWS_WSH_WAR Windshear warning
ND_FO_FAULT F/O ND Fault TAT Total Air Temperature
CP_CPT_FAULT Pilot CP Fault DH_SEL_CPT Selected DH Pilot (COARSE)
FNSG13_FAULT FNSG 1 or 3 Fault DH_SEL_FO Selected DH F/O (COARSE)
Altitude Standard (1013 mb)
CP_FO_FAULT F/O CP Fault ALT_STD
(COARSE)
FNSG23_FAULT FNSG 2 or 3 Fault DH_SEL_CPT Selected DH Pilot (FINE)
HEAD_MAG Magnetic Heading DH_SEL_FO Selected DH F/O (FINE)
IAS Computed airspeed ALT_STD Altitude Standard (1013 mb) (FINE)
PITCH Pitch angle
Table B.14 List of Digital input parameters: Flight Navigation Symbol Generator (FNSG)

Distance Measuring Equipment (DME)

Name Description
DME1_DIST DME distance-left
DME2_DIST DME distance-right
Table B.15 List of Digital input parameters: Distance Measuring Equipment (DME)

Traffic Alert and Collision Avoidance System (TCAS)

Name Description
TCAS_SENS_3 TCAS Sensivity Level (bit 17)
TCAS_SENS_2 TCAS Sensivity Level (bit 16)
TCAS_SENS_1 TCAS Sensivity Level (bit 15)
TCAS_RA Resolution Advisory
Table B.16 List of Digital input parameters: Traffic Alert and Collision Avoidance System (TCAS)

Flight Navigation Control Panel (FNCP)

Name Description Name Description
EFIS Mode Displayed on ND EFIS Mode Displayed on ND F/O
ND_MOD1_CPT ND_MOD2_FO
Captain (ILS VOR NAV) (PLAN ROSE ARC MAP)
EFIS Mode Displayed on ND F/O EFIS Mode Displayed on ND F/O
ND_MOD1_FO ND_NM_FO
(ILS VOR NAV) (NM)
EFIS Mode Displayed on ND
ND_MOD2_CPT FPA_SEL_CPT Selected Flight Path Angle Pilot
Captain (PLAN ROSE ARC MAP)
EFIS Mode Displayed on ND
ND_NM_CPT FPA_SEL_FO Selected Flight Path Angle F/O
Captain (NM)
Table B.17 List of Digital input parameters: Flight Navigation Control Panel (FNCP)

Air Data Computer (ADC)

Name Description
ALT_CPT Pilot Altitude BARO Set (FINE)
ALT_FO F/O Altitude BARO Set (FINE)
ALT_CPT Pilot Altitude BARO Set (COARSE)
ALT_FO F/O Altitude BARO Set (COARSE)
Table B.18 List of Digital input parameters: Air Data Computer (ADC)

Inertial Reference System (IRS)

Name Description
DRIFT Drift Angle
WIN_DIRR Wind Direction
WIN_SPDR Wind Speed
GS Ground Speed
Table B.19 List of Digital input parameters: Inertial Reference System (IRS)

88
 Engine Fuel and Control Systems:
Electronic Engine Control (EEC)
Name Description Name Description
EPR1 EPR Actual Engine #1 EPR2 EPR Actual Engine #2
TLA1 Throttle Resolver Angle Engine #1 EGT1 EGT Engine #1
TLA2 Throttle Resolver Angle Engine #2 N21 N2 Engine #1
EPR_CMD_ENG2 Thrust (EPR) Command Engine #2 EEC2_FAULT EEC 2 Fault
EPR_CMD_ENG1 Thrust (EPR) Command Engine #1 EEC1_FAULT EEC 1 Fault
EPR2 EPR Actual Engine #2 N22 N2 Engine #2
EGT2 EGT Engine #2
Table B.20 List of Digital input parameters: Electronic Engine Control (EEC)

Fuel Quantity Indicator (FQI)

Name Description
FQTT Trim Tank fuel Quantity
Table B.21 List of Digital input parameters: Fuel Quantity Indicator (FQI)

Center of Gravity Control Computer (CGCC)

Name Description
CG Center of Gravity
Table B.22 List of Digital input parameters: Center of Gravity Control Computer (CGCC)

The full ARINC429 and ARINC 717/573 characteristics for the listed parameters are available under
request to [email protected]

89
Appendix C. Superframe Words

Note: Parameter name between brackets [ ]

Superframe #1
Subframe 1, Word 56
Cycle 12 11 10 9 8 7 6 5 4 3 2 1
1 0 0 0 0 0 0 0 0 0 0 0 0
2 0 0 0 1 Trim Tank Fuel Quantity [FQTT]
3 0 0 1 0 0 0 0 0 0 0 0 0
4 0 0 1 1 Oil Quantity Engine #1 [OIQ_1]
5 0 1 0 0 Oil Pressure Engine #1 [OIP_1]
6 0 1 0 1 Oil Temperature Engine #1 [OIT_1]
7 0 1 1 0 Pilot Altitude BARO Set (FINE) [ALT_CPT]
8 0 1 1 1 F/O Altitude BARO Set (FINE) [ALT_FO]
9 1 0 0 0 Selected Flight Path Angle Pilot [FPA_SEL_CPT]
10 1 0 0 1 0 0 0 0 0 0 0 0
11 1 0 1 Selected Heading [HEAD_SEL]
12 1 0 1 1 Selected Course #1 [SEL_CRS1]
13 1 1 0 0 Selected DH Pilot (COARSE) [DH_SEL_CPT]
14 1 1 0 1 Selected DH F/O (COARSE) [DH_SEL_CPT]
15 1 1 1 0 Pres. Pos Latitude (COARSE) [LATP]
16 1 1 1 1 N2 Engine #1 [N21]
Table C.1 Superframe #1

Superframe #2
Subframe 1, Word 57
Cycle 12 11 10 9 8 7 6 5 4 3 2 1
1 0 0 0 0 0 0 0 0 0 0 0 0
2 0 0 0 1 Center of Gravity [CG]
3 0 0 1 0 0 0 0 0 0 0 0 0
4 0 0 1 1 Oil Quantity Engine #2 [OIQ_2]
5 0 1 0 0 Oil Pressure Engine #2 [OIP_2]
6 0 1 0 1 Oil Temperature Engine #2 [OIT_2]
7 0 1 1 0 Pilot Altitude BARO Set (COARSE) [ALT_CPT] 0
8 0 1 1 1 F/O Altitude BARO Set (COARSE) [ALT_FO] 0
9 1 0 0 0 Selected Flight Path Angle F/O [FPA_SEL_FO]
10 1 0 0 1 0 0 0 0 0 0 0 0
11 1 0 1 0 0 0 0 0 0 0 0
12 1 0 1 1 Selected Course #2 [SEL_CRS2]
13 1 1 0 0 Selected DH Pilot (FINE) [DH_SEL_CPT]
14 1 1 0 1 Selected DH F/O (FINE) [DH_SEL_CPT]
15 1 1 1 0 Pres. Pos. Longitude (COARSE) [LONP]
16 1 1 1 1 N2 Engine #2 [N22]
Table C.2 Superframe #2

90
Superframe #3
Subframe 4, Word 41
Cycle 12 11 10 9 8 7 6 5 4 3 2 1
1 0 0 0 0 0 0 0 0 0 0 0 0
2 0 0 0 1 GMT/UTC (Hours) [UTC_HOUR] 0 0 0
3 0 0 1 0 Date (Day) [DAT_DAY] 0 0
4 0 0 1 1 Date (Month) [DAT_MONTH] 0 0 0
5 0 1 0 0 Flight Number (LSH) [FLTNUM]
6 0 1 0 1 Flight Number (MSH) [FLTNUM]
7 0 1 1 0 Flight Leg (Spare)
8 0 1 1 1 A/C type [AC_TYP] 0 0
9 1 0 0 0 A/C number [AC_IDT]
10 1 0 0 1 Fleet Identification [FLIDENT] 0 0 0 0
11 1 0 1 Gross Weight (LSH) [GW]
12 1 0 1 1 Gross Weight (MSH) [GW] 0
13 1 1 0 0 0 0 0 0 0 0 0 0
14 1 1 0 1 DFDAU Program Ident
15 1 1 1 0 FMC Database Update (Day) [NDBUPD] 0 0
16 1 1 1 1 FMC Database Update (Month) [NDBUPD] 0 0
Table C.3 Superframe #3

91
Appendix D. A310: Attitude Angles and Flight Control
Surfaces

Flight Path Vector and Attitude Angles

Figure D.1 Representation of the Flight Path angle, Pitch angle and Angle of Attack [11]

ATA 27: Flight Control Surfaces

PRIMARY CONTROLS SECONDARY CONTROLS
 Flaps: one double slotted inboard flap [6]
 2 Elevators [3], hinged to the Trimmable and on single slotted outboard flap [7]
Horizontal Stabilizer  Slats: three slat sections per wing [8]
 Trimmable Horizontal Stabilizer (THS) [4] associated with a Kruger flap [9] located on
 1 Aileron [1] per wing the leading edge of each wing
 5 outboard spoilers (Nos. 3 to 7) [2], on  Spoilers: there are 7 spoilers (Nos. 1 to 7)
each upper wing surface, as roll spoilers on each upper wing surface which are used
 Rudder [5] as speed brakes [10], roll spoilers [2] or
ground spoilers (all 7 surfaces)
Table D.1 List of the Flight Control surfaces on the Airbus A310

Figure D.2 Representation of the Primary and Secondary Flight Control surfaces on the A310 aircraft [11]

92
Appendix E. A310: Mean Aerodynamic Chord and Center of
Gravity Calculation

The position of the Center of Gravity (CG) is recorded in database version and expressed in terms of
percentage of MAC (Mean Aerodynamic Chord). The MAC is a reference line used in the design of
the wing and its position relative to the wing and the fuselage is accurately known [51]. Figure E.1
indicates the position and dimension of the MAC or AMC (Aerodynamic Mean Chord), which is
5.8287m long. The position of the point located at 25% of the MAC’s length is the reference when it
comes to measuring moments. The CG is calculated with the formula below:

( ) (E.1)

There the and correspond to the distance from station 0 - located

6.3825m forward of the fuselage nose [10] - to the CG and leading edge of the MAC, respectively.
Substituting with the values available the expression becomes:

( ) (E.2)

: 25%MAC

Figure E.1 A310: Airplane Dimensions [52]

93
Appendix F. Flight Phase Profiles

Airbus A310

Figure F.1 Flight Profile with Airbus Flight Phases [21]

SAGEM

Figure F.2 Flight Profile with the SAGEM Flight Phases [24]

94
Appendix G. A310-325: ECM Instrument Reading Example

95
Appendix H. Codes from the Algorithms developed

Code for the extraction of Stability Points

Code available under request to [email protected]

Code for the definition of the Engine Baseline Models and the plotting of
results for Trend Monitoring
Code available under request to [email protected]

96
Appendix I. Baseline Model Results: CS-PWB aircraft

Aircraft Stability Criteria Converged Stability Criteria

(76 Stability Points) (69 Stability Points)

(a) (b)
Figure I.1 CS-PWB Baseline Model Results: (a) Aircraft Stability Criteria (b) Converged Stability Criteria.

97
 Figure I.2 Distribution of Baseline Points by PKS Code

(a) (b)
Figure I.3 Distribution of Baseline Points by (a) Mach Number and (b) Altitude and Search Period

98