See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/324051082

Damage detection with ultrasonic guided wave under operational conditions

(Conference Presentation)

Conference Paper · March 2018

DOI: 10.1117/12.2318696

CITATIONS READS

0 117

3 authors:

Zahra Sharif Khodaei Mohammad Saleh Salmanpour

Imperial College London Imperial College London
81 PUBLICATIONS   562 CITATIONS    9 PUBLICATIONS   81 CITATIONS   

SEE PROFILE SEE PROFILE

M.H. Aliabadi
Imperial College London
594 PUBLICATIONS   8,346 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Reliability Based Analysis with the Boundary Element Method View project

Wi-PaSS: Wireless Passive Sensing System for SHM of Composite Structures View project

All content following this page was uploaded by Zahra Sharif Khodaei on 03 August 2018.

The user has requested enhancement of the downloaded file.

9th European Workshop on Structural Health Monitoring
July 10-13, 2018, Manchester, United Kingdom

Damage detection with ultrasonic guided wave under

operational conditions
M. Salmanpour 1, Z. Sharif Khodaei1,a, M.H. Aliabadi 1,b
1
Department of Aeronautics, Imperial College London, UK
a
z.sharif-khodaei@imperial.ac.uk; bm.h.aliabadi@imperial.ac.uk

Abstract

Typical airliners operate in a range of conditions, hence airborne structural health

monitoring (SHM) components, must withstand the relevant environmental conditions.
Additional to the integrity of the components, the SHM performance (diagnosis and
prognosis) must be robust and reliable under environmental and vibration profiles during
operation. This work investigates the influence of the operational condition (including
temperature, humidity and vibration loads) on the integrity of a piezoelectric based SHM
system in terms of integrity of the system and robustness of the diagnosis for detecting
barely visible impact damage (BVID) on a CFRP panel. Consequently, compensation
techniques are proposed to remove the effect of the environmental loading on the decision-
making algorithm. The validity of the proposed algorithm is demonstrated on a composite
plate for the operational profile defined in MIL-STD 180G standard for airborne
components of a regional aircraft.

1. Introduction
Structural health monitoring (SHM) approach offers a paradigm shift from a schedule
driven maintenance regime to a condition based maintenance philosophy. This family of
approaches have attracted both industrial and academic attention [1, 2]. Using a network
of sensors and/or actuators, SHM systems continuously monitor structural integrity while
the structure is in operation. By monitoring the response of the structure and appropriate
post-processing techniques it can result in impact detection and characterization [3-7] as
well as damage detection and localization [8-11]. Sensorized structures, with embedded or
surface mounted sensors, record the response of the structure during operation and result
in on-line damage detection. Therefore, the SHM has the potential to reduce the downtime
for scheduled inspections and maintenance and offer a condition based maintenance
strategy for lower cost and higher reliability.
The decision to have any SHM system permanently installed on the structure as a
maintenance strategy is directly influenced by the additional weight of the system as well
as its reliability and probability of detection. The additional weight of the onboard
monitoring system can be addressed by optimizing the number and location of the required
sensors [12-15]. The reliability of the SHM system must be tested and assessed under
operational condition of the structure [16]. This is significant, as most of the established
damage assessment techniques require comparison between two system states. However,
the structural response will vary under different environments such as changes in
temperature, load level and humidity. These changes in the structural performance, within
the designed operating range clearly are not a concern, yet these changes can have
significant effects on the SHM system performance [17]. There are many damage detection
techniques based on ultrasonic guided waves, which have been developed and validated in
laboratory conditions. However, for certifying any SHM technique as a maintenance

Creative Commons CC-BY-NC licence https://creativecommons.org/licenses/by-nc/4.0/

strategy, its performance must be tested under operational conditions. The regulatory
bodies (e.g. FAA) have included the use of SHM only as a scheduled maintenance
technique so far, to be used in parallel with existing detection techniques, set at fixed
schedule intervals (ATA-MSG3-rev 2009). This approach is to demonstrate the capability,
reliability and durability of such techniques in an in-service condition.
In this paper, the effect of temperature, humidity and vibration on SHM based damage
detection methodology is investigated. The integrity of the onboard SHM system is tested
under operational condition, whereas the reliability and probability of detection of the
diagnostic system is assessed for on ground conditions.

2. Overview of the SHM system

An effective condition based monitoring system should integrate the SHM
sensors/components into the aircraft without any appreciable adverse effects on the
performance of the host structures or other aircraft systems. The most significant aspects
of the system architecture are the physical and functional attributes:
 Physical architecture: sensors, transducers, wires, data acquisition systems,
amplifiers, other components and subsystems as well as their interconnections.
 Functional architecture: performance requirements (i.e. damage detection
(BIVID), reliability, etc.)
Figure 1 shows the main physical entities of the proposed SHM architecture. Transducers
(Sensors and actuators) and their connections are physical entities that must be built into
each individual SHM enabled part.

Figure 1. The main physical components of the proposed SHM architecture

2.1 Damage detection algorithm

The idea of smart structure is to have SHM systems on board to diagnose its health status
through permanently installed sensors and only trigger for maintenance action only when
a defect is detected. Once the fault is detected and characterized, the remaining useful life
of the component will be predicted using the recorded data. The proposed damage detection
algorithm is based on ultrasonic guided wave with delay and sum algorithm [10, 18, 19]
which is based on baseline subtraction techniques to detect and localize damage based on
the damage reflected wave, i.e. the residual signal. In an ideal situation, the residual signal
only contains damage-reflected features. However, the operational and environmental load
conditions can contribute to the residual signal and result in missed-detection or false
alarm. Temperature fluctuations directly influence the propagation velocities of guided
waves in terms of change in amplitude and/or phased shift.

2
 (a) Aluminum (b) CFRP plate
Figure 2 Delay times for 300 kHz actuation
The methodology presented here to correct the temperature effects on the recorded signals
is the multiple baseline signal stretch (MBSS) [20]. In previous correction methods [21,
22], usually only one correction factor has been proposed (e.g. BSS method). However,
each wave mode is influenced differently by temperature change. In addition, the time
delay is cumulative in time. That is it increases with propagation distance and increases
with time. Although this change is linear for isotropic structures, it is not the case for
composite plates, even if they are quasi-isotropic as shown in Figure 2. Hence, a range of
stretch factors is applied iteratively in order to reduce the residual level from temperature
change over the entire signal domain. The reliability of the damage detection algorithm
with temperature correction is tested with experimental tests in the next section.

3. Experimental set-up
The environmental test profiles that are selected represent the on-ground operational
conditions of a regional aircraft operating in Europe with maximum operating altitude of
50,000 ft. The ground-based test profiles include low temperature, high temperature,
humidity and vibration. The integrity tests are divided into two categories:
 A: Integrity tests of Airborne components and reliability assessment of ground-
based components.
 B: Reliability assessment of Ground-based components.
Tests in category A are to ensure that the on-board equipment can operate under the
environmental condition of the host structure reliably during the operational life of the
structure [16]. Tests in category B are carried out to assess the integrity of the data
acquisition system as well as the reliability of the decision making algorithm, i.e. damage
detection methodology under on-ground operational conditions. This work reports on the
later, i.e. the reliability of the damage detection algorithm under the operational load of the
aircraft as part of ground maintenance action.
3.1 Reliability assessment of damage detection methodology under operational
condition
To investigate the reliability of the diagnosis two composite plates were manufactured to
be tested for damage detection with guided waves. DuraAct transducers (PZT discs 10x0.2
mm) and an in-house manufactured SHM layer by ICL (PZT discs 10x0.25 mm) were

3
permanently attached to two separate CFRP plates of size 225x300 mm. The plates were
manufactured using 16 unidirectional Hexply 914-TS-5-134 plies with stacking sequence
[0,45,-45,90]2s of 2 mm overall thickness, see Figure 3.

Figure 3 CFRP plates with (a) ICL’s SHM layer; (b) DuraAct transducers
The plates were impacted using a drop tower with a 10mm radius hemispherical impactor
head. After the impact C-scan results of the panels were taken to confirm the initation of
barely visible impact damage (BVID). Signals under environmental loading (including
temperature, humidity and vibration) was recorded in an environmental chamber for
pristine and impacted plates to assess the damage detection methodology. The assessment
of the sensor integrity have been verified earlier [16].

3. Results and Discussion

3.1 Influence of temperature and humidity

Since temperature influences the stiffness of the plate, it affects the propagation velocity
of guided waves, which influences the time of arrival (ToA) of the wave, see Figure 4. This
phase shift can be mistaken as a damage residual if not corrected. The damage detection
algorithm without and with MBSS temperature correction were applied to detect and
localize BVID. The results of the diagnosis with temperature correction is shown in Figure
5 for 50 kHz excitation. The baseline was measured at 25C while the post-impact signals
were recorded from -50C to 60C. Temperature compensation was also shown to be
effective for higher frequencies such as 300 kHz excitation. In this case the error for
localization was 2.5 cm for the temperature range of -5 to 35°C. When the temperature
difference increased to outside of this range, the error also increased. The increased error
can also be attributed to the reduced sensitivity of S0 mode for detecting through thickness
BVID. The relative amplitude of the residual at 300 kHz was at least 50% lower compared
to the A0 dominate excitation at 50 kHz.
There were no significant influence due to changes in humidity on the propagation
velocities of guided waves [23]. One reason could be that the duration of the humidity loads
applied (in combination with temperature profile) was not long enough to notice any
adverse effects. Noticeable humidity saturation is reached in time scale in the order of
months and years.

4
 Figure 4 Effect of temperature on Lamb wave propagation in CFRP panel at 300 kHz
∆T=0 oC ∆T=-25 oC ∆T=-50 oC

∆T=10 oC ∆T=20 oC ∆T=35 oC

Figure 5 Impact damage detected on a CFRP panel at 50 kHz with MBSS correction
3.1 Influence of vibration
The influence of random vibration on the recorded guided wave signals is shown in Figure
6. Since the damage detection methodology is based on baseline comparison, for a reliable
diagnosis the effect of the noise has to be successfully removed from the recorded signal.
As a first attempt, a zero-phase Butterworth filter with high-pass cut off frequency above 2
kHz was used to remove vibration noise. However, as shown in Figure 6 the random low
frequency noise can have higher amplitude than the diagnostic signals which results in the
signal to be clipped or even saturated if a the recording voltage range will be close to
recorded signal amplitude.

5
 Figure 6 Signal recorded under random vibration highlighting signal clipping at 300 kHz
To avoid the signal clipping a hardware high-pass filter was designed and used. Damage
localization was then carried out with the filtered signals that were recorded during
vibration to detect BVID. Both 50 kHz and 300 kHz signals were used as depicted in Figure
7. At 50 KHz, localization was only possible using transducers that did not cross the
vibration clamp (in the middle of the plate). It may appear that the effect of vibration on
signals was the main reason for poor performance at 50 kHz. However, boundary
conditions imposed by the vibration fixture had a significant impact on the wave
propagation since the baseline measures were taken outside of the fixture. The effect of the
boundary conditions on the recorded signal inside and outside of the fixture is shown in
Figure 8 where it can be seen that the residuals due to the boundary condition at 300 kHz
excitation (Figure 8 (b)) are an order of magnitude lower than those for 50 kHz (Figure 8
(a)). To avoid the influence of change in the boundary conditions, only paths which did not
cross the fixture clamp were used to localize damage, Figure 8 (c). This shows the influence
of changes in the boundary condition when baseline detection methods are used and that
the baseline signals should be recorded under the same boundary conditions as the plate
under vibration during the experiment.

(a) 50 kHz detection (b) 300 kHz detection (c) 50 kHz detection only
3 transducers
Figure 7 Damage localization for BVID with baseline signals recorded under vibration

6
 (a) 50 kHz excitation (b) 300 kHz excitation
Figure 8 Residual signal due to fixture clamps for vibration setup

4. Conclusions
This paper presented the integrity and reliability assessment of an on-board SHM
components under operational load conditions. The methodology proposed in this work for
structural diagnosis is a baseline comparison method based on ultrasonic guided waves.
Aircraft structures undergo various operational load such as temperature, humidity and
vibration. These factors change the propagation properties of guided wave. Therefore, if
the current state of the structure where the response is recorded varies from the baseline
state, these changes can contribute to missed-detection or false alarm. The effect of
temperature change was shown to have the highest influence on phase shift of recorded
signal. This phase shift is not linear for composite plates in time. Therefore a single
correction factor is not reliable to correct for the temperature effects. The proposed MBSS
technique was shown to successfully detect and localize damage for a wide range of
operating temperatures from -50 to 60 C when the baseline measure was recorded at 25
C. The humidity did not have any significant effect when the loading time was similar to
temperature load. For the humidity to have significant effect on the structure it must be
loaded for months and years which was out of scope of this work. In addition, the signals
were recorded under vibration profile which represented the random vibration profile for a
regional aircraft. Appropriate filters were designed to remove the effect of vibration and
damage localization was then carried out with filtered signals. Due to the different
boundary conditions of the plate when the vibration load was applied, there were residual
which was due to reflection from clamps and decreased the accuracy of the localization.
This shows the importance of boundary conditions when baseline comparison detection
algorithms are applied. However, when using paths which were not affected by the new
boundary conditions, the localization was carried out successfully. Therefore, it was shown
that the proposed SHM methodology is capable of reliably detecting and localizing damage
under on-ground operational conditions.
References
1. Aliabadi, M.H. and Z.S. Khodaei, Structural Health Monitoring for Advanced
Composite Structures Computational and Experimental Methods in Structures, Vol.
Volume 8. 2018: World Scientific Publishing Europe Ltd. 288.
2. Mitra, M. and S. Gopalakrishnan, Guided wave based structural health monitoring: A
review. Smart Materials and Structures, 2016. 25(5): p. 053001.
3. Ghajari, M., et al., Identification of impact force for smart composite stiffened panels.
Smart Materials and Structures, 2013. 22(8): p. 085014.

7
 4. Sharif-Khodaei, Z., M. Ghajari, and M. Aliabadi, Determination of impact location on
composite stiffened panels. Smart Materials and Structures, 2012. 21(10): p. 105026.
5. Morse, L., et al., Reliability based impact localization in comp panels using Bayesian
updating and the Kalman filter. Mechanical Systems & Signal Processong 2018. 99: p.
107-128.
6. Yue, N. and Z.S. Khodaei, Assessment of Impact Detection Techniques for Aeronautical
Application: ANN vs. LSSVM. Journal of Multiscale Modelling, 2016: p. 1640005.
7. Seydel, R. and F.-K. Chang, Impact identification of stiffened composite panels: I.
System development. Smart materials and structures, 2001. 10(2): p. 354.
8. Salmanpour, M.S., Z.S. Khodaei, and M.H. Aliabadi, Instantaneous Baseline Damage
Localization Using Sensor Mapping. IEEE Sensors Journal, 2017. 17(2): p. 295-301.
9. Sharif Khodaei, Z. and M. Aliabadi, A Multi-Level Decision Fusion Strategy for
Condition Based Maintenance of Composite Structures. Materials, 2016. 9(9): p. 790.
10. Sharif-Khodaei, Z. and M. Aliabadi, Assessment of delay-and-sum algorithms for
damage detection in aluminium and composite plates. Smart materials and structures,
2014. 23(7): p. 075007.
11. Michaels, J.E. and T.E. Michaels, Guided wave signal processing and image fusion for
in situ damage localization in plates. Wave Motion, 2007. 44(6): p. 482-492.
12. Thiene, M., Z. Sharif Khodaei, and M. Aliabadi, Optimal sensor placement for
maximum area coverage (MAC) for damage localization in composite structures. Smart
materials and structures, 2016. 25(9): p. 095037.
13. Salmanpour, M., Z. Sharif Khodaei, and M. Aliabadi, Transducer placement
optimisation scheme for a delay and sum damage detection algorithm. Structural
Control and Health Monitoring, 2017. 24(4).
14. Mallardo, V., et al. Optimal sensor positioning for impact localization in smart
composite panels. J of intellig material systems and structures, 2012: p. 1045389.
15. Mallardo, V., Z. Sharif Khodaei, and F.M. Aliabadi, A Bayesian Approach for Sensor
Optimisation in Impact Identification. Materials, 2016. 9(11): p. 946.
16. Salmanpour, M.S., Z. Sharif Khodaei, and M.H. Aliabadi, Airborne Transducer
Integrity under Operational Environment for Structural Health Monitoring. Sensors,
2016. 16(12): p. 2110.
17. Salmanpour, M., Z. Sharif Khodaei, and M. Aliabadi, Guided wave temperature
correction methods in structural health monitoring. Journal of Intelligent Material
Systems and Structures, 2016: p. 1045389X16651155.
18. Konstantinidis, G., B. et al. The temperature stability of guided wave structural health
monitoring systems. Smart Materials and Structures, 2006. 15(4): p. 967.
19. Michaels, J.E., et al. Imaging algorithms for locating damage via in situ ultrasonic
sensors. in Sensors Applications Symposium, 2008. SAS 2008. IEEE. 2008. IET.
20. Salmanpour, M.S., et al. Impact damage localisation with piezoelectric sensors under
operational and environmental conditions. Sensors, 2017. 17(5): p. 1178.
21. Croxford, A.J., et al., Efficient temperature compensation strategies for guided wave
structural health monitoring. Ultrasonics, 2010. 50(4): p. 517-528.
22. Clarke, T. and P. Cawley, Enhancing the defect localization capability of a guided wave
SHM system applied to a complex structure. Structural Health Monitoring, 2010.
23. Salmanpour, M.S., Structural Health Monitoring of Composite Laminates Under
Operational Conditions using Piezoelectric Transducers in Aeronautics2017, Imperial
College London Imperial College Publishing

View publication stats