9th

European Workshop on Structural Health Monitoring

July 10-13, 2018, Manchester, United Kingdom

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

A Flexible-Integrated Impact Monitoring System for Aircraft

Composite Structures

Lei Qiu1, Shenfang Yuan

1, Xiaolei Deng

1, Yongan Huang

2 and Yuanqiang Ren

1

1 Research Center of Structural Health Monitoring and Prognosis, State Key Lab of

Mechanics and Control of Mechanical Structures, Nanjing University of Aeronautics

and Astronautics, P.R. China, lei.qiu@nuaa.edu.cn, ysf@nuaa.edu.cn,

dengxl@nuaa.edu.cn, renyuanqiang@nuaa.edu.cn

2 State Key Lab of Digital Manufacturing Equipment and Technology, School of

Mechanical Science and Engineering, Huazhong University of Science and Technology,

Wuhan 430074, China, yahuang@hust.edu.cn

Abstract

In recent years, large-scale composite structures have been widely applied on large

aircrafts. However, the poor impact resistance of composite structures may lead to

barely visible impact damage and result in stiffness degradation and a significant loss of

structural integrity. Therefore, impact monitoring of aircraft composite structures is an

important issue in the field of structural health monitoring. The impact of a structure is

an instantaneous event so that it needs to be monitored on-line continuously during the

whole service lifetime of the structure. Thus considering the strict restrictions of

aerospace application, an impact monitoring system is required to be low weight, low

power consumption and high reliability. In this paper, a Flexible-Integrated Impact

Monitoring System (FIIMS) with advantages of compactness, ultra-thin, light weight,

low power consumption and high efficiency is developed to meet the strict requests of

on-line application. Differently from conventional impact monitoring systems, the

complex analog circuits are removed and the whole impact monitoring process is

achieved in a digital way by turning the output of the PZT sensors directly into digital

sequences through a two-level mechanism which is realized by combing diode array

with a Field Programmable Gate Array (FPGA). A simple but efficient impact-region

localization method is implemented in the FPGA. In addition, the FIIMS is realized in a

flexible way so that it can be embedded onto the composite structure to be a

mechatronic system to realize impact monitoring on-line continuously. To illustrate the

capability of the FIIMS, it is integrated with a flexible PZT sensor array to construct a

large-scale impact monitoring network. The network is embedded onto the surface of a

large-scale carbon fiber reinforced plate with stiffeners by using a co-curing process and

the impact monitoring ability of the FIIMS is validated.

1. Introduction

Composite materials have been widely used in many areas of engineering applications

due to many advantages compared to conventional metallic materials, especially in

aerospace applications. Their excellent strength-to-weight ratios, resistance to

fatigue/corrosion and flexibility in design are particularly attractive for high

performance light weight aircraft structures. The latest Airbus A380, taken as an

example, is composed of 22% composites, and the Boeing B787 aircraft uses 50% of

Mor

e in

fo a

bout

this

art

icle

: ht

tp://

ww

w.n

dt.n

et/?

id=

2320

9

2

composite materials [1]. However, a major concern that has to be considered following

the widespread implementation of composites is their safety during service.

Impact usually causes inner damages in composite material which in general

cannot be found easily. This undetected, hidden damage is also known in aerospace

applications as Barely Visible Impact Damage (BVID), which can cause significant loss

of strength or stiffness. Because of this, impact monitoring has always been an

important monitoring object in the research area of Structural Health Monitoring (SHM)

for aircraft composite structures.

In recent twenty years, many literatures have reported to use PZT sensor array

based methods to monitor composite impact. Acellent Corporation in USA has

developed a series of integrated SHM systems named SMART Suitcase [2-3], including

an IMGenie instrument to detect impact events. The authors’ team also developed a

PZT network-based integrated multi-channel scanning system called PXI-ISS [4-5].

Since conventional PZT-based impact monitoring systems are developed aiming at

fulfilling a high-precision impact localization in composite structures, these systems

have to consist of many modules, such as charge amplifiers, data acquisition boards,

computer systems and work depending on complex algorithms. These kinds of

monitoring system are usually bulky, heavy, high power consumption and rely on

complex algorithms, which are usually used in lab research.

For damage monitoring, it may not be requested to be performed on line. But

impact is an instant event, and the aircraft structure may suffer from impacts during

both its service and maintenance, which are caused mainly by bird strikes, tool drop,

runway stones or ballistic impacts. These impact events have to be monitored on line.

On-line aircraft application of SHM does not allow too much additional weight, size and

power consumption attachment caused by the SHM systems. The on-line SHM system

applied to aircraft structures should meet following requirements: (1) Small size, light

weight. (2) Low power consumption. (3) Strong Electro Magnetic Compatibility

(EMC). (4) Large-scale structure monitoring.

Based on the requirements for on-line applications, some relevant works have been

reported. Metis Design Corporation in US has developed an impact monitoring system

both with active and passive [6], it is rounded and the diameter is only 40mm. However,

the analog signal acquired by it must be uploaded to PC for further processing. The

authors’ team also developed a miniaturized digital impact monitoring system, it has the

feature of small size (80×40×20mm3), light weight (180g) and low power consumption

(90mW) [7-9]. In spite of this, the size, weight and even the power still have the space

of optimization. Comparator array is used in the digital impact monitoring system to

realize the transformation of analog signal to digital sequence, which needs power

supply and 16 PZT sensors need to use 4 comparator chips. Besides, Altera FPGA is

used as the core processing chip and it is not a low power FPGA.

This paper develops an impact monitoring system of micro size, light weight and

low power consumption. It can support large numbers of piezoelectric (PZT) sensors to

monitor large-scale structures. The comparator array and the Altera FPGA is replaced

by a diode and a low power FPGA. A simple but efficient impact region localization

method is implemented in the low power FPGA of the micro and low power impact

monitoring system to detect and record the impact events. Furthermore, the FIIMS is

realized in a flexible way so that it can be embedded onto the composite structure to be

a mechatronic system to realize impact monitoring on-line continuously.

3

2. The architecture design of the FIIMS and impact region localization

algorithm

2.1 The architecture design of the FIIMS

When an impact occurs on an elastic structure, a stress wave will be caused to propagate

radially across the surface of the structure from the point of impact. The stress wave can

be caught by the PZT sensors bonded on the structure surface or built into the structure.

Based on it, the existing PZT-based SHM systems for impact monitoring are mostly

working in a passive way and the structure of these systems is illustrated in Figure 1(a).

Generally, the outputs of PZT sensors have to be pre-processed through the signal

filtering, amplification and AD converting circuit boards, occupying a large part of the

system and limiting the number of senor channels in the systems. And these impact

monitoring systems are often bus-based and only some traditional interfaces are

attached. Meanwhile, some impact monitoring methods are developed in these systems

to obtain impact monitoring results [10-15]. All these methods aim at high-precision

impact localization results in specific structures and are realized with complex

algorithms. Besides, most of these methods require a significant amount of information

like full impact data or structure characteristics.

In this paper, the requirements in on-line impact monitoring systems make the

method here quite different from the traditional development methods. This method

focuses on achieving approximate location results other than high-precision ones. The

architecture of the new impact monitoring system FIIMS is illustrated in Figure 1(b).

Filter, Amplifier A/D Converter

Co

mm

un

ica

tion

bu

sFilter, Amplifier A/D Converter

CPU/MCU

Massive data

storage

External

communicationPZT array

Complicated impact

monitoring methods

(a) The hardware architecture of the existing integrated impact monitoring systems

Filter, Comparator

Filter, Comparator

EEPROM

External

communication

PZT array

Simple impact monitoring methods

FPGA

(b) The hardware architecture of the FIIMS

Figure 1. The hardware architecture of the impact monitoring systems for impact monitoring

4

In the FIIMS, the signal filtering, amplification and AD converting circuits in

existing impact monitoring systems are replaced by filter array and comparator circuit.

The response signal from a PZT sensor is directly changed into a digital sequence by

comparing it with a pre-set trigger value. Although it reduces a large amount of

information from each sensor, it can access to more sensors and reduce the system

volume, weight and power-demand at the same time. Besides, it also improves the

ability of anti-electromagnetic interference by greatly minimizing the proportion of

analog circuits in it.

The Altera Field Programmable Gate Array (FPGA) is chosen in the development

method. Unlike ordinary CPU/MCU, FPGA has no given processor structure but offers

large amounts of logic gates, registers, SRAM, and routing resources, which makes it

more flexible and lower power request. Complicated logical and arithmetical algorithms

can be readily implemented in FPGA with parallel computation for much better

performance. Typically, thousands of operations can be performed in parallel in FPGA

during every clock cycle. And its compact physical size also makes it applicable in on-

line impact monitoring system.

2.2 The impact region localization algorithm

An illustration of the impact region localization algorithm adopted by the FIIMS is

shown in Figure 2. In Figure 2 (a), nine PZT sensors are placed on a composite

structure, forming four impact monitoring regions. With the conventional SHM system,

we can get the signals from the PZT sensors as shown in Figure 2(b). Then all these

signals are sent into filter, comparator array and FPGA. Achieving nine digital

sequences as a result as shown in Figure 2(c). According to the arrival time of the first

rising edge appeared in each sequence, the first three PZT sensors (PZT1, PZT2, PZT8)

can be recognized. So we can determine that the impact occurred in the left region

surrounded by PZT1, PZT2, PZT8 and PZT9. This algorithm can be applied on a large-

scale structure just by increasing the number of PZT sensors and the algorithm doesn’t

have strict restrictions on the monitoring objects.

(a) Sensor array and the region

5

0 1 2 3 4 5 6 7 8 9 10

PZT1

PZT2

PZT3

PZT4

PZT5

PZT6

PZT7

PZT8

PZT9

Time(ms)

0 1 2 3 4 5 6 7 8 9 10

PZT1

PZT2

PZT3

PZT4

PZT5

PZT6

PZT7

PZT8

PZT9

Time(ms)

3rd arrival rising edge

2nd arrival rising edge

1st arrival rising edge

PZT9

PZT8

PZT7

PZT6

PZT5

PZT4

PZT3

PZT2

PZT1PZT1

V

PZT2

PZT3

PZT4

PZT5

PZT6

PZT7

PZT8

PZT9

Time(ms)Time(ms)

(b) Analog response signals (c) Digital sequences

Figure 2. Illustration of the impact region localization method

For more precise impact localization, the algorithm based on energy-weighted

region location [16] can be used. The algorithm defines a characteristic parameter called

Impact Effect Factor (IEF) generated by the sensor digital sequence to characterize the

extent to which the sensor is affected by impact. IEF is calculated as equations (1), (2)

and (3) below, where DT is the high level duration time and ORE is the order of the first

rising edge. After calculating the IEF values of all digital sequences in the sensor array,

add the values of the four sensors that consist an impact monitoring region to

characterize the extent to which the region is affected by impact. The region which has

the largest Earea is the algorithm-defined area of impact. The flow chart is shown in

Figure 3.

1

n

i

i

DT DT

=

=∑ (1)

DTIEF

ORE= (2)

4

1

area j

j

E IEF=

=∑ (3)

Multiple

channels are

triggered by

impact

Generate

digital

sequences

Count OREs

and DTs

Calculate the

Earea of each

sub-region and

compare

Compute

IEFs

Locate sub-

region with the

maximum Earea

Generate

localization

result

Figure 3. The implementation process of the impact region localization method based on energy-

weighted region location

3. Hardware and software design of the FIIMS

Figure 4 shows the detail hardware structure of the FIIMS, which mainly consists of

filter and diode array module, core processing module, serial communication module

and storage module and power supply module.

6

E2PROM

Storage

Low

Power

FPGA

Flash*Freeze

Clock

Reset

LED

Processing

Core

Module

… …

Filter and

Diode Array

PZT1

RS232

Serial

5V

3.3V

2.5V

1.2V

Power Supply

PZTn

…

Figure 4. Hardware architecture of the FIIMS

3.1 Filter and diode array

A second-order passive high-pass filter array is reserved to reduce the load and

aerodynamic noise of low frequency. Comparator array is realized by a diode array

because the advantage of zero power consumption and small size. 25 diodes are

included in FIIMS to enable the access up to 25 PZT sensors. The outputs of the diode

array are provided as the inputs to the low power FPGA, through low power FPGA pin

level standard to transfer the analog signal into digital sequence, and then do impact

region localization.

3.2 Processing core

At the center of the FIIMS is the processing core which contains function modules for

data collection, processing and communication control. A low power FPGA is selected

as the main chip in processing core. A 10MHz crystal oscillator is adopted to provide

the system clock to the processing core and is divided in FPGA for different modules

such as storage module and serial module as shown in Figure 5. The FLASH-based

FPGA adopted in FIIMS will not lose all the configuration information after power

down, so external configuration chip is not needed, which reducing the power

consumption. An easy reset circuit is used for resting the program and giving the initial

value. Besides, the low power FPGA offers unique Flash*Freeze technology, allowing

the device to enter and exit ultra-low power Flash*Freeze mode when no power is

consumed by the I/O banks, clocks, JTAG pins, or PLL, and the device consumes as

little as 2µW in this mode.

3.3 Storage and serial communication

An external 64k byte CMOS-based EEPROM is added to the FIIMS design for the

storage of impact monitoring results. The EEPROM transmits and receives data via an

I2C port with the FPGA, in which an I2C IP core is created with a rate of 300kbits per

second. And the package of the EEPROM chip is nano DFN, with the size of

2×3×0.75mm. And the serial module is used to communicate with the PC for the

download of impact monitoring results.

7

Serial Module

Control Module

Digital

Sequence

Acquisition

Characteristic

Parameter

Extraction

Local

Storage

Storage Module

PZTs

Global

Clock

FPGA

validation

Led Module

PC

Write

DataRead

Data

Data

Package

Module

Send Order

Communication

Clock Module

EEPROM Clock

Serial Clock

Figure 5. Software architecture of the FIIMS

3.4 The FIIMS

Figure 6 shows the developed FIIMS. The FIIMS is capable of: (1) Micro size and

flexible (66.5×28×3mm3), light weight (3g) and low power consumption (30mW). (2)

Access up to 25 PZT sensors. (3) Real-time response and rapid impact region

localization and storage.

Figure 6. The developed flexible impact monitoring system

4. Functional validation of the FIIMS

To illustrate the capability of the FIIMS, it is integrated with a flexible PZT sensor array

to construct a large-scale impact monitoring network. The network is embedded onto

the surface of a large-scale carbon fiber reinforced plate with stiffeners by using a co-

8

curing process, which is shown in Figure 7. The impact monitoring ability of the FIIMS

is validated by applying multiple impacts in each monitoring regions and comparing

monitoring results with actual impacts. The monitoring regions and the impact locations

in each region are shown in Figure 8. Apply 5 impacts in each location. Thus there are

25 impact times in each region and 400 impact times on the whole structure.

Filter and Diode Array

FPGA

EEPROM

Serial Chip

Power and

Communication

Monitoring Region PZT Sensor Figure 7. Carbon fiber reinforced plate and large-scale impact monitoring network

1

2 3

45

CH1

CH2

CH3

CH4

CH5

CH6

CH7

CH8

CH9

CH10

CH11

CH12

CH13

CH14

CH15

CH16

CH17

CH18

CH19

CH20

CH21

CH22

CH23

CH24

CH25

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

Stiffener Impact location

Impact location in a region

Figure 8. Illustration of impact monitoring regions and impact locations

Figure 9(a)-Figure 9(b) shows the impact response digital sequences and Earea of

each region when impact was in region 1. It is clear that the order of the first rising edge

of digital sequences (CH2, CH1 and CH7) is correspond to the impact location.

Meanwhile the region which has the largest Earea is the same as the actual impact region.

Figure 10(a)-Figure 10(b) shows the impact response digital sequences and Earea of each

region when impact was in region 11. It is also clear that the order of the first rising

edge of digital sequences (CH13, CH14, CH18 and CH19) is correspond to the impact

location. Meanwhile the region which has the largest Earea is the same as the actual

impact region.

Table 1 shows the impact localization results. The accuracy rate on the whole

structure is around 95.5%. The results show that the FIIMS has the ability to realize

large-scale impact monitoring for composite structures when integrated with a PZT

sensor array and the accuracy rate of the impact monitoring is high.

9

Sub-region No.

3 4 5

Earea

6 7 8 10 11 12 13 14 1591 2 16

(a) Digital sequences (b) Earea of impact monitoring regions

Figure 9. Impact localization results

Sub-region No.

3 4 5

Earea

6 7 8 10 11 12 13 14 1591 2 16

(a) Digital sequences (b) Earea of impact monitoring regions

Figure 10. Impact localization result

Table 1. Impact localization results and accuracy rate

Region number Impact times Correct times Accuracy rate

1 25 25 100%

2 25 25 100%

3 25 25 100%

4 25 23 92%

5 25 22 88%

6 25 25 100%

7 25 25 100%

8 25 24 96%

9 25 21 84%

10 25 25 100%

11 25 24 96%

12 25 23 92%

13 25 24 96%

14 25 24 96%

15 25 25 100%

16 25 22 88%

Total 400 382 95.5%

10

5. Conclusions

In FIIMS, all the complex analog circuits are turned into digital sequences by diode

array and FPGA pin level standard. A low power FPGA is adopted as the processing

core to detect and record the impact events. All these characters in the FIIMS show its

potential for real applications in aircraft with advantages of compactness, light weight,

low power consumption and high efficiency. Our current effort is to acquire more

information such as the impact force level from the digital sequences besides the impact

region location results. And improvement is needed in the ruggedness and robustness of

the FIIMS for real applications. In addition, more tests are also needed to fully validate

the FIIMS.

Acknowledgements

This work is supported by the National Natural Science Foundation of China (Grant

No.51635007), the State Key Laboratory of Mechanics and Control of Mechanical

Structures (Nanjing University of Aeronautics and Astronautics in China) (Grant No.

MCMS-0517G01), Priority Academic Program Development of Jiangsu Higher

Education Institutions and Qing Lan project.

References

1. WJ Staszewski, S Mahzan and R Traynor, “Health monitoring of aerospace

composite structures–Active and passive approach”, Composites Science and

Technology, 69(11), pp 1678-1685, 2009.

2. XP Qing, HL Chan, SJ Beard and A Kumar, “An active diagnostic system for

structural health monitoring of rocket engines”, Journal of intelligent material

systems and structures, 17(7), pp 619-628, 2006.

3. www.acellent.com.

4. L Qiu and S Yuan, “On development of a multi-channel PZT array scanning

system and its evaluating application on UAV wing box”, Sensors and Actuators

A: Physical, 151(2), pp 220-230, 2009.

5. L Qiu, S Yuan, Q Wang, Y Sun and W Yang, “Design and experiment of PZT

network-based structural health monitoring scanning system”, Chinese Journal of

Aeronautics, 22(5), pp 505-512, 2009.

6. SS Kessler and EB Flynn, “Hybrid passive/active impact detection & localization

for aerospace structures”, Proc. 9th Int. Workshop on Structural Health

Monitoring, pp 2218-24, 2013.

7. S Yuan, P Liu and L Qiu, “A miniaturized composite impact monitor and its

evaluation research”, Sensors and Actuators A: Physical, 184, pp 182-192, 2012.

8. L Qiu, S Yuan, P Liu and W Qian, “Design of an all-digital impact monitoring

system for large-scale composite structures”, IEEE Transactions on

Instrumentation and Measurement, 62(7), pp 1990-2002, 2013.

9. P Liu, S Yuan and L Qiu, “Development of a PZT-based wireless digital monitor

for composite impact monitoring”, Smart Materials and Structures, 21(3), pp

035018, 2012.

11

10. ED Niri and S Salamone, “A probabilistic framework for acoustic emission (AE)

source localization in plate-like structures Smart Mater”, Smart Materials and

Structures, 21, pp 035009, 2012.

11. S Ahmari and M Yang, “Impact location and load identification through inverse

analysis with bounded uncertain measurements”, Smart Materials and Structures,

22, pp 085024, 2013.

12. L Qiu, B Liu, S Yuan and Z Su, “Impact imaging of aircraft composite structure

based on a model-independent spatial-wavenumber filter”, Ultrasonics, 64, pp 10-

24, 2016.

13. HK Jung, G Park, MD Todd, “Structural impact detection with vibro-haptic

interfaces”, Smart Materials and Structures, 25(7), pp 075041, 2016.

14. F Ciampa, S Boccardi and M Meo, “Factors affecting the imaging of the impact

location with inverse filtering and diffuse wave fields”, Journal of Intelligent

Material Systems and Structures, 27(11), pp 1523-1533, 2016.

15. Y Ren, L Qiu, S Yuan and Z Su, “A diagnostic imaging approach for online

characterization of multi-impact in aircraft composite structures based on a

scanning spatial-wavenumber filter of guided wave”, Mechanical Systems and

Signal Processing, 90, pp 44-63, 2017.

16. S Yuan, Y Ren, L Qiu and H Mei, “A multi-response-based wireless impact

monitoring network for aircraft composite structures”, IEEE Transactions on

Industrial Electronics, 63(12), pp 7712-7722, 2016.