viewed here

viewed here
Project: ERG103.015
Document Identification No.: ERG103.015-DOC-3000-0001-Fokker-Issue 1
WP3 Report
“Evaluation”
ERG103.015 - AHMOS II
AHMOS
ACTIVE HEALTH MONITORING SYSTEM
INTENTIONALLY LEFT BLANK
Fokker Services B.V.
Title
: ERG103.015-DOC-3000-0001-Fokker-Evaluation
Controlled : AHMOS II TAMG
Copies
AHMOS II Industry partners
For Continued Competitive Operation
Lucas Bolsstraat 7, 2152 CZ Nieuw-Vennep
P.O. Box 231, 2150 AE Nieuw-Vennep
The Netherlands
Issue date
Issue no.
Report no.
Order no.
Security class
Company / dept.
:
:
:
:
:
:
19 Dec. 07
1
MSG-E-WP3
D27006
UNCLASSIFIED
FS/AMM
REPORT
Summary
In this report, an evaluation is given of the results of the work that has been performed in Work Packages 1 and
2. This evaluation is done against the following predefined requirements: functional, qualification and data
presentation. Furthermore, a benchmarking tool and a decision support tool has been developed and included
in the AHMOS II website. Finally, the next steps are identified and suggestions are given for improvements and
exploitation.
prepared/department:
checked/department:
I.P.A. van Veen/ S. A. Gungor FS/AMM
Seyit Gungor
original issue date:
approved/department:
19 Dec. 07
Seyit Gungor
All rights reserved. Reproduction or disclosure to third parties of this document or any part thereof,
or the use of any information contained therein for purpose other than provided for by this document,
is not permitted without prior and express written permission of Fokker Services B.V.
page 1 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
EXECUTIVE SUMMARY
To achieve the project goals, the work has been conducted in two Work Packages: WP 1, the technology path
and WP2, the flight test path.
In Work Package 1, the technology path, new sensors have been tested and or evaluated for their damage
detection capability and some preselected existing sensors have been further developed. In addition, to improve
data transfer capacity an ethernet network for AHMOS II has been developed and the microcontroller module
has been improved. Data analyses methods have been studied. Data presentation to the maintenance
personnel by using a Graphical User Interface via a notebook has been developed. In Work Package 1, to prove
the diversity in application of the damage detection systems, several tests have been performed. CASA and
ALENIA have performed flight test trials as part of Work Package 1.
In Work Package 2, ground tests and flight tests have been performed to demonstrate the feasibility of operating
Structural Health Monitoring Systems (SHMS) in a flight environment utilizing an actual military air vehicle. The
focus of the ground tests was on both damage detection (compare system output to actual damage evolution)
and system operation using qualified systems.
Flight tests have been done using a Hawk aircraft (BAE Systems) and a F-18 (Patria) aircraft. In the flight tests,
the Acoustic Emission, Strain Gauge, Fiber Optic, SWISS and Lamb Wave sensors have been tested together
with their complete network. The notebook with the GUI has been used on both the ground as the flight tests.
In order to learn the progress in Technology Level, the Technology Readiness Levels of all used systems have
been determined. The highest level achieved is TRL 7 for AE, crack in metal structure, for FBG, strain
measurement, for the Graphical User Interface, for Strain Gauge, crack in metal structure (if sensors are
correctly positioned), metal stringer disbond (if sensors are correctly positioned), system network and modules
and for the Central Computer.
The achievements of AHMOS II and the next steps
Acoustic Emission (AE)
The AE detection technology is real-time. It is designed to operate while the platform is in service undergoing
operational loads. Damage detection capability of the AE sensors has again been proven. The capability of
damage localization and extent is however not yet proven. Initially, further developments for the in service use
of this sensor should focus on “hot spot’ monitoring.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 2 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
SWISS
SWISS (Smart Wide area Imaging Sensor System) is based on ultrasound. In principle, it has the potential to
detect all such damages (e.g. corrosion, cracks, delaminations etc.) that can be detected by ultrasound in
conventional inspections. The self-test capability of sensor, sensor-coupling to metallic specimen after
instrumentation and under in-service like conditions has been demonstrated. Also, the known-good ultrasound
response of known-good (healthy) structure has shown sufficient repeatability to serve as ultimate self-test of
the system functioning. Damage has been detected when damage occurred. Localized damage (cracks) in the
structure has been detected, localized and quantified. Depending on the specific installation and
implementation, SWISS has demonstrated applications specific potential to deal with unexpected (non-lab-like)
behavior of hot-spot areas. In summary, SWISS has been used on the ground and the flight tests with
application specific success.
Strain Gauges
Strain Gauge technique can be used for damage detection in metallic aircraft structure. Strain Gauge sensors
are only suitable for monitoring known ‘hot-spots’. The Strain Gauge technique has been used on the ground
and the flight tests.
Fibre Optic Bragg Grating (FOBG)
Structural Strain Variation monitoring technique by means of FOBG is real time. It is designed to operate with
the platform in operation. FOBG sensors have the potential to detect delamination and debonding in composite
material, cracks in metal structure, to do strain measurements and to do fatigue life consumption calculations.
FOBG sensors have been tested on the ground tests and flight tests. On ground tests, the damages have been
found after post processing of the data.
Lamb Wave
The capabilities of the Ultrasonic Guided Wave (Lamb Wave) system are limited to monitoring of the structural
integrity of a region of structure local to the transducer. Therefore, this system is useful only to monitor ‘hotspots’ for debonding of joints, cracks and possibly corrosion. Test results show that the system has detected the
debond, in real time, during flight-testing, without any false calls.
Eddy current array sensor
The investigation showed that the high-frequency absolute sensor is suited for the detection of surface cracks
located just beneath the sensor, and the low-frequency reflection ring is suited for the detection of sub-surface
cracks in riveted joint structures. It is recommended to further investigate both sensor types using different and
realistic specimen configurations under fatigue testing.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 3 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
Long period gratings (LPG’s)
The investigation concludes that low cost, commercially available equipment based on HRI coated LPG’s can
provide advanced chemical sensing and liquid monitoring. This basic research has shown the potential of LPG’s
for SHMS purposes. However, extensive further research and development is needed to prove the usability for
SHMS purposes and to become sufficiently mature to be used on aircraft.
Corrosion sensor
This technique is based on the principle that the surrounding structure corrodes at the same rate as the sensor.
A corrosion sensor has been used on the Hawk flight test without any result due to not occurring of corrosion
damage. Extensive further research and development is needed for this sensor to become sufficiently mature to
be used on aircraft. One of the issues to tackle is in-service data validation.
Fiber optic sensor using reflectometric approach.
This work involved testing COTS components and telecommunications as fibres and optical devices to verify the
“reflectometer” concept for the detection of the presence of a liquid.
This is a preliminary study based on defining low cost and robust components, and the range of potential
measurements likely. The work also includes some practical investigations on the influence of the fibre optic cut
quality on the performance of an eventual sensor. This basic research has shown the potential of ‘reflectometric
approach’ for SHMS purposes.
Therefore, extensive further research and development is needed to prove the usability for SHMS purposes and
to become sufficiently mature to be used on aircraft.
System network and modules
All parts and modules of the system network that have been used and tested during the AHMOS II are
experimental or prototype parts. During the ground and flight-testing it has been proven that they fulfill the
functional and performance requirements as needed for the AHMOS II project but that they are not yet fit for
definitive in-service installation into aircraft. In order to achieve this, these parts should be further developed
(including miniaturization) and further tested on-ground and in-flight for qualification.
Data presentation
In this project, a Graphical User Interface (GUI) by means of a Notebook System has been developed by
(NIRAS) DEMEX in order to present damage data and more to the maintenance personnel. The GUI has been
used during a ground test and flight tests.
In three levels, information is given about the health status of system and structure to the user. During ground
testing and flight testing it has been proven that the level 1 and level 2 functionality of the GUI works as
expected. Level 3 information, as far as available, has not completely been used.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 4 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
Software qualification
In AHMOS II, because the SHM Systems are introduced for testing purposes only and the failure of the system
will not influence the operation of the aircraft, Software Criticality Level E (no effect on the system) has been
selected. However, at the commercial introduction of a SHMS in (military) aircraft the Software Criticality Level
of at least C should be considered.
Hardware qualification
During AHMOS II extensive amount of hardware qualification testing has been performed.
All the system parts that have been installed in the Hawk pod have successfully been tested on the ground and
in flight either by suppliers or by BAE Systems. All the system parts that have been installed in the F-18 pod
have also been ground and flight-tested. Except for an EMI problem, the system passed all tests.
From the AHMOS II project results the following next steps have been identified:
•
Proving 100% reliability of damage detection in the monitored area. This requires extensive testing and
validation.
•
Further development of damage localization capability.
•
Further development of damage extent determination capability.
•
Further development of software to provide Level 3 damage presentation data to the Graphical User
Interface. In first instance, for one known structural part with damages the 3D presentation should be
further developed, tested and validated.
•
Hardware and software Qualification of the actual system to be installed.
•
Prove 100% reliability of the GUI software.
•
Wireless technology.
•
Data reduction capability and enhancement of download speed of the Central Computer.
•
Standardisation of interfaces (both hardware and software).
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 5 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
Table of Contents
1
INTRODUCTION.......................................................................................................................................................... 8
2
TECHNOLOGY PATH (WP1)..................................................................................................................................... 9
2.1
SENSOR AND NETWORK TECHNOLOGIES (WE1.1) .................................................................................................... 9
2.1.1
New Sensors (WE1121-2)................................................................................................................................. 9
2.1.2
Existing Sensors (WE1121-3)......................................................................................................................... 12
2.1.3
System Network and Modules (WE1122) ....................................................................................................... 13
2.1.4
Data analysis and Data fusion (WE1123)...................................................................................................... 14
2.1.5
Data Presentation (WE1124) ......................................................................................................................... 15
2.2
GROUND TESTS TO PROVE DIVERSITY IN APPLICATION OF DAMAGE DETECTION SYSTEMS (WE1.2) ........................ 18
2.2.1
Wing Full Scale test (WE1210) ...................................................................................................................... 18
2.2.2
EF Slat fatigue test (WE1220, WE1290 and WE12A0) .................................................................................. 18
2.2.3
Marine Trial with AE (WE1230) .................................................................................................................... 19
2.2.4
Landing Gear attachment Upper fuselage (WE1240) .................................................................................... 20
2.2.5
Embedding trials for FBGS (WE1250) .......................................................................................................... 20
2.2.6
Trial on instrumentation practicability of SWISS (WE1260) ......................................................................... 20
2.3
FLIGHT TEST TRIALS CASA (WE1.3) ..................................................................................................................... 21
2.4
FLIGHT TEST TRIALS ALENIA (WE1.4).................................................................................................................... 21
3
FLIGHT TEST PATH (WP2)..................................................................................................................................... 22
3.1
PREPARATIONAL GROUND TESTS (WE2.4) ............................................................................................................. 22
3.1.1
Damage Detection.......................................................................................................................................... 22
3.1.2
System Operation ........................................................................................................................................... 26
3.1.3
Ground test results ......................................................................................................................................... 27
3.1.4
User comments ............................................................................................................................................... 36
3.2
HAWK FLIGHT TEST (WE2.5) .................................................................................................................................. 38
3.2.1
Acoustic Emission .......................................................................................................................................... 39
3.2.2
Strain Gauge .................................................................................................................................................. 40
3.2.3
Fibre Optics ................................................................................................................................................... 41
3.2.4
SWISS ............................................................................................................................................................. 41
3.2.5
Lamb wave ..................................................................................................................................................... 43
3.2.6
Graphical User Interface ............................................................................................................................... 44
3.2.7
Conclusion ..................................................................................................................................................... 44
3.3
F-18 FLIGHT TEST (WE2.6)..................................................................................................................................... 45
3.3.1
Strain Gauge .................................................................................................................................................. 46
3.3.2
SWISS ............................................................................................................................................................. 47
3.3.3
Graphical User Interface ............................................................................................................................... 48
3.3.4
Conclusion ..................................................................................................................................................... 48
3.4
QUALIFICATION ASPECTS (WE2.1)......................................................................................................................... 49
3.4.1
Software qualification .................................................................................................................................... 49
3.4.2
Hardware qualification .................................................................................................................................. 50
4
TECHNOLOGICAL READINESS LEVEL OF SYSTEMS ................................................................................... 52
4.1
4.2
4.3
TRL RATING........................................................................................................................................................... 52
TRL EVALUATION .................................................................................................................................................. 53
CONCLUSIONS ......................................................................................................................................................... 60
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 6 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
5
BENCHMARKING TOOL AND DECISION SUPPORT TOOL........................................................................... 61
6
IDENTIFICATION OF NEXT STEPS FOR EXPLOITATION AND IMPROVEMENTS................................. 62
6.1
INTRODUCTION ....................................................................................................................................................... 62
6.2
SENSOR SYSTEMS ................................................................................................................................................... 62
6.2.1
Acoustic Emission (AE) .................................................................................................................................. 62
6.2.2
SWISS ............................................................................................................................................................. 63
6.2.3
Strain Gauges................................................................................................................................................. 64
6.2.4
Fibre Optic Bragg Grating (FOBG) .............................................................................................................. 64
6.2.5
Lamb Wave..................................................................................................................................................... 65
6.2.6
Eddy current array sensor.............................................................................................................................. 65
6.2.7
Long period gratings (LPG’s)........................................................................................................................ 66
6.2.8
Corrosion sensor............................................................................................................................................ 66
6.2.9
Fiber optic sensor using reflectometric approach ......................................................................................... 67
6.3
SYSTEM NETWORK AND MODULES .......................................................................................................................... 67
6.4
DATA PRESENTATION .............................................................................................................................................. 68
6.5
QUALIFICATION ...................................................................................................................................................... 68
6.5.1
Software qualification .................................................................................................................................... 68
6.5.2
Hardware qualification .................................................................................................................................. 69
7
REFERENCES............................................................................................................................................................. 70
8
ABBREVIATIONS AND ACRONYMS .................................................................................................................... 72
9
APPENDICES .............................................................................................................................................................. 73
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 7 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
1
INTRODUCTION
The purpose of AHMOS II is:
•
To demonstrate the feasibility of an operating Structural Health Monitoring Systems (SHMS) in a flight
environment utilizing an actual military air vehicle.
•
Further development of:
o
damage detection capabilities
o
sensors
o
systems
o
presentation of damage data
o
qualification aspects
To achieve the AHMOS II goals the work is divided in two Work Packages: Work Package 1 “Technology Path”
and Work Package 2 “Flight Test Path”. In Work Package 3, the results of these work packages are evaluated
and the steps for exploitation and improvements are identified.
Evaluation is performed following the AHMOS II project structure, except the WE activities that were
preparational work for either the technology path or the flight test path. Additionally, Technology Readiness
Levels have been determined for the systems as developed and tested in this project.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 8 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
2
TECHNOLOGY PATH (WP1)
2.1
Sensor and Network technologies (WE1.1)
2.1.1
New Sensors (WE1121-2)
The work consists of separate investigations highlighting relevant damage detection approaches not included in
the prototype AHMOS SHMS. There were four major activities in this work element (Ref. 3):
-
Long period gratings
-
Reflectometric approach for chemical sensing
-
Develop corrosion sensor
-
Develop eddy current array sensor
Long period gratings (LPG’s) (WE1121-21, Ref. 4)
This work involved numerical and experimental investigations on the effects of High Refractive Index (HRI)
coatings of different thicknesses on LPG devices. The objective was to develop species-specific opto-chemical
sensors based on nanoscale HRI polymeric sensitive overlays. The coating material used was thin film
syndiotactic polystyrene (sPS) in 150 nm and 300 nm thick layers. This material exhibits a nanoporous structure
able to reversibly absorb certain analytes whose size and shape fit nanocavities in the sPS to establish
host/guest interactions, the main effect being a strong increase in density and thus refractive index. Theoretical
and experimental analysis showed good agreement, and suggested that sensitivity can be correlated with
cladding thickness. The investigation concludes that low cost, commercially available equipment (including a
spectrometer with a wavelength resolution of 0.1 nm) based on HRI coated LPG’s, can provide advanced
chemical sensing for VOC’s, gases, and liquid monitoring in light of the easy multiplexing capability and lower
complexity compared to alternatives such as Plasmon resonance or waveguide sensors.
This basic research has shown the potential of LPG’s for SHMS purposes. However, extensive further research
and development is needed to become sufficiently mature to be used.
Reflectometric approach (WE1121-21, Ref. 4)
This work involved testing COTS components and telecommunications as fibres and optical devices to verify the
“reflectometer” concept for the detection of the presence of a liquid. A “reflectometer” is a fibre optic sensor
based on measuring the reflective index difference between a cut glass fibre and the surrounding environment.
The reflection coefficient inside the glass dramatically decreases when the open fibre end is contaminated by
liquid. This is a preliminary study based on defining cheap and robust components, and the range of potential
measurements likely. The work also includes some practical investigations on the influence of the fibre optic cut
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 9 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
quality on the performance of an eventual sensor. Plastic deformation in the core zone may induce case-bycase variability in the reflected light intensity, leading to a lack of absolute measures for the sensor output.
This basic research has shown the potential of the ‘reflectometric approach’ for SHMS purposes. However,
extensive further research and development is needed to become sufficiently mature to be used.
Corrosion sensor (WE1121-22, Ref. 5)
The AHMOS II project includes a corrosion sensor development task as part of the research oriented Work
Package 1. BAE Systems was tasked with bringing appropriate technology to the project. Due to parallel
developments within BAE Systems, a relatively mature corrosion sensor (or more accurately paint degradation
sensor) could be used during this project. The UK programme within AHMOS to flight test a number of structural
health monitoring technologies on board a BAE Systems Hawk jet has presented an opportunity to deploy an
example of these sensors to raise their Technology Readiness Level to the status of ‘demonstrated in a
representative environment’.
The corrosion sensor was mounted in the AHMOS flight instrument package (ref. paragraph 3.2 Hawk flight
test). It was bonded to the outer surface of the BAE Systems Acoustic Emission instrument box. Leads attached
the sensor to a battery powered data logger located inside the instrument box. The data logger is a ‘stand alone’
unit and completely self-contained. It is not part of the integrated damage detection architecture of the Hawk
flight test system.
Figure: Family of corrosion and environment sensor designs. The ‘Sentinel’ corrosion sensor as used in
the AHMOS flight test pod is shown top right with windows of varying size in the paint layer. Control
samples are also shown in the bottom row
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 10 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
The technique is based on the assumption that the surrounding structure corrodes at the same rate as the
sensor.
The corrosion sensor has been used on the Hawk flight test. Because there was no occurrence of corrosion
damage, the detection capability could not be verified. Advanced lab trials at BAE Systems, however, showed
the corrosion detection capability of this sensor. On other military aircraft outside AHMOS, trials are performed
with prototype systems.
Eddy current array sensor (WE1121-23, Ref. 6)
Eddy current array sensors have been investigated by NLR for their potential applicability of local monitoring of
hot spots in metal aircraft structures. In reference 6 the detailed results are given.
Flexible sensor for high-frequency surface crack detection
Eddy current inspection (EC) is currently the primary technique for in-service inspection of metallic aircraft
components. The technique is based on the response of induced currents that are caused in a conductive
material when that material is subjected to an alternating electromagnetic field. Material defects that interrupt
the EC flow can be detected by a change in the EC response. An advantage of the technique is that no special
coupling medium is required between the inspection probe and test material.
Two prototypes of EC sensors were selected for the present investigation: a thin and flexible high-frequency
absolute sensor for the detection of surface cracks, and a low-frequency reflection ring sensor for the detection
of sub-surface cracks in fastener holes in multi-layer aluminum structure. Both sensor types work with a
TM
MultiScan MS 5800-E
of Olympus NDT. This is an EC multi-channel acquisition unit with 4 independent probe
inputs and up to 64 multiplexed inputs (16 time slots). The sensors were evaluated using a range of test
specimens with artificial defects and real fatigue cracks, including specimens under real fatigue loading.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 11 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
The investigation showed that the high-frequency absolute sensor is suited for the detection of surface cracks
located just beneath the sensor. This sensor is flexible and can be bent down to small radii; the sensor
produces reliable EC signals for radii down to 7 mm. The low-frequency reflection ring sensor can be used for
subsurface crack detection in riveted joint structures. The detectable crack length for this sensor depends on the
inspection configuration, the depth of the defects and the test frequency. As a rough indication, defects from
about 8 mm at a depth up to 3 mm in lap joints of Glare 4B-0.4 material are detectable. Smaller defects (length
< 6 mm) and defects at larger depth (depth > 4 mm) are not detectable anymore.
Both prototype sensors are sensitive to changes in ambient temperature. This should be considered (and
compensated for) when performing SHM measurements at different temperature, for example during in-flight
and on-ground conditions.
The feasibility and potential applicability of both EC sensor types for the local detection of cracks in metal
aircraft structures has been demonstrated. However, much further work is needed. For example, it is
recommended to increase the monitoring capabilities of the MultiScan equipment in order to use the system for
practical SHM applications. The system should allow continuous operation for periods much longer than one
day (main areas of attention are the acquisition rate and the cursor length in the strip chart display of the
instrument). It is also recommended to evaluate the multi-sensor capability of the MultiScan equipment using a
specified 16-channel switch box (in the present investigation only single sensor measurements were
performed).
It is recommended to further investigate both the high-frequency absolute sensor and the low-frequency
reflection ring sensor using different and realistic specimen configurations under fatigue testing. The actual
limits of flexibility of the high-frequency absolute sensor should be investigated using specimens with smaller
radii (< 7 mm). The low-frequency ring sensor should be further investigated for the detection of sub-surface
corrosion in aluminum alloy structures.
2.1.2
Existing Sensors (WE1121-3)
This work element consists of separate investigations regarding the capability enhancement of existing sensor
systems over the course of the AHMOS project. The report gives an overview of the results of the following two
activities:
- Improve functional performance of SGI
- Miniaturize electro-optical interface
Improvement of functional performance of SGI (WE1121-32, Ref. 7)
The strain gauge interface developed in AHMOS I was considered to perform relatively well in structural
damage detection (TRL7) and quantification (TRL5) applications. However, the prototype hardware used in the
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 12 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
first AHMOS project was too fragile to be implemented in the harsh (realistic) environment demanded by WP2
(“Flight test path”) of the AHMOS II project. For this reason, there were key objectives within WP1 (“Technology
path”) of AHMOS II for new hardware and software for the Strain Gauge Sensor System (SGSS), using the
AHMOS I prototype hardware as the base for this development.
The two major thrusts of development effort were to implement a new communication standard, and to assess
and respond to the environmental requirements specified for operation as in-flight equipment. In general, the
first of these (i.e. the new communication standard) has been a great success resulting in an increase in
communication speed and easier interfacing to other equipment. However, the challenge of addressing the
environmental requirements has been less straightforward and has resulted in more substantial design
alterations.
The SGSS has now, however, seen the full envelope of F-18 during total of ~11 test-hours and has been
operating faultlessly. This means that SGSS is fit for installation in a flight environment.
Miniaturization of electro-optical interface (WE1121-33, Ref. 8)
The space enclosure requirements defined in WP2 (“Flight test path”) of the AHMOS II project required a
substantial reduction in size of the FBG interrogator used compared to the AHMOS I program. The focus of the
WP1 (“Technology path”) program in this case is to address the main issues that prevent the technology being
commercially applied under operating conditions (size, ruggedisation, operating speed, communications, price),
rather than any issues remaining concerning the sensing concept as such.
The key driver in the development was to reduce the volume of the final design by optimizing existing
components or by replacing with newly available technology. The miniaturization led to a decrease in volume of
the interrogator from 23.2 liters (AHMOS I) to 2.4 liters. This is an acceptable size for use on test aircraft.
2.1.3
System Network and Modules (WE1122)
Two sub-sub-tasks are defined in this work element (Ref. 9):
- Review and improve system network.
- Review and improve system modules.
Review and improve system network (WE1122-1)
In AHMOS I, a CAN-network was used to transfer data between measuring modules and the central laptop. The
data transfer capacity of the CAN-network, however, is limited, due to the relatively slow transfer speed and
short message size.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 13 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
A recent trend in data communication is the use of Ethernet-network with Internet related protocols. This
technology was also adopted by the AHMOS-group, and new communication technology has been selected to
be the backbone of the AHMOS II project. This technology is the Ethernet-network with UDP/IP-protocol on top
of it. It has successfully been used in WIWEB and Risø test benches. It has also been tested over the Internet
and it has been implemented into the Hawk-pod. No major problems have occurred.
Review and improve system modules (WE1122-2).
In AHMOS I, the microcontroller modules and the whole system were designed to run in the laboratory only and
no special attention was paid to flight environmental requirements. The environmental and EMC-requirements of
the F-18 and Hawk pod made the hardware design of all subsystems challenging. Especially the vibration
caused problems when a multiboard design is used. Damages and malfunctions caused by vibration were
subcomponent level, not electrical component level ones: board-to-board connector, power module to printed
circuit board (PCB), etc. The individual electrical components (integrated circuits, resistors, transformers, etc.)
are mainly so lightweight that their mount to the PCB is lasting. For that manner, the vibration level is anyway far
from the level causing internal damages to electrical components. EMC was also an issue in both pods: the
level of conducted emissions exceeded the standard (MIL-STC-461E/CE102) in both pods. However, in both
cases the emission level was reduced to acceptable level by adding a filter to the power feed. Concluding, the
system modules have sufficiently been improved for flight-testing.
2.1.4
Data analysis and Data fusion (WE1123)
This work element gives an overview of different data analysis methods that can be used for SHM techniques
(Ref. 2). Data analysis - or signal processing - methods are essential to any generic SHM system to extract
features from different types of sensors and to translate this information into a diagnosis of the presence (yes/no
detection) and characterisation (type, location, dimensions and severity) of damage. The selection of a signal
processing technique is dependent on the specific SHM application and the specific data type.
Basic steps in signal processing for a SHM system include:
•
Data pre-processing.
- Smoothing / noise reduction (filtering, fitting, averaging).
- Data compression / dimensionality reduction.
•
Data analysis.
- Time-domain (statistical parameter analysis, time series models).
- Frequency domain (Fourier transform, Hilbert transform).
- Time-frequency domain (short time Fourier transform, Cohen distribution, Hilbert-Huang transform).
- Time-scale domain (wavelet analysis).
•
Data post-processing.
- Pattern recognition (neural networks, feature extraction/selection).
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 14 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
- Data fusion (multi-sensor architecture).
This table gives an overview of the applicability of data analysis methods for the different SHM techniques:
Data analysis method
Time domain analysis
Frequency domain
analysis
Time-frequency analysis
Time-scale analysis
Learning techniques
AE
++
0
+
+
+
+
FBG
+
0
++
+
+
++
SHM technique
LW
SWISS
+
++
0
++
++
+
++
++
++
++
+
++
Statistical parameter
Time-series
Statistical parameter
Fourier transform
Hilbert transform
Short-time Fourier
transform
Cohen distribution
+
+
+
Hilbert-Huang
+
+
++
Wavelet transform
++
+
++
Neural network
+
+
0
Data Fusion
+
+
+
0 - less applicable, + - applicable, ++ - well applicable
SG
+
0
++
+
+
+
++
0
++
0
++
+
+
++
+
+
The individual data analysis methods of different sensing techniques needed improvement and have been
improved significantly. Until now, it was not possible to work on data fusion. However, work on data fusion
should certainly be done in a future project to make optimum use of the available data to enhance the reliability
of inspection (increasing the POD and reducing the number of false calls).
2.1.5
Data Presentation (WE1124)
In this project a Graphical User Interface (GUI) by means of a Notebook System has been developed by
(NIRAS) DEMEX in order to present damage data and more to the maintenance personnel. The GUI has been
used during the ground tests and the flight tests of F-18 and Hawk aircraft.
In three levels information is given about the health status of the system and the structure to the user (Ref. 10):
level 1 gives an indication about the status of the system (by means of lights), level 2 gives an indication about
the health status of the structure (whether or not a damage or damage growth has been detected) and level 3
gives a graphical presentation of the damage (type, location and dimension). In level 3, also the detailed results
of each individual sensing technique, graphs, images etc. can be obtained.
Fokker Services has investigated the possibility of using original design/production (3D) drawings for the
presentation of the damage on the monitored structural part. This with the aim to give an as real as possible
picture to the maintenance personnel.
Investigation showed that, due to the great amount of 3D CAD (CAM) file formats that aircraft manufacturers
use for their aircraft structural drawings and presentations, it is impossible to write a common interface in
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 15 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
Labview to cope with all OEM structural drawings. Therefore, a solution would be to make an inventory of
converters with one file format as output, for which only one interface with Labview should be created. Most
programming in Labview is done in C/C++, so therefore the “to c” converters may be of most interest for
creating a standard interface in Labview and use converters for the different industry standard 3D model
formats.
Risø Ground Test (Ref. 22)
The notebook program could not be installed on a Risø laptop; therefore, the system was accessed via a
DEMEX supplied computer on which the GUI program was pre-installed. This system was used only to connect
to the network and download data for analysis elsewhere (directly by the sensor suppliers and by the notebook
program integrated evaluation tool). Occasionally, it was not possible to make a notebook connection to the
network, and this necessitated a direct FTP from the Central Computer data directory. None of the higher
function data evaluation tools of the GUI was used during the ground test.
Because during the ground test not all systems were running simultaneously it was not possible to correlate the
readings of the different sub-systems with the lab controlled loading. The change from testing all sub-systems in
parallel (one common test set-up) to testing the sub-systems in series (separately) meant that the lab personnel
had to cope with at least four individual set-ups during the workshop. To handle this change, the notebook
system configuration files had to be drawn from scratch by DEMEX.
During the ground test, the notebook system was connected to the Central Computer (CC) for a considerable
amount of time during data download after each test sequence. It became clear that for a full operational system
there is a need for a quicker download speed. This is a requirement, which pinpoints that the amount of data
gathered by the different measurement sub-systems should be reduced to a minimum. Furthermore, the
interaction between the CC and the notebook system should be optimized. A possibility could be to use an
interchangeable data storage medium onboard, in which case the time consuming data transfer procedure
would be minimized/eliminated. Furthermore, it became clear that there is a need for storing (and presenting)
only data sequences where sensor malfunctions or damage registrations occur.
Hawk Flight Test (Ref. 23)
According to the BAE partners there has been no problems using the program. The overall feedback is of a
quick user-friendly program, which has shown no signs of breakdown during the test. This is in respect of the
fact that the program was mainly used for quick and safe data transfers from the onboard equipment to the
laptop. From earlier tests, the program as a whole has been useful, accompanied with suggestions to
improvements.
F18 Flight Test (Ref. 24)
The DEMEX GUI software has operated smoothly during the tests. However, there was no need for making
many inspections and relatively few people have used the system.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 16 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
Not all the indications of the software were in harmony with the measured strains. The reason for this turned out
to be different interpretations of the definitions. This minor non-uniformity can be easily corrected from the
Demex or AHMOS-HW-SGSS unit software.
Questionnaire
After the development of the GUI, Fokker Services and DEMEX evaluated the system with the end-users
(maintenance personnel of the Royal Netherlands Air force and Fokker Services). The end users were quite
pleased with the GUI, especially with the Level 3 information, if fully developed.
After the completion of the AHMOS II WP2 tests, DEMEX sent a questionnaire to the users of the GUI notebook
during the Risø ground tests, Hawk and F-18 flight tests. From the results of the questionnaire the following can
be concluded:
- The users were overall satisfied with the performance of the Notebook System.
- The notebook system has primarily been applied for data download and secondarily for an overall health
status check; i.e. applications that already from the design phase were meant for the “ordinary”
user/maintainer (Level 1 and 2).
- Level 3 facilities, information on damage location and extent have - except for the F-18 test - not been used.
- None of the end-users has used the notebook system configuration facility during the test period. The fact, that
the configuration task has been left with DEMEX whenever needed, indicates that it is too big a responsibility
to give to even super-users. Due to the potential huge consequences, regarding health status misinformation if
configuration is done improperly this feature should therefore be restricted as indicated in the figure below.
Restricted Area
SHMS
Health
Status
Structure
Configure
SHMS
Test
SHMS
Notebook
System
Maintainer
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 17 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
Conclusion
The GUI notebook has primarily been used to check the level 1 and level 2 status. Damage visualization at level
3 has not been used. The GUI notebook was used however to download the level 3 data from the central
computer. This download in its current form takes too long for practical use in an operational environment. The
use of an onboard interchangeable data storage medium has been proposed. Another solution would be to
transfer only relevant data sequences.
The end-users during the AHMOS II tests are not fully representative for the end-user that will operate the SHM
system in the future. During the AHMOS II tests, a verification of the level 3 data on a low level was necessary,
also for building confidence in the system.
The creation of the required setup files takes a high level of expertise and was done by the supplier of the GUI
notebook (DEMEX).
2.2
Ground tests to prove diversity in application of damage detection systems
(WE1.2)
The purpose of the ground tests is to extend the test experience for potential monitoring tasks, to broaden the
knowledge on feasibility of integrated damage detection techniques for a variety of platforms or components
under practical conditions (Ref. 11).
2.2.1
Wing Full Scale test (WE1210)
The test part has been successfully manufactured. At least 25 of the FBGs are operative after the co-bonding of
rd
th
the test part. Due to delays, the structural test is scheduled for the 3 /4 week of November 2007. The
executive summary of this report will be included in an appendix to this report.
2.2.2
EF Slat fatigue test (WE1220, WE1290 and WE12A0)
The scope of the activities in these work elements can be summarized as follows (Ref. 12, 13, 14 and 15):
- Definition of the FBGS locations to be monitored.
- Description of the test load introduction system.
- Definition of the fatigue spectrum.
- Checking of the structure integrity.
- Fatigue measurements.
- Recording and processing of the measurements.
- Conclusions.
The FBGS have been located in areas where high strain levels are expected. The tracks and actuator cutout
areas have been selected since a high stress level is expected in these areas due to the hole concentration
effects. A continuous data recording has been performed during the fatigue test development. This dynamic
data capture will be used to obtain the real fatigue sequences, cycle-by-cycle, at the FBGS locations.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 18 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
These fatigue sequences have been developed in terms of strains at the FBGS locations, but it can easily be
converted into stress sequences applying Hooke’s Law. These stress sequences could be used directly to
obtain the cumulative damage information at the FBGS locations.
The objective of fatigue tests is to monitor the behavior of the structure under the acting fatigue loads.
Therefore, it is necessary to do inspections during the fatigue test to check that the structure is free of cracks. It
is well known that in the case of the appearance of a crack in the structure, this crack causes changes in the
load paths in the area surrounding to this crack. Therefore, by checking if there are changes or not in the strain
measurements in the instrumentations installed close to the critical locations could be used of the indication of a
possible crack in this area. Both FBGS static data capture or/and standard strain gauge data could be used to
check this structural integrity.
The executive summary of this report will be included in an appendix to this report.
2.2.3
Marine Trial with AE (WE1230)
To be able to extract an AE response due to a change in structural state from the normal, legitimate structural
response it is essential to gain knowledge about the AE (background) profile under normal operational
conditions. The objective with the AE marine trial has therefore been to demonstrate that it is possible to
generate an AE profile that varies with and is significant to the vessels operating environment (Ref. 30 & 31).
The AE measurements were carried out with a portable, dual channel AE system, Pocket AE equipped with two
R15a AE sensors. The system records an extensive number of AE Hit parameters at every signal exceeding a
pre-selected AE Hit threshold value, including Peak Amplitude, Energy and Duration. The system was chosen
for these initial trials/investigations due to its all round ability to provide common AE parameters, but it would not
be suitable as a permanently mounted system. A final full operational and permanently mounted marine SHM
application should be tough, clever and cheap (simple).
During an eight-hour sail with a Danish Navy Standard Flex 300 vessel AE measurements were successfully
taken against the vessels’ composite hull while the marine vessel was operated at various speeds and
maneuvers.
It is concluded from the marine trial measurements that the AE profile varies with the operating environment and
furthermore with the measurement location (hull section), which indicates that it is possible to “map” the AE
(background) profile of a vessel. This observation suggests that is possible to identify structural damage through
an “exception analysis”, i.e. changes in the AE profile.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 19 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
Two generations of passive AE sensors have been developed and tested. The second generation AE sensor,
while being completely minimized in terms of cost, seems to give a very good representation of the acoustic
energy in a structure.
2.2.4
Landing Gear attachment Upper fuselage (WE1240)
C27 Platform vibration level simulation test applied on the FOBG instrumentation to verify the correct
functionality of FOBG instrumentation before applying in flight on C27 Platform (Ref. 11).
Validation tests results about the crossing inboard/outboard of FOBG sensors through the C27 platform upper
fuselage forward wing trap door. The critical point of this application is the ingress of the 4 external (outboard)
FOBG inside the aircraft (inboard) in order to connect these sensors to the optical instrumentation; the FOBG
crossing happens through the upper fuselage forward wing trap door.
The executive summary of this report will be included in an appendix to this report.
2.2.5
Embedding trials for FBGS (WE1250)
These trials deal with the manufacturing of carbon fiber reinforced epoxy specimens to assess the viability of
FBGS monitoring technology for composite materials structures (Ref. 16). The aim of such an assessment is
threefold:
1. To know how the insertion of FO sensors affects characteristic mechanical properties of host material.
2. To estimate the variation of FO performance through the evaluation of the characteristic ‘K’ coefficient, that
converts wavelength units to deformation (or other) units for embedded sensors submitted to different type of
efforts and for different material conditions in a temperature range that is representative for aeronautical
structures in service.
3. To check if the entire interface of the optical fiber from the laminate egress to the optical connector can resist
habitual fatigue and vibration envelopes in aeronautical environments.
All these three aims have been evaluated by the accomplishment of ‘test matrixes’ within the AHMOS II project.
It could be seen that in all the different types of specimens, the used embedding techniques gave good results
and the optical fibers are operative after the manufacturing of the host material. In the special case of the
ruggedized specimens it could be seen that the embedded optical sensors and the fiber optic interface with its
optical connector survived all the fatigue, vibration and acceleration conditions that simulates operative loads of
military fighter aircraft.
2.2.6
Trial on instrumentation practicability of SWISS (WE1260)
SWISS Imaging Ultrasound sensors have been applied permanently to a selection of two hidden damage
susceptible hot spot areas of a complex weapon system structure section (Ref. 17). Difficulties in the
practicability of the instrumentation have been identified and overcome and sufficient quality and reliability of the
installed sensors have been proven.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 20 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
Fatigue tests have been conducted with the result that no false alarms occurred, as was confirmed by later
conventional inspection.
Future improvements should address the ease of installation and the reliability of sensor bond for various
materials (such as composite, aluminum, titanium and steel). Furthermore, real applications would require
reliable means of over-seal, practical means of connecting cables with sensors and to ground-interface
connectors at low weight and reduced cost (miniaturization!).
2.3
Flight test trials CASA (WE1.3)
No results were received. Preparational work is reported in ref. 18 and 19.
The executive summary of this report will be included in an appendix to this report.
2.4
Flight test trials Alenia (WE1.4)
No results were received.
The executive summary of this report will be included in an appendix to this report.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 21 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
3
FLIGHT TEST PATH (WP2)
To demonstrate the technological maturity of a prototype integrated structural health monitoring system under
operational conditions two flight test programmes have been executed (Ref. 20). A limited damage simulation
was included in these flight tests to demonstrate the functional performance and the reliability of the SHMS
under fast jet flight conditions for those sensing techniques that have to operate during flight.
One SHMS was installed on a pod of a Hawk test bed aircraft of BAE Systems (WE 2.5 "Flight Demonstration
Hawk") and one on a pod of a F-18 aircraft of FiAF (WE 2.6 "Flight Demonstration F18"). The SHM systems
were operated similar to a realistic scenario. The Hawk test bed was intended to offer the possibility to deeply
investigate and control the flight test activities, but for a limited number of flights. The F-18 aircraft was planned
to be exposed to independent in-service operation, but for a significantly larger period of flights.
Because the limited damage simulation in the flight test programmes was not deemed suitable to demonstrate
practicability and applicability in terms of functional performance, a ground test (WE 2.4 “Ground Test
Demonstration”) with realistic structural tests was performed to prove the damage detection and monitoring
function for realistic structural forms on subcomponent level. Several simultaneously operated structural tests
then allowed to demonstrate the complete SHM for a set of realistic monitoring tasks.
3.1
Preparational Ground tests (WE2.4)
During the ground tests at Risø, several independently but simultaneously operated sub component structural
tests were performed during which the complete system with all sensing techniques was monitoring the damage
development (Ref. 21 and 22).
The focus of these tests was on both damage detection (compare system output to actual damage evolution)
and system operation.
3.1.1
Damage Detection
Sub-component and test definitions
Sub-component 1a (tensile test):
A large metal riveted panel with some structural detail
2
Description: <1.5mm thick, 1m Aluminium
Sensor type(s): AE and SGI
Damage form: Cracking at the rivets
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 22 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
Figure 2: Showing the fully instrumented sub-component 1a
Instrumentation of sub-component 1a consisted of seven SGI sensors and five AE sensors. Figure 2 shows the
overall distribution of sensors on the finished sub-component.
Test specifications for sub-component 1a
Mean load
Initial load level of 2kN
Ultimate load level of 22kN
Amplitude
+/- 0.5kN
Fatigue
1 Hz
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 23 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
Sub-component 1b (peel test):
A large metal panel with bonded stringers
2
Description: <1.5mm thick, 1m Aluminium
Sensor type(s): AE, FO, Lamb Wave and SGI
Damage form: stiffener debonding
Figure 3: Showing the fully instrumented sub-component 1b
Instrumentation of sub-component 1b consisted of a single SGI sensor, five AE sensors, two FO strain sensors,
and a lamb wave sensor. Figure 3 shows the overall distribution of sensors on the finished sub-component.
Test specifications for sub-component 1b
Mean load applied
300-600 N
Displacement Amplitude
0.15mm - 0.50mm
Frequency
1Hz
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 24 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
Sub-component 2 (four point bend test):
A CRFP panel stiffened by a bonded CRFP top-hat stiffener
2
Description: Quasi-isotropic CFRP 1m panel
Sensor type(s): Lamb Wave, AE, and FO
Damage form: Disbonded stringer
Figure 4: Showing the fully instrumented sub-component 2
Instrumentation of sub-component 2 consisted of five AE sensors, two FO strain sensors, and two lamb wave
sensors. Figure 4 shows the overall distribution of sensors on the finished sub-component.
Test specifications for sub-component 2
Mean load applied
200-540 N
Displacement Amplitude
0.25mm – 0.75mm
Frequency
1Hz
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 25 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
3.1.2
System Operation
Network task
Prior to deploying the system for the ground test, the main network task was to assemble the full SHMS and test
its’ function. This task was led by Emmecon and supported by NLR (Central computer) and DEMEX (Notebook),
with input from each of the sensor module suppliers. This task has successfully been accomplished.
Figure 5: Diagram showing the AHMOS SHMS network, the test components and the data output from the
Ground Test Demonstration
In the schematic diagram, the Central Computer Subsystem (CCSS) connected to each of the four on-line
measurement subsystems: Fibre Optic (FOSS), Strain gauge (SGSS), Acoustic Emission (AESS) and Lamb
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 26 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
Wave (LWSS). These subsystems are all connected via an Ethernet switch. The AHMOS Notebook can also be
connected to this network in order to download and interpret the sensor data collected by the Central Computer.
Each of the four on-line sub-systems receives data from sensors attached to one, two, or three subcomponents, as they are loaded in the test machines. The sensor data is analyzed and transferred to the
Central Computer for later retrieval by the Notebook.
In addition to the on-line systems, there is the SWISS sub-system. This is an off-line measurement system that
generates output once each test cycle is completed (while the aircraft is “on the ground”). The notebook
connects directly to the SWISS SS to receive this data output.
The data generated during the Ground Test consists of the complete sensor output received by the AHMOS
Notebook, the fatigue loading data from the test machines, and the visual inspections carried out on the three
sub-components.
In addition to the working SHMS, a status monitor was included with the Central computer in order to provide a
live indication of the status of the system throughout the ground test.
Notebook task
The objectives of the Notebook task in the Ground tests included supporting the network working group during
set-up of the SHMS, demonstrating the primary functionality of the notebook in connecting and downloading the
SHMS data gathered and stored in the Central Computer, and instructing Risø personnel to be able to operate
the notebook hardware during the ground testing.
A manual was prepared specifically for the ground tests defining the AHMOSII SHMS notebook system.
Detailed information about the Ground test Notebook system is contained in reference 28. This manual also
defines the terms used in classifying returned damage for each sensor sub-system.
3.1.3
Ground test results
Notebook system
The Ground test was the first time all sub-systems were connected and collecting real data. Because not all of
the systems were running simultaneously it was not possible to correlate the readings of the different subsystems with the lab controlled loading.
Data analysis / output summary from the ground test
The change from testing all sub-systems in parallel (one common test set-up) to testing the sub-systems in
series (separately) meant that the lab personnel had to cope with at least four individual set-ups during the
workshop. To handle this change, the notebook system configuration files had to be drawn from scratch by
DEMEX.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 27 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
Hardware/software issues highlighted as a result of the workshop and/or ground test
During the ground test, the notebook system had to be connected to the central computer for a considerable
amount of time for data downloading after each test sequence. For a full operational system, there is need for a
quicker download speed. This can be e.g. accomplished by reducing the amount of data gathered by the
different measurement sub-systems to a minimum. Furthermore, the interaction between the central computer
and the notebook system should be optimized. A possibility could be to use an interchangeable data storage
medium onboard, in which case the time consuming data transfer procedure can be minimized/eliminated.
Changes made because of the experiences during the workshop and/or ground test
Some optimization regarding the download speed of the system was programmed after the workshop.
Furthermore, it became clear that there is a need for storing (and presenting) only data sequences where
sensor malfunctions or damage registrations occur.
Development of the system readiness because of the workshop/ground test
At Risø, the AHMOS system reached a stage where it was for the first time possible to run the notebook system
on non-simulated data. The download part of the notebook system was proven to work and fulfilled its purpose
during the workshop. The other features of the notebook system were not addressed/tested, and therefore the
readiness of these parts was not advanced during the workshop.
Central Computer
Hardware/software issues highlighted as a result of the workshop and/or ground test
One by one, the sensor systems were integrated with the Central Computer. Apart from some minor network
issues, like cabling and duplicate IP addresses, the integration of the Central Computer with the strain gauge,
fiber optical, and lamb wave sensor systems was successful. During the workshop, it was not possible to
establish a network connection with the acoustic emission sensor. The problem was identified and verification of
the solution was postponed to a later time. Data exchange with the notebook system was also verified and
worked as expected. The Central Computer web interface proved a useful tool for online verification of level 1/2
sensor data.
Changes made because of the experiences during the workshop and/or ground test
No changes were required in the Central Computer hardware and software because of the experiences during
the workshop. Network and system parameters are stored in a configuration file and they can be easily adapted
to changes in network topology or sensor system behavior.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 28 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
Development of the system readiness because of the workshop/ground test
The integration of the Central Computer with the AHMOS II system components on network and data exchange
level was successful. Most of the network issues were solved during the workshop and the remaining matters
were closed during the follow-up meeting at QinetiQ.
After the workshop no major issues were left that could block successful integration of the AHMOS II system
components.
Strain Gauge
Data analysis / output summary from the ground test
The SGSS operated as planned and was successfully re-tuned by Risø.
Conclusions concerning the sub-component 1a (the tensile rivet test; 7 SG channels):
The anticipated failure mode (rivet hole cracking) was not achieved. Instead, the failure occurred around the
lower grip region. The seven SG channels were however instrumented to capture cracking from rivet holes.
Therefore, no results on damage detection capability.
Conclusions concerning the sub-component 1b (the peel test; 1 SG channel):
The anticipated failure mode (peeling) was achieved. The SG was working as anticipated to indicate there is
damage for the anticipated failure mode (disbond).
Hardware/software issues highlighted as a result of the workshop and/or ground test
SGSS hardware used at Risø was the same, which was later, used in F18 and Hawk- test benches, no issues
identified. In addition, the same software was later used in flying test benches, in Hawk-POD with minor
modifications to support specimen characteristics. SGSS hardware and software functioned satisfactorily.
Development of the system readiness because of the workshop/ground test
The system itself worked as planned and the minor network problems were solved later.
The damage detection capability of the Strain Gauge system is summarised in the table below:
Damage Properties
Number
Specimen
of sensors
Damage Detected
Extent
Location
During Test 1, 2
Composite
7
No4
No4
No3,4
1
Yes
No
Yes 3
Disbond
Bonded Metallic
Peel
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 29 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
Notes:
1. All methods of damage detection using strain gauges (Fibre Optic or Electrical) require prior knowledge
about what kind of damage is likely to occur and the expected location.
2. Post processing is included in the run time firmware of the Strain Gauge system.
3. The number of sensors and the spacing between them determines the accuracy of the location.
4. The anticipated failure mode (rivet hole cracking) was not achieved.
Lamb waves
General comments on the workshop and ground test and data analysis / output summary from the
ground test
The LW system operated flawlessly on setup and during the first few hours of operation. It then developed an
overheating problem. Several factors contributed to this problem. Firstly, the heat sinks for the transmitter
power amplifiers were enclosed within the unit and not connected to the walls of the box, causing heat to build
up inside. Secondly, the transmitter inhibit line which was designed to switch off the transmitter amplifiers
temporarily during the receive phase was not activated by the control software. Thus, the transmitters were
running continuously. As overheating did not occur when operating with the box lid removed, it was agreed that
ground testing would continue without the box lid in place and that the overheating problem would be rectified
after completion of the tests.
The composite Test Sample 2 was first tested in a four-point bend designed to cause disbonding of the stringer.
Throughout the test, the LW system correctly transmitted both Level-2 and Level-3 data over the Ethernet
network to the Central Computer (CC). Level-2 data took the form of a transmission parameter, an integer in
the range 1 – 100 indicating the normalised transmission amplitude across the bonded joint and indirectly the
extent of disbond, whilst the Level-3 data comprised the received waveforms, which were transmitted at a rate
of approximately three per second. Initially, the normalised transmission amplitude was stable at 100 % +/- 3 %
until the disbond entered the region monitored by the LW system when it began to fall, as the disbond
progressed.
When the normalised transmission amplitude fell below the 85% threshold, damage was reported to the CC.
The 85% damage threshold was chosen to represent an ‘Amber’ damage warning where the majority of the
bond was intact but damage had been detected.
When the normalised transmission amplitude eventually fell below the 50% threshold, critical damage was
reported to the CC. The 50% damage threshold was chosen to represent a ‘Red’ damage level where the
disbond had reached a critical level. When reaching the Red, or critical, damage level, the Level-2 damage was
reported as 255, as specified in the AHMOS II NETWORK AND COMMUNICATION SPECIFICATION.
The metallic specimen Test Sample 1b was subjected to a peel test in a tensile test machine. At intervals during
the test, the point of tensile load application was adjusted by moving it to different positions along the stringer.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 30 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
The test was carried out over a period of eight days, during which the LW system correctly transmitted both
Level-2 and Level-3 data over the Ethernet network to the CC. Finally, the disbond extended to the region of
stringer monitored by the LW system when, once again, the system immediately indicated the damage through
the falling transmission amplitude. Waveforms from the test were also successfully stored over the network.
Once again, when the normalized transmission amplitude fell below the 85% threshold, damage was reported to
the CC. The 85% damage threshold was chosen to represent an Amber damage warning where the majority of
the bond was intact but damage had been detected. When Amber damage had been detected, the damage field
of the SAMH header was set. Level-2 damage continued to be transmitted whilst in the Amber damage state.
When the normalized transmission amplitude eventually fell below the 50% threshold, critical damage was
reported to the CC.
Thus, at the conclusion of the Ground Test, the LW sub-system had demonstrated disbond detection on both
CFC and aluminum-alloy specimens.
Hardware/software issues highlighted as a result of the workshop and/or ground test
The following hardware and software issues were encountered in the course of the workshop:
The analogue board of the LWS experienced overheating problems, which were temporarily mitigated
by operating the LWS with the lid removed from the LWS enclosure.
Minor inconsistencies with reporting the state of the LWS were observed and rectified.
Movement of the LWS into the test hall caused a patch lead to become unsoldered from the analogue
circuit board. This obviously has implications for any vibration testing.
Significant problems were encountered with transferring FPGA programs to the flash RAM of the logic
board. This resulted in substantial delays, and raises questions about the reliability of the current logic
board. It was possible that the overheating of the LWS contributed to this.
Significant delays were encountered due to the need to modify the FPGA software for each new time
delay.
Only a limited range of transmit and receive gains are currently available, and no filtering was available.
This was the result of deviations from the original design and the selection of incorrect components. On
the receiver amplifier this problem was mitigated by fitting a variable resistor, however only one
transmitter gain was available, which seriously compromised the signal-to-noise ratio. It was not
possible to mitigate the problems with the filtering.
Changes made because of the experiences during the workshop and/or ground test and development of
the system readiness because of the workshop/ground test
Since the Ground Test, before flight-testing, all of the outstanding problems highlighted above were addressed
and largely resolved. Improvements to the digital control software improved the downloading of FPGA programs
to the flash RAM, and changes to the analogue gain control were implemented. To overcome the problem of
transceiver overheating, the two output power amplifiers were moved to a position mounted on the transceiver
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 31 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
casing where heat could be dissipated in the case wall. More importantly, the control software was modified to
utilize the power amplifier inhibit line, radically reducing their duty cycle and consequently their heat output.
The reason for failure of the first sensor installed on the carbon composite specimen was investigated. One
previous similar example of sensor failure had been encountered during testing. This sensor was first examined
and electrical continuity tests carried out, but no fault was identified. The sensor was then disassembled and the
electromechanical coupling coefficient of the piezoelectric layer was measured and found to be normal.
Ultimately, no clear explanation for the failure could be found, though it was conjectured that the sensor bonding
could have been the cause.
These modifications were essential to the successful performance of the system throughout the subsequent
flight-testing.
The damage detection capability of the Lamb Wave system is summarised in the table below:
Damage Properties
Number
Specimen
of sensors
Composite
Damage Detected?
During Test
Post processing
Extent
Location
2
Yes
Yes
Yes 1
Yes 2
2
Yes
Yes
Yes 1
Yes 2
Disbond
Bonded Metallic
Peel
Notes:
1. The indicated damage extent was verified by visual and NDT measurements.
2. The system was configured to monitor a specified bond and the location of disbonding was known a-priori.
Acoustic Emission
The BAE Systems acoustic emission system was used in the ground tests at Risø and the Hawk pod flight test.
The flight tests in the Hawk under wing pod provided a realistic environment for the system but imposed
limitations on the test structure and duration. The ground test at Risø provided more realistic test structures but
took place in a relatively benign environment.
The ground test was performed approximately half way through the AHMOS II program. Thus elements of the
acoustic emission (AE) system, both hardware and software were still under development. For instance,
Emmecon’s signal processing board was developed in two phases. This allowed other system components to
be developed without limiting its final functionality. The second phase was delivered later on in the program. It
was not available for the ground tests at Risø. Thus, these tests were performed with the first phase device.
Despite the use of interim rather than final versions of equipment and software, the system used at Risø was a
good functional representation of the final devices destined for flight trial. The results from the Risø trials
increased confidence in the overall system approach adopted for the AE system and decreased risk for the
subsequent flight trials.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 32 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
Description of tests and main results
This section describes data captured during the AHMOS ground tests at Risø in Denmark. The acoustic
emission system recorded data from two test structures as damage was grown in them. The first structure
tested was a carbon fiber skin co-bonded to a top hat stiffener. The stiffener was cut width-ways across its
centre and the panel was loaded to encourage the bond to fail at the point where the stiffener was cut.
The second specimen was a riveted metallic structure put under load to induce the growth of metallic cracks.
Each specimen carried multiple sensors (e.g. strain gauge, fiber optic, Lamb wave etc.) from other AHMOS
partners
The response of the AE system was checked before testing began by making a pattern of pencil lead breaks on
both structures. The lead breaks (Hsu-Nielson test) simulate acoustic emission. This process is to ensure that
the sensors were functioning correctly. It also allowed the speed of sound to be determined by recording
differences in time of arrival of the lead break signals to reach each of the sensors. This timing information was
used as calibration data for the AE event location algorithm.
Specimen testing took place over a two-week period. As only one AE system was available, the coupons were
tested in sequence.
Examination of the AE waveform data gathered during the ground test revealed periodic interference picked up
by the system. The source of this signal was not known at the time of the ground test. While this low level
interference did seriously impede the acoustic emission system its elimination would improve the system’s
signal / noise performance.
The interference signal was eventually traced to the switched mode power supply. A minor modification to the
circuit removed the interference.
Removing the interference signal also allowed the sensor signal gain to be increased enhancing the overall
sensitivity.
Further shortcomings were observed with the AE system Built-In-Test facility. The acoustic emission system
was designed with a self- testing facility, which uses the ability of the sensor to also be used as sound
generators. One sensor is used to generate a simulated acoustic emission, which can then be detected by the
other sensors. Thus, the system can determine if a sensor is functioning correctly. The self-test facility failed to
function correctly during the ground test and as result; design modifications were made which proved successful
on subsequent trials (including the Hawk flight trials)
The Risø ground tests served as a dress rehearsal for later flight tests of the AE system. The tests increased
confidence of the overall system capability and highlighted some issues that could then be corrected. As such,
the whole exercise reduced the risk for the final flight test programme.
The tests at Risø demonstrated particularly well the AE system’s ability to detect damage in a failing composite
bonded joint. The key features of these results are:
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 33 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
•
Detection of multi-site damage originating from damage growth either side of the cut stiffener.
•
By tracking, the clustering of events concentrated at the active crack faces, the progression and hence
extent of disbond could be measured.
These composite disbond results are a valuable supplement to the overall evaluation of the AHMOS AE system
and complement the evaluation results obtained during the flight trials which focused on a different form of
damage detection: i.e. metallic crack detection.
Results from the riveted metallic specimen were not so conclusive mainly because the specimen failed in an
unexpected fashion. Instead of damage occurring along the stiffener, the specimen failed at the load input point
near one end. The positioning of sensors on the specimen meant that this region was not well monitored.
Despite this, subsequent (post trial) analysis of the data yielded a location map with a cluster of events close to
the crack albeit with a low statistical significance. This demonstrates some ability of the acoustic emission
system to find damage in unexpected locations.
The issues discovered with interference and the built-in-test facility allowed improvements to the system to be
made, which would not necessarily have occurred without the Risø trials.
The damage detection capability of the Acoustic Emission system is summarised in the table below:
Damage Properties
Number
Specimen
of sensors
Composite
5
Damage Detected?
During Test
Post processing
Extent
Location
Not attempted
Yes
Yes (by
Yes
Disbond
tracking
point of
disbond)
Bonded Metallic
5
Not attempted
Yes
No
Yes
Peel
Fibre optic
Data analysis / output summary from the ground test
Strain measurements were made on the bonded composite and bonded metallic subcomponents. There was
no hardware, network of other interruptions to the data. Changes in the strain signals correlated well with
damage events noticed during the test period.
The interrogator was configured to output data at the maximum rate possible, i.e., 8 sensors on each of 4
channels. Where less than 8 sensors were present on a channel, the rest of the spaces was to be filled with
zeros.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 34 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
Hardware/software issues highlighted as a result of the workshop and/or ground test
Optical parts of 4-channel instrument were assembled using COTS 1 x 2 splitters cascaded into a 1 x 4
arrangement and fitted in a standard splice cassette. Assembly was fiddly and the bare fibers exiting the
cassette were vulnerable.
Periodic variation in the noise of data files, also reproduced in ‘null’ data (where the interrogator expects a
sensor but none is connected). Spikes repeat every 64 points.
The noise levels varied from sensor to sensor. Some were near to our ideal level of 2 micro strain but many
were much higher, around 20 micro strain.
The samples were instrumented with single FBG, which were then spliced together. The excess fiber was
coiled and taped in position on the sample. While this was an easy, flexible method, it was, as expected a little
untidy in that there was quite a lot of excess fiber to tidy up, approx. 500 mm for each pair of sensors.
Changes made because of the experiences during the workshop and/or ground test
Optics for flight instrument to be made from components with 900 mm protected fibers wired in place on a flat
glass-phenolic PCB-type board for ease of assembly and handling. Optics also to be potted with silicone resin.
Future 4-channel instruments to be made with a custom 1 x 4 splitter.
Periodic spikes traced to an overflow error and corrected.
Output was limited to 2 sensors on one channel to reduce file sizes for the flight tests.
Some improvements were made to the signal processing which led to a reduction in the noise level for the flight
tests. (Subsequent work has demonstrated that we can reliably achieve 5 micro strain resolutions and 2 micro
strains are possible with further improvements).
No splices to be used in the flight coupon sensors. It was decided to use an array of 2 FBG rather than
separate ones and to terminate this array directly into a connector.
Based on the experience gained processing the ground test data, LabView programs were written to process,
check and display the flight test data, making the analysis much more efficient.
Development of the system readiness because of the workshop/ground test
We were greatly reassured by the successful integration of a working fiber-optic system into the AHMOS II
system. Useful data was recorded and its analysis suggested a few improvements, some of which could be
made in the Hawk flight test hardware and some of which were implemented in later versions of the hardware.
The damage detection capability of the Fibre Optic Strain gauge system is summarised in the table below:
Damage Properties
Number
Specimen
of sensors
Composite
4
Damage Detected 1 ?
During Test 2
Post processing
Extent
Location
Not attempted
Yes
No
Yes 3
Disbond
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 35 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
Damage Properties
Number
Specimen
of sensors
Bonded Metallic
3
Damage Detected 1 ?
During Test 2
Post processing
Extent
Location
Not attempted
Yes
No
Yes 3
Peel
Notes:
1. All methods of damage detection using strain gauges (Fibre Optic or Electrical) require prior knowledge
about what kind of damage is likely to occur and the expected location.
2. The algorithms used for post processing can be implemented in the firmware of the Fibre Optic
Interrogator to facilitate in-test detection.
3. The number of sensors and the spacing between them determines the accuracy of the location.
3.1.4
User comments
According to the Risø personnel involved in the ground testing, this prototype system did not (yet) display the
level of integration, robustness, development or completeness intended in the final system. Examples of this
lack include the necessity for unique software to connect directly with each sub-system in turn, the fragility of the
communications network, the lack of facilities in the de jure interface, the lack of local data handling, etc.
The sub-component testing did not take place simultaneously, with the combined SHMS capable of detecting all
damage types of interest. Due to the strong need for individual suppliers to validate their particular sub-system,
meant the focus of the Ground Test changed from being a trial of a system comprising several damage sensor
types, to being a test of various sensor sub-systems that shared a common data storage and download tool.
Network
Occasionally the network would become inaccessible and it was necessary to power cycle all the components
and let the system reboot. The network environment was not robust and problems occurred if sub-systems (and
other IP addresses) were disappearing and reappearing on the network. It was necessary to maintain the strict
IP allocations in order to have a moderately stable network.
Central Computer
The Central Computer appeared well advanced in terms of hardware development.
The status monitor provided for use during Ground Test turned out to be vital, supplying “live” information of the
composition of the network, the state of the system, and the operational condition of each sub-system. It was
also the preferred control interface for effecting system mode changes.
The amount of data transferred to the Central Computer (especially by the Fibre Optic sub-system) seemed to
be excessive and necessitated very long download times after loading periods.
Unexpected power outages (due to network inaccessibility for example) and data download without first
changing operating modes could result in corrupted or improperly “closed/truncated” data files.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 36 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
The Central Computer higher functionality with respect to configuration files, data directories, and log file
access, etc. was not integrated with the user interface (Notebook) requiring a direct connection to make any
changes.
Notebook
The notebook program was not successfully installed on a Risø laptop, and the system was accessed via a
DEMEX supplied computer with the GUI program pre-installed. This system was used only to connect to the
network and download data for analysis elsewhere (directly by the sensor suppliers and by the Notebook
program integrated evaluation tool). Occasionally it was not possible to make a Notebook connection to the
network, and this necessitated a direct FTP from the Central Computer data directory. None of the higher
function data evaluation tools was used during the ground test.
SGSS
The SGSS appeared very well advanced in terms of hardware development. In addition, this sub-system
included a local LED “status indication” display for the eight sensor channels (similar to the live status indication
provided for all the sub-systems by the Central Computer status monitor).
It was possible for the 8-channel SGSS to monitor different sub-components (structural sections) that were
being tested simultaneously. However, it was noted that the user was required to manually set “warning” and
“damage” limits via direct connection to the sub-system and using unique software tools. Also noted was the
fact that any changes in loading condition must also be reflected in a change to the user input limits, in order to
continue to obtain good damage information from the sub-system. Furthermore, the SG sensors are vulnerable
to “zero-drift” if the sub-component is placed in an unloaded state or handled for whatever reason, requiring
manual connection/correction again to ensure channels are still giving valid data.
The manual supplied to assist in using the SG interrogator was excellent, and after being guided through the
set-up once, it was then a simple (if time consuming) process to subsequently tune the channels.
Lamb Wave
The Lamb Wave module appeared to be the least developed hardware prototype used during the Ground Tests,
and required the most technician attention during the workshop to prepare it for the testing. During the Ground
Test however, this sub-system performed very well and supplied the most useful information to the status
monitor allowing a good correlation between the damage growth and the sensor response to be displayed “live”
during the testing.
The Lamb Wave sub-system higher functionality, with respect to configuration files for different structural
material and tests, was not integrated with the user interface (Notebook) and required a direct connection to be
made to the module with software tools provided by the sensor supplier.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 37 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
SWISS
Unfortunately, it was not possible to integrate the SWISS sub-system during the workshop, and the hurried
fatigue testing of sub-component 3 finally resulted in static failure, instead of damage growth. However, during
cycling the ultimate self-test function of the imaging system worked: reflections from the specimen boundary
have been repeatedly localized as a stable means of function prove, until when the arrays started to de-bond as
the data showed, while it was not optically visible in the test rig. Pushing the installation mechanically then
revealed de-bond. Nevertheless, the arrays were still working and have been reinstalled with an improved
process of bonding. Finally, the arrays survived further cycling, but finally also the arrays have been broken by
the static crack that occurred at about 10.000 microstrain. In summary, the tests did not give the expected good
results, apart from highlighting the importance of a correct sensor bonding procedure. It was also noted that the
SWISS system is operated entirely separately from the rest of the SHMS
Fibre Optic
The fibre optic hardware used in the Ground Test was not the system intended for use in the final SHMS. Off all
the sensor sub-systems, the FOSS required the least attention from Risø during the Ground Test. It provided no
information to the Central Computer status monitor beyond indicating its presence on or absence from the
network. However it was very noticeable during data downloads as the FOSS flooded the CC (and Notebook)
with high information content “level 3” data continuously, regardless of the state of the sub-components
instrumented.
Acoustic Emission
The AESS module appeared well advanced in terms of hardware development. The acoustic emission subsystems higher functionality with respect to configuration files for different structural materials and sensor
distributions was not integrated with the user interface (Notebook) and required a direct connection to be made
to the module with software tools provided by the sensor supplier.
During the Ground Test, the AESS displayed an occasional “sensor damaged” warnings on the Central
Computer status monitor, these warnings were inaccurate and were generated by an automatic sub-system
diagnostic routine.
3.2
Hawk flight test (WE2.5)
BAE Systems primary aim for WE2.5 of the AHMOS II programme (ERG103.015) was to demonstrate for the
first time, successful, operation of integrated automated damage detection systems, in- flight, aboard a BAE
Systems Hawk aircraft (Ref. 23). A further aim was to demonstrate the viability of selected damage detection
technologies to a Technology Readiness Level of seven (TRL7). Following on from the successful trial of an
automated damage detection system in RTP 3.20 AHMOS, the consortium focused on adapting the technology
to the Hawk flying test bed demonstrator programme within ERG 103.015.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 38 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
The system was designed to monitor damage growth on two structural items located (metallic crack growth and
metallic disbond) inside an underwing Hawk instrument pod during flight. The system will function as part of an
integrated SHM architecture comprising a number of SHM technologies from project partners. Ground-based
trials of the technology, utilizing other partners’ facilities, will underpin and supplement the flight trials. The
success of this programme will provide evidence for all stakeholders that an integrated, automated damage
detection SHM system is proven to TRL7.
The down-selected sensor technologies were:
- Acoustic Emission, BAE Systems ATC
- Strain Gauge, Emmecon
- Fibre Optics, Smart Fibres
- SWISS, KT Systems
- Lamb Wave, QinetiQ
Also contained within the Hawk pod system, was a communication Ethernet Hub Switch to allow the sensor
equipment boxes and the central computer, utilized for data storage, to communicate with each other. The
central computer data could then be downloaded to the laptop both for analysis and to generate a real-time
presentation of structural health provided via the Graphical User Interface (GUI).
3.2.1
Acoustic Emission
The acoustic emission systems has demonstrated the ability to detect the growth of cracks in aluminum allow
structure in flight aboard a fast military jet while undergoing unrestricted flight maneuvers at a range of altitudes.
Moreover, the trials have indicated a promising resilience to ‘false positives’ as demonstrated by clear
delineation in systems response to damaged and undamaged specimens.
The acoustic emission system operated successfully throughout the flight tests. Very weak acoustic emission
events were detected in a noisy, high vibration, high acceleration, and high- altitude environment. Acoustic
emission events were also detected despite high levels of electrical interference from other on-board systems.
The processes put in place to screen out spurious signals were only partially successful necessitating some
post analysis of data. Examination of the spurious signals that were not rejected will help us to refine our
screening process. This work is currently ongoing.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 39 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
The damage detection capability of the Acoustic Emission system is summarised in the table below:
Damage
Number
Specimen
Metallic, fully
of
Damage Detected?
Properties
sensors
During Test
Post processing
Extent
Location
3
Not attempted
Yes
No
Yes
instrumented
(localized
to flight
specimen)
Metallic, partly
3
Not attempted
Yes
No
Yes
instrumented
(localized
to flight
specimen)
3.2.2
Strain Gauge
The SGSS developed in AHMOS II has proven to be rigid and reliable in extreme conditions by mechanical and
electrical construction and by software. Emmecon’s software platform used also in acoustic emission and lambwave systems has proven to be flexible and reliable. SGSS is versatile and could be modified for loads
monitoring based life assessment with software modifications.
The detection of damage based on strain measuring requires solid understanding and/or modeling of the
structure. The sensors have to be located so that the strain caused by the normal loads and changes in strain
caused by the damages can be isolated. In ground and flying test benches of the AHMOS II, the cracks were
reliably captured with SGSS in others but not in Hawk-pod.
The damage detection capability of the Strain Gauge system is summarised in the table below:
Damage
Number
Specimen
Metallic, fully
of
Damage Detected
Properties
sensors
During Test 1,2
Extent
Location
3
No 3
No 3
No 3
n.a.
Not attempted
No
No
instrumented
Metallic, partly
instrumented
Notes:
1. All methods of damage detection using strain gauges (Fibre Optic or Electrical) require prior
knowledge about what kind of damage is likely to occur and the expected location.
2. Post processing is included in the run time firmware of the Strain Gauge system.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 40 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
3. Damage tolerance parameters of the damage assessment algorithm were set too tight which
caused false alarms. If specimen used in Hawk pod had been tested in laboratory in shaker with
representative spectrum, the Hawk test would most likely had been successful, too.
3.2.3
Fibre Optics
The interrogator hardware was qualified for flight worthiness in time for the flight programme. Flight-testing was
successful and delivered high quality data.
There were no significant failures, however there were problems obtaining the desired FO connectors in time for
the pod sensor install, the lower spec alternatives caused some problems, which meant we had to actively align
connectors whenever new test specimens were installed. The testing and analysis required to successfully
implement damage detection algorithms were under-estimated at the outset of the project. Meaning that the
equipment has not completely fulfilled its potential in this aspect.
The damage detection capability of the Fibre Optic Strain gauge system is summarised in the table below:
Damage
Number
Specimen
Metallic, fully
of
Damage Detected 1 ?
Properties
sensors
During Test 2
Post processing
Extent
Location
2
Not attempted
No 3
No
No
1
Not attempted
Yes 4
No
No
instrumented
Metallic, partly
instrumented
Notes:
1. All methods of damage detection using strain gauges (Fibre Optic or Electrical) require prior
knowledge about what kind of damage is likely to occur and the expected location.
2. There is no reason why the algorithms used for post processing could not now be implemented in
the firmware of the Fibre Optic Interrogator to facilitate in-test detection.
3. Method attempted during post processing was to measure differences in strain fields of two
sensors at different distances from the area of crack growth. We learnt that to successfully apply
this method would likely require more than two sensors.
4. Method used during post processing was to measure changes in the resonant frequency of the
specimen when excited by aircraft maneuvers.
3.2.4
SWISS
Complementary to the other successful damage detection techniques the imaging ultrasound technique and the
associated SWISS equipment has been used to support the objective of this project to “Demonstrate operational
prototype SHMS based on mature enough sensing techniques by flight test demonstration supported by
independent full scale on-ground damage detection tests” (ref. contract ERG103.015).
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 41 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
For this purpose, pre-integrated broadband ultrasound matrix arrays have been installed to a metallic specimen
in the Hawk test bed mini-lab to detect and localise damage that would emerge as result of vibrational loads.
A crack of length 6 ± 2 mm has been detected, localised and quantified.
The pre-integrated arrays allow quick and practical installation, avoiding soldering or similar on a weapon
system platform’s structural item. To qualify for flight tests the given type of installation has undergone an
airworthiness test programme, with only minor deviations, which were acceptable. An improved surface
preparation method guarantees extraordinary life times.
Throughout this programme the weight of cabling issue remained, because better alternatives were not
available and if, they are for industrial applications only (like the analogue bus system of the KTS220).
The off-board equipment, necessary for operating the sensors installed on board has been miniaturised
considerably (from one cubic metre to portable), but not to the extent that would be technologically possible (as
has been implemented for a KTS220) because the risk that necessary electronic components could fail was
considered to be too high without prior analysis of maturity. Compared to AHMOS II technology the KTS220
integrates substantial parts of electronics with transducers and thereby eliminates the cabling issue and
improves the system performance: higher sensitivity due to lower noise, less cross-talk, reduced susceptibility to
EMI.
In general permanently mounted imaging ultrasound arrays offer the same or better capabilities of ultrasound
monitoring, but also a breakthrough in cost reduction. 112 channels have been used in the case here involving
tremendous costs – not for the sensor, but for the wiring. Industrial applications (KTS220) offer between 16 to
up to 1000 channels and more. Every added channel costs only 50€ so that adding redundancy to the
monitoring sensor network becomes affordable.
Since natural echoes, such as holes, corners or boundaries offer an ultimate self-test capability a false alarm
rate of 0% has been demonstrated here and a theoretical maximum of 3% estimated thereof. A partial
degradation, non-functioning or loss of some elements in a multi-element network is to some extent backed-up
by the remainder of the sensor-network and does not imply that the hole system fails due to local impairments.
It seems reasonable, to also exploit the same technology to measure deformations in areas, where sensors
cannot be placed (e.g. hidden corners), to detect plastic deformations or measure deformations on-line.
Future activities should address the need to specify actual monitoring tasks which possibly could replace
existing visual, eddy-current, ultrasound or visual inspection, offering similar capabilities as NDI and requiring
similar expertise, but eliminating the need for disassembly, re-assembly and rechecking. For these monitoring
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 42 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
needs, validations –similar to existing NDI – should be conducted by challenging industrial next generation
versions of imaging ultrasound equipment for use on military platforms.
The damage detection capability of the SWISS system is summarised in the table below:
Damage
Number
Specimen
of
Damage Detected 3
Properties
sensors
During Test
Post processing
Extent
Location
Metallic, fully
112 in
Never, by
Yes 2
Yes
Yes
instrumented
three
principle1
6 mm
groups
Notes:
1. Imaging Ultrasound with SWISS is an interval based inspection technique. Data is acquired during
a test or at intervals. There from independent, the imaging process and subsequent interpretation
of the data is done in a post-data-acquisition point in time and allows post-measurement choice of
wave types and material parameters in contrast to conventional imaging equipment that requires
selecting these parameters at the time of operation.
2. A crack of 6mm has been detected as the specimen was damaged after the last inspection interval.
A 2 mm accuracy is given by the physics of the sensor system used here to detect a crack between
1 and 3 mm as requested. Previous inspections revealed only minor changes, which suggest that
the damage was below 1 to 3 mm.
3. The damage has been found as a spatial increase (of about 6mm) of the original echo that resulted
from the notch that was in the specimen. Whether this quantification is correct could not be
checked yet, because other installations (strain gauges) cover the area of interest. In another hot
spot area on the specimen (opposite side) it has not been as clear due to physical reasons, which
seem reasonable and are documented in the report. The ultimate self-test capability (imaging of a
given echo - the original notch) has been proven and a limited tolerance, if some individual sensor
element channels fail has been demonstrated.
3.2.5
Lamb wave
The Lamb-wave system was applied to three specimens where it performed without false call throughout the
flight test programme and successfully detected and tracked the disbonding in two of the specimens. It is
considered that the performance of the Lamb-wave system would have provided sufficient evidence to support a
Technology Readiness Level of 7 for this particular application to bondline monitoring, but that this must be
reduced to 6 owing to the temperature-compensation issue.
The configuration of the trial SHM system isolated within an aircraft underwing pod proved highly successful.
The Lamb-wave specimen design, together with the ability to specify the profile of each flight test, successfully
enabled a disbond to be propagated within the short duration of a single sortie. The scheduling of the flight test
programme was beyond the control of the AHMOS team and the success of the Lamb-wave flight testing relied
heavily on the ability to quickly change specimens in the short turn-round servicing period between flights.
However, the very close scheduling of flight tests meant that there was no time for data analysis or remedial
action between flights following Flight 3. Project resources were sufficient for only three instrumented Lambissue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 43 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
wave specimens and since the high-altitude flying was scheduled for Flight 7 the last specimen had to be
preserved through Flights 5 and 6. No damage was induced in this final specimen and, since there was also
insufficient time to repair the damaged temperature-compensation sensor, Flights 5, 6 and 7 provided neither
damage detection nor temperature-compensation data, so their value in respect of the Lamb-wave system
assessment was marginal. Further specimens and more time between flights could have circumvented these
difficulties and provided greater benefit from flight-testing.
The damage detection capability of the Lamb Wave system is summarised in the table below:
Damage
Number
Specimen
Lamb-wave
Damage Detected 1 ?
of
Properties
sensors
During Test 2
Post processing
Extent
Location
2
Yes
Yes
No 1
Yes 2
2
Yes
Yes
No 1
Yes 2
2
No 3
No 3
No 3
No 3
specimens
(FL1)
Lamb-wave
specimens
(FL2)
Lamb wave
specimens
(FL3)
Notes:
1. The system was unable to determine the extent of disbond during flight owing to a problem with the
temperature compensation scheme.
2. The system was configured to monitor a specified bond and the location of disbonding was known
a-priori.
3. No damage was generated during flight-testing.
3.2.6
Graphical User Interface
According to the BAE partners there has been no problems using the program (Ref. 32). The overall feedback is
of a quick user-friendly program, which has shown no signs of breakdown during the test. This is in respect of
the fact, that the program mainly was used for quick and safe data transfers from the onboard equipment to the
laptop. From earlier tests, the program as a whole has been useful, accompanied with suggestions to
improvements.
3.2.7
Conclusion
It has been shown that SHM systems can function in an aviation environment. Seven flights were performed.
Strain Gauge sensors and Swiss sensors flew only for two flights.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 44 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
AE operated successfully during all flights. Damage was detected. Interference occurred from outside factors
(overseal separation of SWISS sensor detected, presence of Fibre Optic, Strain Gauge and SWISS sensor
increased dampening effect on the specimen, damage occurring not on specimen but elsewhere in the pod).
Strain Gauge controller successfully operated during all flights, functioned and detected damage when the
sensors were installed. Sensor location is critical.
Fibre Optics successfully operated during all flights; damage was detected but not quantified. This requires
further development of algorithms.
SWISS was installed on the first two flights. Damage was detected and quantified.
Lamb wave successfully operated during all flights. Damage was detected but not quantified due to a
temperature-compensation algorithm that must be further developed.
3.3
F-18 flight test (WE2.6)
The purpose of the F-18 AHMOS-pod flight testing was (Ref. 24):
1.
Test the operation of the strain gauge and SWISS systems in actual military aviation environment.
Because of the nature of the systems (COTS components used, simplified ground testing etc.) and the
test environment (severe conditions at the wing tip) it was possible that problems were encountered. In
these cases, reasons for the problems and possible solutions to them were studied within the resources
available for the project.
2.
Test the capability of the systems to detect cracks in these circumstances. The fatigue test set-up has
been extensively tested on the ground, so there was a relatively good prediction how and when the test
specimens will crack. This is why the flight test was only used to show that the systems operate similarly
compared to ground testing. Emergence of cracks was to be estimated comparing the ground test
results, traffic light indications, strain measurements and SWISS sensor images. Also, it was planned, if
necessary, to detect the cracks with other NDT methods.
3.
Collect practical experience from the FiAF personnel of the operation of the systems. Because the
nature of the flight-testing was development testing, not certification testing, no special operational
requirements were set for the tested systems.
The F-18 AHMOS-pod, which is designed for the AHMOS-project, is a look-alike of the AIM-9 CATM missile,
which is the training version of the AIM-9 Sidewinder air-to-air missile. Inside the AHMOS-pod, there are test
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 45 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
specimens equipped with strain gauges and SWISS sensors, equipment to load the specimens, AHMOS-HWSGSS strain gauge measuring and analysis unit and data network.
The specimens were designed to have a fatigue life of approximately 100 hours, which was assumed to be
roughly divided in 50 flight hours and 50 hours of activities when the electrical system is switched on but the
aircraft is not flying (inspections, engine start-up, taxing etc.).
3.3.1
Strain Gauge
The operation of the AHMOS-HW-SGSS strain gauge measuring and analysis unit has been monitored using
the traffic lights and the data recorded by the system. During the first five flights, monitoring was possible from
the traffic lights as long as the aircraft started to taxi for take-off. During the first flight, just as the plane started
to move, the traffic lights showed that the system fell down and started to reboot immediately. This was also
visible from the recorded data. Based on the recorded data there is no other indication that the system has
fallen down.
The damage detection capability of the Strain Gauge system is summarised in the table below:
Damage Properties
Number
Number
Metallic, fully
of
of
Damage Detected
instrumented
sensors3
channels3
During Test 1,2,3,5
specimen no.
25
4
2
LEFT
RIGHT
Extent
Location
No damage
Detected
No4
No4
No4
No4
No4
No4
No4
No4
Damage,
26
27
4
4
2
2
Not detected
Detected
damage
Damage,
No damage
Not detected
damage
28
4
2
Detected
Detected
Damage,
Damage,
Notes:
1. All methods of damage detection using strain gauges (Fibre Optic or Electrical) require prior
knowledge about what kind of damage is likely to occur and the expected location.
2. Post processing is included in the run time firmware of the Strain Gauge system.
3. Each specimen is instrumented with four strain gages. Two upper side strain gages form a half
bridge connected into one measuring channel of the SGSS. Two lower side strain gages form a
half bridge connected into one measuring channel of the SGSS. Upper side LEFT and lower side
LEFT strain gages are in same position but on opposite sides of the specimen. Upper side RIGHT
and lower side RIGHT strain gages are in same position but on opposite sides of the specimen.
4. The number of sensors and the spacing between them determines the accuracy of the location.
5. [24, chapter 3.4.4]
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 46 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
3.3.2
SWISS
The hardware for making the SWISS system inspections were delivered to Patria in June and the software in
July 2007, more than a year later than planned in the schedule. This is why the flight-testing was started without
the system. The first attempts to carry out SWISS system inspections were done in August. However, the first
inspection was succeeded in early December using other equipment than originally intended for the AHMOS
project. This meant that no data was available from earlier inspections to be used as a reference, which would
be necessary to detect and localize damage smaller than 2 to 3 mm.
The inspections done in early December were done only with SWISS sensors of an earlier generation, not with
the next generation sensor. When making the inspections it was noticed that one sensor did not respond. The
reason was found to be a faulty cable. Because the pod was disassembled it was possible to make the
inspection using another sensor’s cable.
The SWISS sensors were on test specimens 25, 26 and 27. In general, it seemed that the sensors had survived
the flight-testing. The SWISS system inspections indicated that if there are cracks they are bigger in specimen
25 than in 27 because the ultrasound changes were very small. The image did not clearly indicate possible
cracks or the artificial notches. The reasons for this are not yet understood. The changes identified by the
SWISS sensors in the ultrasonic response indicate that the damage is very asymmetric, i.e. probably on one
side only.
The results also indicate a possible debond of the sensor on specimen 26 because the changes in the
ultrasound response were unexpectedly high - although a larger crack would be expected to result in a similar
change. However, conventional NDT only reported small cracks. Only imaging could bring up further help to
interpret the data (similar to Hawk tests), but first attempts did not result in a satisfying image. The sensor does
not move and the bonding seems intact so it is possible that the debonding originates to the bonding process.
No conclusions can be done of the status of this test specimen, because the changes are significantly larger
than the changes found in 25 and 27.
The damage detection capability of the SWISS system is summarised in the table below:
Damage
Number
Specimen
of
Damage Detected 2
Properties
sensors
During Test
Post processing
Extent
Location
Metallic, fully
16 in
Never, by
No 2
No
Small
instrumented
one
principle1
<2
changes
mm
near
group
sensor
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 47 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
Damage
Number
Specimen
Damage Detected 2
of
Properties
sensors
During Test
Post processing
Extent
Location
Metallic, fully
16 in
Never, by
No 2
No
Small
instrumented
one
principle1
<2
changes
mm
near
group
sensor
Metallic, fully
16 in
Never, by
instrumented
one
principle1
Yes 2
group
No,
Large
Large
changes
crack
near
or
sensor
debond
or both
Notes:
1. Imaging Ultrasound with SWISS is an interval based inspection technique. Data is acquired during
a test or at intervals. There from independent, the imaging process and subsequent interpretation
of the data is done in a post-data-acquisition point in time and allows post-measurement choice of
wave types and material parameters in contrast to conventional imaging equipment that requires
selecting these parameters at the time of operation.
2. A 2 mm accuracy is given by the physics of the sensor system used here to detect a crack between
1 and 3 mm as requested. Inspections revealed only minor changes, which suggest that the
damage was below 1 to 3 mm.
3.3.3
Graphical User Interface
The DEMEX GUI software (Ref. 33) has operated smoothly during the tests. However, there has not been a
need to make many inspections and relatively few people have used the system.
Not all the indications of the software were in harmony with the measured strains, because after flight thirteen
the green light was shown for both sensor and structural health status. This was not logical for such a large
change. The reason for this turned out to be different interpretations of the definitions. Because of this, the
behaviour of the Demex software traffic lights and the pod traffic lights was not analogous. This minor nonuniformity can be easily corrected from the Demex or AHMOS-HW-SGSS unit software.
3.3.4
Conclusion
The Strain Gauge measuring and analysis unit operated normally during the flight-testing. The strains indicated
the emergence of four of the six cracks found with traditional NDT methods.
The SWISS sensors survived the flight-testing and strain levels in the test specimens. The SWISS system did
not detect the cracks with certainty because they were below 1,6 mm which is below detectability of the 2 MHz
transducer.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 48 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
Except for an EMI problem (interference on the cockpit audio system) on 3 out of 4 aircraft used, the system
passed all tests. The EMI problem could not be resolved in the timeframe of this project because of the slow
progress of the flight-testing.
3.4
3.4.1
Qualification Aspects (WE2.1)
Software qualification
Whenever a system is developed for commercial or military aircraft the safety of the crew and occupants is a
factor in the development of the system. Airworthiness authorities established various procedures to qualify that
the system meets the safety critical objectives.
At the highest level, the Authorities require that the aircraft is qualified and that the qualification includes all
different systems. Within the systems, any implemented software must be qualified.
The qualification plan is the document that establishes the criticality of the system and its various components.
This determines the level of design assurance required by the regulations. It also establishes the software
approval levels as required by the software development standard RTCA Inc. DO-178B.
As part of the qualification process, the system safety assessment is used to determine and categorize the
failure conditions and the criticality of a system. The objective is to identify requirements that will preclude or
limit the effect of faults, and may result in fault detection and fault tolerance requirements being added.
The failure condition categories used by DO-178B are:
A.
Catastrophic.
B.
Hazardous/Severe-Major.
C.
Major.
D.
Minor.
E.
No effect.
A DO-178B criticality level is assigned to the software under development based on the contribution that the
software may make to potential failure conditions, as determined by the system safety assessment process.
Table 3.4.1 shows the relationship between the system safety definition and the software criticality level and
how it relates to the verification objectives as defined by DO-178B section 6.4.4.2.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 49 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
Table 3.4.1 DO-178B Software Criticality Levels
Software Criticality
Definition
Associated Structural Coverage
Level
Level objectives (DO-178B
6.4.4.2)
Level A
Level B
Software that could cause or contribute to the
Modified Condition/Decision
failure of the system resulting in a catastrophic
Coverage, Decision Coverage &
failure condition.
Statement Coverage
Software that could cause or contribute to the
Decision Coverage & Statement
failure of the system resulting in a hazardous
Coverage
or severe-major failure condition.
Level C
Software that could cause or contribute to the
Statement Coverage
failure of the system resulting in a major failure
condition.
Level D
Software that could cause or contribute to the
Not required
failure of the system resulting in a minor failure
condition.
Level E
Software that could cause or contribute to the
Not required
failure of the system resulting in no effect on
the system.
The verification objectives for DO-178B are set in place to detect and report errors that may have been
introduced during development processes. Software verification objectives are satisfied through a combination
of reviews and analyses, the development of test cases and procedures, and the execution of those test
procedures. Reviews and analyses provide an assessment of the accuracy, completeness, and verifiability of
the software requirements, software architecture, and source code.
The Software Criticality Level of at least C should be considered at the commercial introduction of a SHMS in
(military) aircraft. This means that Statement Coverage applies for the Structural Coverage Analyses Objective.
Statement Coverage means that every statement in the program has been invoked at least once.
During software development in this project, however, the failure of the SHMS will not have any influence on the
aircraft safety, which means that software criticality level E (refer to table 1) applies. We advise, however, to do
software qualification for a limited part of the software as developed by the various partners during the project in
order to experience the associated work.
3.4.2
Hardware qualification
During AHMOS II extensive amount of hardware qualification testing has been performed.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 50 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
All the system parts that have been installed in the Hawk pod have successfully been tested on the ground (Ref.
29) and in flight either by suppliers or by BAE Systems.
All the system parts that have been installed in the F-18 pod have also been ground (Ref. 29) and flight-tested.
Except for an EMI problem (interference on the cockpit audio system) on 3 out of 4 aircraft used, the system
passed all tests. The EMI problem could not be resolved in the timeframe of this project because of the slow
progress of the flight-testing.
All test results are documented in the Qualification Program Plan (Ref. 26).
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 51 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
4
TECHNOLOGICAL READINESS LEVEL OF SYSTEMS
Another task in WP3 is to include an overview of the Technology Readiness Level (TRL) of the different sensor
techniques and sub-systems. All AHMOS II partners were invited to define their key system development(s) in
the AHMOS II programme (e.g. AE on composite/metal structures, embedded/surface mounted FBG, Lamb
wave, SGI, GUI, etc.) together with main the detection/research objective (impact damage, surface crack, subsurface crack around fasteners, skin-stiffener debond, residual stress, chemical sensing, etc.), TRL rating,
substantiation of the TRL (e.g. in cases where a relatively low TRL was caused by failure of other sub-systems),
main advantages, main limitations, and future needs. The key systems can be systems developed and
evaluated in laboratory environment, during ground tests (e.g. at Risø), and during the 4 flight tests in UK,
Finland, Spain and Italy. Therefore, it can include research and development activities in both WP1 and WP2 of
the AHMOS II project.
4.1
TRL Rating
GAO/NSIAD-99-162 Best Practices (Appendix 1) definition of TRL, which is presented in table 4.1, has been
used (Ref. 25). The discussion between experts showed that the given definitions of TRL levels could be
interpreted in various ways. Thus it has been rather difficult to agree on a unique TRL for some subsystem
modules, meaning that an error of plus or minus 0.5 is still expected. Still this is a good indication of the overall
TRL, which can range from one to nine.
Table 4.1 Definition of Technology Readiness Level (TRL)
Technology readiness level
Description
1. Basic principles observed and
Lowest level of technology readiness. Scientific research begins to be
reported.
translated into applied research and development. Examples might
include paper studies of a technology’s basic properties.
2. Technology concept and/or
application formulated.
Invention begins. Once basic principles are observed, practical
applications can be inverted. The application is speculative and there is
no proof or detailed analysis to support the assumption. Examples are
still limited to paper studies.
3. Analytical and experimental
Active research and development is initiated. This includes analytical
critical function and/or
studies and laboratory studies to physical validate analytical predictions
characteristics proof of
of separate elements of the technology. Examples include components
concept.
that are not yet integrated or representative.
4. Component and/or
Basic technological components are integrated to establish that the
breadboard validation in
pieces will work together. This is relatively 'low fidelity' compared to the
laboratory environment.
eventual system. Examples include integration of "ad hoc" hardware in a
laboratory.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 52 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
Technology readiness level
Description
5. Component and/or
Fidelity of breadboard technology increases significantly. The basic
breadboard validation on
technological components are integrated with reasonably realistic
relevant environment.
supporting elements so that the technology can be tested in a simulated
environment. Examples include "high fidelity” laboratory integration of
components.
6. System/subsystem model or
Representative model or prototype system, which is well beyond the
prototype demonstration in a
breadboard tested for TRL 5, is tested in a relevant environment.
relevant environment.
Represents a major step up in a technology’s demonstrated readiness.
Examples include testing a prototype in a high fidelity laboratory
environment or in a simulated operational environment.
7. System prototype
Prototype near or at planned operational system. Represents a major
demonstration in an
step from TRL 6, requiring the demonstration of an actual system
operational environment.
prototype in an operation environment, such as in an aircraft, vehicle or
space. Examples include testing the prototype in a test bed aircraft.
8. Actual system concept and
Technology has been proven to work in its final form and under expected
"flight qualified' through test
conditions. In almost all cases, this TRL represents the end of true
and demonstration.
system development. Examples include development test and evaluation
of the system in its intended weapon system to determine if it meets
design specifications.
9. Actual system “flight proven"
Actual application of the technology in its final form and under mission
through successful mission
conditions, such s those encountered in operational test and evaluation.
operations.
In almost all cases, this is the end of the last 'bug fixing' aspects of true
system development. Examples include using the system under
operational mission conditions.
4.2
TRL Evaluation
The result of the inventory is given in table 4.2. The table shows that the TRL varies from value 3 (analytical and
experimental critical function and/or characteristics proof of concept) to value 7 (system prototype
demonstration in an operational environment). TRL 7 was reached for the AE (crack detection in metal
structure), corrosion sensor, FBG (strain measurement), SGI and SWISS sensor systems. Also, the system
network and modules, GUI and central computer were rated with a TRL 7. This is an appreciable improvement
in comparison with the TRL values obtained in the predecessor project AHMOS I. In that project the maximum
TRL value was only 5 (component and/or breadboard validation on relevant environment), reached for the
surface mounted FBG and the SWISS system.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 53 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
Table 4.2 AHMOS II System Evaluation
AHMOS
partner
EADS-G /
WIWEB
SHM
System
SWISS
Detection/research
objective
Localisation of both
defects and selected
boundaries to check
structural integrity at
required confidence in
hidden and/or
uninstrumented areas.
TRL
7
Covers otherwise called
“phased array
ultrasound” or similar
techniques.
Substantiation of TRL
Advantages
Limitations
Future needs
7:
Hawk flight trials
- Localisation of artificial
defect (hole)
- Repeatability of
undamaged boundary
(hole)
- Robust instrumentation
- Retrofit
- Low cost multiplexing
- Comparable NDI needs
- Potential to deal with
unexpected damage
evolution
- Potential to detect as
sensitive as existing NDI
- Potential to merge with
techniques that require
similar hardware
- Expert independent
automatic offline multimode interrogation
- Expert post-analysis on
demand, without need for
availability of platform
- No on-board software
changes necessary
- Redundancy: even if some
individual elements fail, the
monitoring capability may
not be affected
- Robustness
- Multiplexing
- EMC immunity
- Robustness
- Multiplexing
- Sensitivity
- Versatility depending on
the material choice
- Relative range-reduction
for monitoring of nonlinearities
- Requires more
acquisition time than NDI
(~min,~h)
- Requires built-in growth
potential at time of retrofit
(e.g. 28 channels, when
only 11 are needed) to
cover extended spots,
conventionally covered by
moving NDI
- Only physical data
evaluation. No empirical
- Global material changes
can only be detected if
they result in a significant
change of the scatter
distribution
- Improvement of
sensor material and
structural miniature
design (high
technological risk)
- Improvement of
off-board electronics
to exploit second
breakthrough in
range
- Improve software to
automate parameter
selection
- Alternative
approach to currently
known global (non
NDI) monitoring
concepts: requires
elaboration of future
case specifications in
parallel to low TRL
development
Point technology
Capability to monitor
an entire area of a
component
- One-chip
engineering
- Monitoring on
selectable chemicals
for specific
applications
Remark:
TRL 7 applies to above
determined test case and
used implementation.
Implementation readiness
is by definition
significantly lower.
Significantly improved
implementation concepts
start at lower TRL (e.g. 3).
Alenia
FBG
Strain measurement
5
CIRA
FBG
Strain measurement
5
FO,
reflectometr.
and LPG
approach
Chemical sensing
6
- Qualification test for on
ground application
- Flight test on C27
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
No real weakness
- Packaging
- Reliable interrogation
systems
page 54 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
AHMOS
partner
BAE
Systems
QinetiQ
SHM
System
AE
Detection/research
objective
Crack in metal
structures
TRL
Substantiation of TRL
Advantages
Limitations
Future needs
7
Hawk flight trials of AE
system
- Damage sizing difficult
- Requires live load to
‘activate’ the acoustic
emissions so that an online system is required
-Methodology for ‘in
service’ use
- In-service
validation of data,
confer conventional
inspection methods
Metal stringer disbond
6
Risø lab tests
- Wide area coverage
- No dependence on
defining initial undamaged
state
- Less susceptible than UT
based systems to sensor
degradation (i.e. method
depends largely on AE
signal frequency
characteristics rather than
absolute signal magnitudes)
As above
As above
6
Corrosion
sensor
Composite stringer
disbond
Indication likely
occurrence of corrosion
7
Metal stringer disbond
6/7
Assumes surrounding
structure corrodes at the
same rate as the sensor
- Resolution is dependent
on the size of the sensor
- Structurally specific
- Requires a baseline of the
initial undamaged structure
In-service validation
of data
LW
Risø ground tests and Joint
lab tests with QinetiQ
Advanced lab trials plus
other military aircraft trials
outside of AHMOS
System was successfully
proven on the BAE Hawk
flight test. LW system
operated in flight for the
duration of all of the
flights and correctly
classified damage to the
specimens in real time
As above although
continuous monitoring can
plot progress of linear
disbonds as they propagate
hence some sizing possible
As above
As above
Needs no on-line
instrumentation (passive
witness sensor)
- Active system, can operate
on ground
- Can be configured for
plug-in ground operation
with reduced weight
- Can be configured to
detect other defect types
- System does not require
continuous operation to
monitor the structure (can be
employed on periodic basis)
- Low cost of system; low
cost & weight of sensors
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
As above
- Improved temp.
compensation
- Sensor robustness
page 55 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
AHMOS
partner
Smart
Fibres
SHM
System
FBG
Detection/research
objective
Composite stringer
disbond
TRL
6
FBG strain
measurement
7
Crack in metal
structures
N/A
Substantiation of TRL
Advantages
Limitations
Future needs
System was successfully
proven on the Risø ground
test. LW system operated
without error for the
duration of the test and
correctly measured and
classified damage of the
specimen. Since the
composite specimen was
not flight tested, the TRL
is lower than the Metal
stringer disbond
Strain measurements made
with FBG sensors during
all 7 Hawk flight tests
As above
As above
As above
- Strain/load measurements
can enhance the capability
of other SHM techniques
- Sensors are small, immune
to EMI, intrinsically safe
and several can be
multiplexed on a single
optical fibre if desired
- Cabling is lighter than
copper cable and less is
needed for a given number
of sensors
- Performance is as good as
or better than comparable
electrical strain gauges
See above
- Interrogator currently has
maximum temperature
rating of 55 °C
- Minimum bend radius of
optical fibre may limit
installation flexibility, but
this can be worked around
- Active cooling of
laser
- Reduction in size
and weight.
Capability of flighttested model was 18
FBG/kg. Current
model is 27 FBG/kg,
target is 50 FBG/kg.
- Investigate
implementation of
sensors in bendtolerant fibre (5 mm
allowable bend
radius)
- Study of optimal
sensor placement for
damage detection
- Develop algorithm
or process to help
designer choose
number and location
of sensors
- Investigate
possibility of
Playback of data on ground
revealed strain changes
associated with crack
growth. Algorithm to
perform automatic
detection was suggested
but not implemented in
Hawk flight tests.
However, the hardware
was flight-tested
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
A priori knowledge of
crack location required
page 56 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
AHMOS
partner
SHM
System
Detection/research
objective
Metal stringer disbond
Airbus (S)
FBG based
damage
sensing on
CFRP
EADSCASA
FBG bonded
on metallic
surface
Composite stringer
disbond
Composite stringer
disbond
- Strain/Stress
measurement
- Calculation fatigue life
consumption
- Load path failure
TRL
4/5
4/5
Substantiation of TRL
Advantages
Limitations
Future needs
Disbond detected during
Risø ground tests. System
was a prototype, not
ruggedised for flight,
hence lower TRL (4 for
interrogator and 5 for
sensors)
See above
See above
Disbond needs to be close
to a sensor. Note that many
sensors can be multiplexed
on one fibre to allow that
fibre to detect and track a
significant length of
disbond
combining
strain/load
measurements
directly with another
SHM technique to
produce a composite
damage indication
- Demodulation system
requires certain level of
spectral measurement
capability (secondary peak
detection or full spectrum
measurement capability),
not required for strain
measurement systems
- Disbond needs to affect
the bonding line area
where the FBG is
embedded. Note that many
sensors can be multiplexed
on one fibre to allow that
fibre to detect and track a
significant length of
disbond
Improve
optoelectronics and
algorithms to reduce
/ avoid coupling of
longitudinal/transvers
e and its effect on the
signal to noise ratio
- Equipment needs
auxiliary laptop
- Use of FBG sensor to
monitor temperature (due
to the difficult uncoupling
of load and temperature
effects)
- Fragility of the fibre
(several loops of fibre were
necessary to be installed in
the Ground Test)
- Analysis new
installation to use
FBG sensors to
monitor temperature
- Temperature
compensation
- Comparison of
FBG and strain
gauge measurements
- FBG sensors to
monitor damages
4/5
Disbond detected during
ground tests. System used
is a standalone equipment
7
- Ground Test “EF-2000
Slat Fatigue Test”
- Flight test trials on C101
See above
- Embedded optical path and
sensors (high level of
integration-protection)
- Compatible with strain
measurement systems
- Strain/load measurements
can enhance the capability
of other SHM techniques
- Sensors are small, immune
to EMI, intrinsically safe
and several can be
multiplexed on a single
optical fibre if desired
- Cabling is lighter than
copper cable and less is
needed for a given number
of sensors
- Performance is as good as
or better than comparable
electrical strain gauges
- Multiplexing of several
sensors per fibre
- EMC immunity
- Small size of sensors
(allowing to monitor areas
that actually can not been
monitored with strain
gauges)
- Easy sensor installation in
comparison with strain
gauges
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 57 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
AHMOS
partner
INTA
SHM
System
FBG
embedded in
composites
Detection/research
objective
Strain measurement
NIRAS
DEMEX
GUI
Development of GUI
software for retrieving,
joining and presenting
the structural health
status data to end user
7
AE –
Marine trial
Zonal profiling
inspection demo under
normal operation
3
AE –
Anakron
flight
AE activity output
integrated with standard
flight measurements
5
AE - turbine
Continuous
(12months+) zonal
profiling under normal
operation
6
Crack in metal
structures
7
Metal stringer disbond
7
Risø
VTT
SGI
TRL
5
Substantiation of TRL
Advantages
Limitations
Future needs
INTA Composite Material
Department organization
and facilities do not aim at
reaching a higher TRL.
The aim of INTA
participation is to provide
results and experiences so
that Airbus España and/or
EADS-CASA can go
beyond in TRL
- Not susceptible to EMI
- Size of sensors
- No significant influence on
host material properties
- Reliability in extreme
operative environment
conditions
-High robustness of the
embedded sensors
(including cables and
connectors)
- Movable 3D-illustration
- Statistics
- COTS, and price
- User friendly
- Adjustable to different
sensors and flight types
SWE output correlated with
location and operating
environment
- Cross sensitivity of
temperature/strain
- Influence of temperature
and humidity on strain
sensitivity
- Limited repeatability
observed in acrylic-coated
sensors (due to hysteresis
after thermal cycling,
minor problems due to
humidity)
- Off-board use only
- Implementation of
damage sizing
- Temperature
compensation
- Test campaigns on
embedded sensors
with other coating
materials different
from acrylic
- COTS
- Manual inspection only
- Further analytical
and experimental
function
- Installation system
development
- Integrated system
- Signal
discrimination
- Planer localisation
- Planer localisation
- Signal
discrimination
- Improved data
integration
Proof-of-concept demo
with commercial system
only
- Instant SWE output visible
on flight controls
- Hard landing assessment
Operating environment
only
Modified commercial
system only
SWE profile history for the
structures correlate with
operating environment
- Exception analysis only
- Zonal
- AHMOS-I (Hawk centre
fuselage test)
- Patria table tests
- Hawk under wing pod
flights
- F-18 wingtip pod flights
- Risø ground test
- Proven technology
- COTS support from
multiple sources
(component providers)
Limited detection range
(structural “hot spots” and
failure mechanisms must
be known prior to
installation)
See above
See above
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
- Implementation of
damage location and
sizing algorithm
- Upgrading of
software speed
page 58 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
AHMOS
partner
Emmecon
SHM
System
System
network and
modules
Detection/research
objective
SGI
NLR
TRL
Substantiation of TRL
Advantages
Limitations
Future needs
7
- WIWEB, Risø and
QinetiQ lab tests
- Hawk and F18 pod-tests
Quality and mechanical
layout of the commercial
Ethernet switch is not good
enough.
Electromechanical
layout in most of the
equipment
7
- WIWEB, Risø, QinetiQ
and Patria lab tests
- Hawk and F18 pod-tests
- Environmental tests
(EMC, temp, pressure,
vibration, shock)
- Risø ground tests
- Hawk flight tests
- Standard HW and
communication interface
(TCP/IP). Expansion of the
network is PnP
- Modules are “stand alone”
and thus not continuously
network dependent
- Rugged design
- Stand alone operation with
data storage
- Suitable for loads
monitoring
- Multichannel (8)
- On-board data storage
- Web based real-time
monitoring
- Flexible sensor for radii
Some obsolete components
used
Electronics and
enclosure redesign
- Use of COTS equipment
- Memory less
sensitive to vibration
- Limited detection range
(sensor size)
- Temperature sensitivity
- Limited depth of
penetration (about 3 mm)
- Temperature sensitivity
- Temperature
compensation
- Time monitoring
capabilities of
MultiScan (> 1 day)
Central
computer
Data storage of SHM
sensor systems
7
EC sensor
Surface crack in metal
structures
4
System only evaluated in
laboratory environment
Sub-surface crack
around fasteners
4
See above
- Sensitivity of crack
detection
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 59 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
4.3
Conclusions
An overview of the Technology Readiness Level (TRL) of the different sensor techniques and sub-systems
evaluated in the AHMOS II project is given. The TRL varies from value 3 to value 7. TRL 7 was reached for the
AE (crack detection in metal structure), corrosion sensor, FBG (strain measurement), SGI and SWISS sensor
systems. Also, the system network and modules, GUI and central computer were rated with a TRL 7.
The TRL values are considerably higher than those obtained in the predecessor project AHMOS I. In that
project the maximum TRL value was only 5, reached for the surface mounted FBG and the SWISS system.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 60 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
5
BENCHMARKING TOOL AND DECISION SUPPORT TOOL
A Java based tool has been developed and implemented in the AHMOS II website, see the figure below.
By selecting the material and the damage type that is to be detected, the tool gives an advice on suitable
detection technologies.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 61 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
6
IDENTIFICATION OF NEXT STEPS FOR EXPLOITATION AND IMPROVEMENTS
6.1
Introduction
In this chapter the next steps will be defined to facilitate the introduction of a SHM System on an aircraft, using
the before mentioned results (achievements and failures) of the AHMOS II work. The following areas will be
discussed in detail: sensor systems, system network and modules, damage presentation (graphical user
interface) and qualification of the systems.
6.2
Sensor Systems
In AHMOS II, the following sensors have been tested during ground testing and or flight-testing: Acoustic
Emission, SWISS, Strain Gauge, Fibre Optic Bragg Grating (FOBG), and Lamb Wave (LW).
Furthermore, theoretically or in laboratory testing the following sensors have been analyzed: eddy current array
sensor, long period gratings (LPG’s), corrosion sensor, and fibre optic sensor using a reflectometric approach.
6.2.1
Acoustic Emission (AE)
The AE detection technology is real-time. It is designed to operate while the platform is in service undergoing
operational loads. Sensors are surface mounted. AE sensors have been used in ground testing as well as in the
flight test with the Hawk.
Damage detection capability of the AE sensors has again been proven. The capability of damage localization
and extent, which is potentially one of the major benefits of AE sensors, is however not yet proven. Due to the
short period of testing, aspects such as the reliability, robustness could not sufficiently be challenged. Although,
the AE sensors claim to be global sensors, meaning that they are capable of detecting damages in large areas,
this has not been tested in this project. Initially, further developments for the in service use of this sensor should
focus on “hot spot’ monitoring.
The claimed Technology Readiness Level for this sensor is 7. Considering the achievements in AHMOS II this
claim can be acknowledged. Nevertheless, many more tests in representative environments are necessary to
bring this system to its final form for introduction. This can be achieved by extensive development and testing of
complete systems in accurate operational conditions (ground and flight tests).
Further work is needed in the following areas:
•
Proving 100% reliability of damage detection in the monitored area. This requires extensive testing
and validation.
•
Further development of damage localization capability.
•
Further development of damage extent determination capability.
•
Ground testing to prove robustness of the system.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 62 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
•
Further development of software to provide Level 3 damage presentation data to the Graphical User
Interface.
6.2.2
SWISS
SWISS (Smart Wide area Imaging Sensor System) is based on ultrasound. In principle, it has the potential to
detect all such damages (e.g. corrosion, cracks, delaminations etc.) that can be detected by ultrasound in
conventional inspections. The self-test capability of sensor, sensor-coupling to metallic specimen after
instrumentation and under in-service like conditions has been demonstrated. Also the known-good ultrasound
response of known-good (healthy) structure has shown sufficient repeatability to serve as ultimate self-test of
the system functioning. Damage has been detected when damage occurred. Localized damage (cracks) in the
structure has been detected, localized and quantified. Depending on the specific installation and
implementation, SWISS has demonstrated applications specific potential to deal with unexpected (non-lab-like)
behavior of hot-spot areas. In summary, SWISS has been used on the ground and the flight tests with
application specific success.
The sensors are surface mounted. The SWISS system is not a real time system, which means that the system
is not active during the operation of the aircraft. Interrogation of the structure for damages is done on the ground
only.
The claimed Technology Readiness Level for this sensor is 7 (“flight proven”). The same level of TRL does not
apply to the prototype off-board equipment, since the technological lower criticality for off-board, off-line
operated equipment demanded less allocation of resources within this project to the off-board hardware in favor
of proving the robustness of on-board sensor instrumentation. Considering the achievements in AHMOS II an
overall TRL 7 (including off-board hardware) could not be achieved. Nevertheless, many more tests in
representative environments are necessary to bring this system to its final form for introduction. This can be
achieved by extensive development and testing of complete systems in accurately defined monitoring tasks
under operational conditions (ground and flight tests).
Further work is needed in the following areas:
•
Proving 100% reliability of damage detection in the monitored area. This requires extensive testing
and validation.
•
Further development of automatic damage localization and interpretation capability.
•
Further development of damage extent determination capability (more channels, but same weight &
cost).
•
Ground testing to prove robustness of the system (including off-board/on-board hardware).
•
Further development of software to provide Level 3 damage presentation data to the Graphical User
Interface.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 63 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
6.2.3
Strain Gauges
The Strain Gauge technique can be used for damage detection in metallic aircraft structure. Strain Gauge
sensors are surface mounted and are only suitable for monitoring known ‘hot-spots’. Because of the limited
coverage area of the strain gauges, the positioning of the sensors is extremely important. This requires a good
knowledge of the structure to be monitored and its failure mechanism.
Although the Strain Gauge technique and system are on a quite high technology readiness level (TRL 7) still
much more work is to be done with testing on the ground and in flight, in accurate operational conditions, in
order to quantify the reliability of the system, probability of damage detection, etc.
The claimed Technology Readiness Level for this sensor is 7. Considering the achievements in AHMOS II this
claim can be acknowledged. Nevertheless, many more tests in representative environments are necessary to
bring this system to its final form for introduction. This can be achieved by extensive development and testing of
complete systems in accurate operational conditions (ground and flight tests).
Further work is needed in the following areas:
•
Proving 100% reliability of damage detection in the monitored area. This requires extensive testing
and validation.
•
Further development of damage extent determination capability.
•
Further development of software to provide Level 3 damage presentation data to the Graphical User
Interface.
•
6.2.4
Wireless technology.
Fibre Optic Bragg Grating (FOBG)
The Structural Strain Variation monitoring technique by means of FOBG is real time. It is designed to operate
with the platform in operation. FOBG sensors have the potential to detect delamination and debonding in
composite material, cracks in metal structure, to do strain measurements and to do fatigue life consumption
calculations. The FOBG sensors can be surface mounted as well as embedded in composite structures.
FOBG sensors have been tested on the ground tests at RISØ, on the Hawk flight-testing, on ground and flighttesting in Spain (EF-2000, C101) and Italy (C27).
The claimed Technology Readiness Level for this sensor varies from 4 to 7. Considering the achievements in
AHMOS II these claims can be acknowledged. Many more tests in representative environments are necessary
to bring this system to higher TRL’s. This can be achieved by extensive development and testing of complete
systems in accurate operational conditions (ground and flight tests).
Further work is needed in the following areas:
•
Proving 100% reliability of damage detection in the monitored area. This requires extensive testing
and validation.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 64 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
•
Ground testing to prove robustness of the system.
•
Further development of software to provide Level 3 damage presentation data to the Graphical User
Interface.
•
Improve capability to monitor larger areas.
•
Reduction in size and weight
•
Improve bend radius of optical fiber or investigate the possibility of bend tolerant fiber
•
Active cooling of laser
•
Further improvement of optoelectronics and algorithms to reduce/avoid coupling effects.
•
Temperature compensation.
•
Coating materials
6.2.5
Lamb Wave
The capabilities of the Ultrasonic Guided Wave (Lamb Wave) system are limited to monitoring of the structural
integrity of a region of structure local to the transducer. Therefore, this system is useful only to monitor ‘hotspots’ for debonding of joints, cracks and possibly corrosion. The system requires to be specifically configured
for each monitoring region with respect to both the defect type and structural acoustic properties. UGW sensors
are surface mounted. This system has been used on the Risø ground testing and on the Hawk flight-testing.
Test results show that the system has detected the debond, in real time, during flight-testing, without any false
calls.
The claimed Technology Readiness Level for this sensor is 6. Considering the achievements in AHMOS II this
claim can be acknowledged. Nevertheless, many more tests in representative environments are necessary to
bring this system to its final form for introduction. This can be achieved by extensive development and testing of
complete systems in accurate operational conditions (ground and flight tests).
Further work is needed in the following areas:
•
Improvement of the robustness of the system.
•
Further development of software to provide Level 3 damage presentation data to the Graphical User
Interface.
•
6.2.6
Improvement of the temperature compensation
Eddy current array sensor
The investigation showed that the high-frequency absolute sensor is suited for the detection of surface cracks
located just beneath the sensor. This sensor is flexible and can be bent down to small radii; the sensor
produces reliable EC signals for radii down to 7 mm. The low-frequency reflection ring sensor can be used for
subsurface crack detection in riveted joint structures. The detectable crack length for this sensor depends on the
inspection configuration, the depth of the defects and the test frequency. As a rough indication, defects from
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 65 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
about 8 mm at a depth up to 3 mm in lap joints of Glare 4B-0.4 material are detectable. Smaller defects (length
< 6 mm) and defects at larger depth (depth > 4 mm) are not detectable anymore.
Both prototype sensors are sensitive to changes in ambient temperature. This should be considered (and
compensated for) when performing SHM measurements at different temperature, for example during in-flight
and on-ground conditions.
The feasibility and potential applicability of both EC sensor types for the local detection of cracks in metal
aircraft structures has been demonstrated. However, much further work is needed. For example, it is
recommended to increase the monitoring capabilities of the MultiScan equipment in order to use the system for
practical SHM applications. The system should allow continuous operation for periods much longer than one
day (main areas of attention are the acquisition rate and the cursor length in the strip chart display of the
instrument). It is also recommended to evaluate the multi-sensor capability of the MultiScan equipment using a
specified 16-channel switch box (in the present investigation only single sensor measurements were
performed).
It is recommended to further investigate both the high-frequency absolute sensor and the low-frequency
reflection ring sensor using different and realistic specimen configurations under fatigue testing. The actual
limits of flexibility of the high-frequency absolute sensor should be investigated using specimens with smaller
radii (< 7 mm). The low-frequency ring sensor should be further investigated for the detection of sub-surface
corrosion in aluminum alloy structures.
TRL for this technique is correctly set at 4. Therefore, extensive further research and development is needed to
become sufficiently mature to be used on aircraft.
6.2.7
Long period gratings (LPG’s)
The investigation concludes that low cost, commercially available equipment (including a spectrometer with a
wavelength resolution of 0.1 nm) based on HRI coated LPG’s, can provide advanced chemical sensing for
VOC’s, gases, and liquid monitoring in light of the easy multiplexing capability and lower complexity compared
to alternatives such as Plasmon resonance or waveguide sensors.
This basic research has shown the potential of LPG’s for SHMS purposes. The claimed TRL 6 could not be
acknowledged. TRL 4 is more appropriate. Therefore, extensive further research and development is needed to
prove the usability for SHMS purposes and to become sufficiently mature to be used on aircraft.
6.2.8
Corrosion sensor
This technique is based on the principle that the surrounding structure corrodes at the same rate as the sensor.
The Corrosion sensor has been used on the Hawk flight test without any result due to not occurring of corrosion
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 66 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
damage. Advanced lab trials at BAE Systems, however, showed the corrosion detection capability of this
sensor. On other military aircraft outside AHMOS, trials are performed with prototype systems. Based on the
results of these trials BAE claims TRL 7 for this sensor. This could not be acknowledged because this
information has not been provided to the AHMOS II community. Therefore, extensive further research and
development is needed to become sufficiently mature to be used on aircraft. One of the issues to tackle is inservice data validation.
6.2.9
Fiber optic sensor using reflectometric approach
This work involved testing COTS components and telecommunications as fibres and optical devices to verify the
“reflectometer” concept for the detection of the presence of a liquid. A “reflectometer” is a fibre optic sensor
based on measuring the reflective index difference between a cut glass fibre and the surrounding environment.
The reflection coefficient inside the glass dramatically decreases when the open fibre end is contaminated by
liquid.
This is a preliminary study based on defining low cost and robust components, and the range of potential
measurements likely. The work also includes some practical investigations on the influence of the fibre optic cut
quality on the performance of an eventual sensor. Plastic deformation in the core zone may induce case-bycase variability in the reflected light intensity, leading to a lack of absolute measures for the sensor output.
This basic research has shown the potential of ‘reflectometric approach’ for SHMS purposes.
The claimed TRL 6 could not be acknowledged. TRL 4 is more appropriate. Therefore, extensive further
research and development is needed to prove the usability for SHMS purposes and to become sufficiently
mature to be used on aircraft.
6.3
System network and modules
All parts and modules of the system network that have been used and tested during the AHMOS II are whether
experimental or prototype parts. During the ground and flight testing it has been proven that they fulfill the
functional and performance requirements as needed for the AHMOS II project but that they are not yet fit for
definitive in service installation into aircraft. In order to achieve this, these parts should be further developed
(including miniaturization) and further tested on ground and in flight for qualification.
Further development areas:
•
Data reduction capability (on CC store only data that is relevant)
•
Standarisation of interfaces (both hardware and software)
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 67 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
6.4
Data presentation
In this project a Graphical User Interface (GUI) by means of a Notebook System has been developed by
DEMEX (NIRAS) in order to present damage data and more to the maintenance personnel. The GUI has been
used during the ground tests, the flight tests of the F-18 and the Hawk aircraft.
In three levels information is given about the health status of system and structure to the user: level 1 gives an
indication about the status of the system (by means of lights), level 2 gives an indication about the health status
of the structure (whether or not damage has been detected) and level 3 gives a graphical presentation of the
damage (type, location and dimension). In level 3, also the detailed results of each individual sensing technique,
graphs, images etc. can be obtained.
During ground testing at Risø and flight-testing with the Hawk and F-18 aircraft it has been proven that the level
1 and level 2 functionality of the GUI works as expected. Level 3 information as far as available has not
completely been used.
Following developments are required before the GUI is suitable for in service use:
•
Further improvement of the Level 3 functionality by providing 3D pictures of the monitored structure on
which the detected damages are shown. In first instance, for one known structural part with damages
the 3D presentation should be further developed, tested and validated.
•
Prove 100% reliability of the GUI software.
•
Improve the Robustness of the Notebook.
•
Enhance download speed of data from the Central Computer.
•
Data reduction by storing only relevant data.
6.5
Qualification
For qualification of the SHM Systems both the software and the hardware of the Systems should be qualification
tested.
6.5.1
Software qualification
The verification objectives for DO-178B are set in place to detect and report errors that may have been
introduced during development processes. Software verification objectives are satisfied through a combination
of reviews and analyses, the development of test cases and procedures, and the execution of those test
procedures. Reviews and analyses provide an assessment of the accuracy, completeness, and verifiability of
the software requirements, software architecture, and source code.
As part of the software qualification process, the system safety assessment is used to determine and categorize
the failure conditions and the criticality of a system.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 68 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
In AHMOS II, Software Critically Level E (Software that could cause or contribute to the failure of the system
resulting in no effect on the system) has been selected, because the SHM Systems are introduced for testing
purposes only and the failure of the system will not influence the operation of the aircraft.
However, at least the Software Criticality Level C (Software that could cause or contribute to the failure of the
system resulting in a major failure condition) should be considered at the commercial introduction of a SHMS in
(military) aircraft. This means that Statement Coverage applies for the Structural Coverage Analyses Objective.
Statement Coverage means that every statement in the program has been invoked at least once.
6.5.2
Hardware qualification
During AHMOS II extensive amount of hardware qualification testing has been performed.
All the system parts that have been installed in the Hawk pod have successfully been tested on the ground and
in flight either by suppliers or by BAE Systems.
All the system parts that have been installed in the F-18 pod have also been ground and flight-tested. Except for
an EMI problem (interference on the cockpit audio system) on 3 out of 4 aircraft used, the system passed all
tests. The EMI problem could not be resolved in the timeframe of this project because of the slow progress of
the flight-testing.
All test results are documented in AHMOS II WE reports.
Further development areas:
•
Qualification testing of the actual system to be installed.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 69 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
7
REFERENCES
[1]
ERG103.015-DOC-1100-0001-NLR-Draft, Improvement of sensor and network technologies
[2]
ERG103.015-DOC-1120-0001-NLR-Issue 1, Development of new technologies
[3]
ERG103.015-DOC-1121-2000-CIRA-New Sensors-Draft
[4]
ERG103.015-DOC-1121-2101-CIRA-Issue 1, Long period gratings and reflectometric approach for
chemical sensing
[5]
ERG103.015-DOC-1121-2201-BAE Systems-Draft, Develop corrosion sensor
[6]
ERG103.015-DOC-1121-2301-NLR-Issue 2, Development of an eddy current array sensor
[7]
ERG103.105-DOC-1121-3201-Emmecon-Improve Function Perf of SGI-Issue1
[8]
ERG103.015-DOC-1121-3301-Issue1-SF, Report on the Miniaturisation of the Electro-Optic Interface
[9]
ERG103.015-DOC-1122-0001-Emmecon-Issue1, System network and modules
[10] ERG103.015-DOC-1124-1001-Fokker-Review-validate-and –update-existing-data-presentation-issue1-doc
[11] ERG103.015-DOC-1200-0001-DEMEX-Ground-Tests-Report-Issue1
[12] ERG103.015-DOC-1290-1001-EADSCASA-Technical-Report1-FBGS-Ground-Test-Specification-Issue1
[13] ERG103.015-DOC-1290-2001-EADSCASA-Technical-Report2-FBGS-Static-Data-Capture-Issue1
[14] ERG103.015-DOC-1290-3001-EADSCASA-Technical-Report3-FBGS-Dynamic-Data-Capture-Issue1
[15] ERG103.015-DOC-12A0-1001-EADSCASA-Technical-Report4-Check-of-Structural-Integrity-Using-Issue1
[16] ERG103.015-DOC-1250-0001-INTA-Embedding-trials-for-FBGS-Issue1
[17] EUC103.015-DOC-1260-0001-EADS-G-Trial-for-realistic-instrumentation-Issue1
[18] ERG103.015-DOC-1310-0001-EADSCASA-Technical-Report1-FBGS-Flight-Test-Specification-Issue1
[19] ERG103.015-DOC-1320-0001-EADSCASA-TechnicalReport2-FBGS-Design-Installation-Issue1
[20] ERG103.015-TRP-2000-0001-BAESystems-Draft
[21] ERG103.015-DOC-2410-0000-RISOE-Definition-of-Structural-Tests-Issue1
[22] ERG103.015-DOC-2440-0000-RISOE-Ground Test Demonstration (Evaluation)
[23] ERG103.015-TRP-2500-0001-BAESystems-Issue1
[24] ERG103.015-DOC-2600-0001-Patria-F-18-AHMOS-pod-flight-tests-Draft 1
[25] ERG103.015-DOC-3100-0001-NLR-Issue 1, Functional requirements and technology readiness
[26] ERG103.015-DOC-2110-0001-Fokker-Qualification Program Plan-Issue 2
[27] ERG103.015-DOC-2300-0001-Fokker-Verification of Compliance with applicable QPP items-Issue 2
[28] ERG103.015-DOC-2400-0001-DEMEX-User-Manual-for-the-SHMSNS-used-in-the-RISOE-Tests-Issue1
[29] ERG103.015-DOC-2200-0001-VTT-Issue1, Development of an Operational Structural Health Monitoring
System Prototype
[30] ERG103.015-DOC-1230-0001-DEMEX-Marine-Trial-Issue1
[31] ERG103.015-DOC-1230-0001-ANAKRON-Test-of-passive-Annakron-AE-sensors-Issue1
[32] ERG103.015-DOC-2500-0001-DEMEX-User-Manual-for-the-SHMSNS-used-in-the-HAWK-Tests-Issue1
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 70 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
[33] ERG103.105-DOC-2600-0001-DEMEX-User-Manual-for-the-SHMSNS-used-in-the-F-18-Tests-Issue1
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 71 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
8
ABBREVIATIONS AND ACRONYMS
AE
Acoustic Emission
CAN
Controller Area Network
CC
Central Computer
COTS
Commercial Off The Shelf
EC
Eddy Current
EMC
Electro Magnetic Compatibility
EMI
Electro Magnetic Interference
FBG
Fibre Bragg Grating
FBGS
Fibre Bragg Grating System
FO
Fibre Optic
FOBG
Fibre Optic Bragg Grating
FTP
File Transfer Protocol
GUI
Graphical User Interface
HRI
High Refractive Index
LPG
Long Period Grating
LW
Lamb Wave
NDT
Non Destructive Testing
NS
Notebook System
OEM
Original Equipment Manufacturer
PCB
Printed Circuit Board
POD
Probability Of Detection
SGI
Strain Gauge Indicator
SGSS
Strain Gauge Sensor System
SHM
Structural Health Monitoring
SHMS
Structural Health Monitoring System
SS
Sensor System
SWISS
Smart Wide area Imaging Sensor System
TRL
Technology Readiness Level
VOC
Volatile Organic Compound
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 72 of 73
temp. 0572 12-03 00.01
Fokker Services B.V.
9
APPENDICES
APPENDIX 1
Wing Full Scale test (WE1210).
Report is not yet available.
APPENDIX 2
Flight test trials CASA (WE1.3).
Reported in ERG103.015-DOC-1340-0002-EADSCASA-Technical Report-Issue1:
EADS-CASA participates on the European project ERG103.015 “Prototype Demonstration of Structural Health
Monitoring System for Military Platforms (AHMOS II)” performing activities related to the following WEs, WE 1.2
“Fatigue Test FBGS” and WE 1.3 “Flight Test Trial”. Also supports Airbus on some activities related to WE 1.1
“Improvement of Sensors and Network Technologies”.
EADS-CASA participates in ERG103.015 developing the FBGS technology based in Bragg´s law. WE 1.2 work
was devoted to monitor a ground component with a FBGS system. WE 1.3 work goes further being its scope to
verify the ability of the FBGS system to measure in flight conditions.
Results obtained during the flight test campaign performed by EADS-CASA with a C-101 aircraft showed that it
is possible to use optical fiber with integrated FBGS, as strain/stress sensors, to develop a Structural Health
Monitoring System for Military Platforms.
A good correlation between the stress spectra, generated for the FBGS data recorded, and the vertical
acceleration spectrum, obtained with the inertial data device, has been obtained.
The results given by the recording system design, the flight test campaign and the subsequent analysis is
considered as a success.
APPENDIX 3
Flight test trials Alenia (WE1.4).
Report is not yet available.
issue date
issue no
19 Dec. 07
1
security class
report no.
UNCLASSIFIED
ERG103.015-DOC-3000-0001-Fokker
page 73 of 73
temp. 0572 12-03 00.01
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement