Composite Structures 145 (2016) 226–241

Contents lists available at ScienceDirect

Composite Structures
journal homepage: www.elsevier.com/locate/compstruct

Ablation damage assessment of aircraft carbon fiber/epoxy composite

and its protection structures suffered from lightning strike
F.S. Wang a,⇑, Y.Y. Ji b, X.S. Yu a, H. Chen a, Z.F. Yue a
a
Advanced Material Testing Center, School of Mechanics, Civil Engineering and Architecture, Northwestern Polytechnical University, Xi’an 710129, PR China
b
Chengdu Aircraft Design and Research Institute of Aviation Industry Corporation of China, Chengdu 610041, PR China

a r t i c l e i n f o a b s t r a c t

Article history: Aircraft carbon fiber/epoxy composite material is sensitive to lightning strike. Its damage assessment and
Available online 8 March 2016 protection design suffered from lightning strike is becoming increasingly important. Four different types
of carbon fiber/epoxy composite laminates are selected, which are without protection, with full spraying
Keywords: aluminum coating, with local spraying aluminum coating and with spraying aluminum coating on glass
Carbon fiber/epoxy composite cloth pasted to fastener head, respectively. Impulse electrical current tests were performed by imple-
Lightning strike menting electrical current waveforms with different peak values with regard to different lightning zon-
Impulse electrical current test
ings. Three-dimensional finite element models of composite laminate and its protection structures are
Ablation damage
The spraying aluminum coating
accurately built to assess lightning ablation characteristics based on the coupled thermal/electrical/struc-
tural analysis and element deletion method, in which different electrical and thermal physical properties
of the elements are defined depending on different temperature conditions. The results show that simu-
lation results are good agreement with experimental results. Fiber damaged area, the damaged area and
the maximum damaged thickness increases with the increase of electrical current peak. Aluminum coat-
ing has a good effect on anti-lightning strike. The thicker aluminum layer and the better to anti-lightning
strike.
Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction
Composite materials have offered many advantages compared
with conventional materials especially at where high strength or
stiffness to weight ratio is concerned, which are widely applied
in various fields such as aerospace structures, pressure vessels
and automotive parts [1], etc. Lightning strike is a recurring prob-
lem as aircraft may encounter one strike between certain flight
hours. For example, a commercial aircraft may encounter one
strike about once a year [2]. Lightning strike is also a common
adversary for aircraft flight safety and usually occurs on aircraft
taking off, landing or passing through storm cloud [3]. Unfortu-
nately, carbon fiber reinforced polymer composites are unable to
conduct the high lightning electrical current and cannot suffi-
ciently prevent electromagnetic force from destroying. Lightning
strikes maybe result in embrittlement, delamination and failure
of composite materials [4]. It has a very important significance to
study damage behavior of composite materials suffered from light-
⇑ Corresponding author. Tel.: +86 29 88431002.
E-mail address: fswang@nwpu.edu.cn (F.S. Wang).
ning strike and improve its safety by employing appropriate pro-
tection systems.
Effects of lightning strike on aircraft can be classified into two
main categories, in which direct effects are associated with physi-
cal damage occurring at the attachment point and indirect effects
concern the interference due to electromagnetic coupling with sys-
tems or cabling [5]. The coupling direct effects of composite mate-
rials suffered from lightning strike attribute to the very complex
physical process and are difficulty to be completely simulated by
the numerical calculation such as finite element (FE) method at
present. Due to the high peak current and action integral, short
duration and extreme requirement on experimental operating con-
ditions and devices, there is little research on effects of composite
materials suffered from lightning strike. In recent years, study on
lightning damage of carbon fiber reinforced polymer composites
through experimental and numerical methods has only an initial
progress [1,6,7].
Feraboli and Miller [8] used a simulated generator to carry out
lightning strike experiment, in which strike peaks up to 50,000 A
and 28,000 V were inflicted on both pristine specimens and speci-
mens containing a Hilok stainless steel fastener. Damage area was
evaluated via ultrasonic scanning and further understanding of

http://dx.doi.org/10.1016/j.compstruct.2016.03.005
0263-8223/Ó 2016 Elsevier Ltd. All rights reserved.
 F.S. Wang et al. / Composite Structures 145 (2016) 226–241 227

damage pattern was also gained via the advanced optical micro- material is demonstrated when suffered from lightning strike.
scopy. Hirano et al. [9] examined damage evolution of graphite/ Influence of the sprayed aluminum coating on composite protec-
epoxy composite laminates suffered from lightning strikes. In tion structures is also evaluated in order to anti-lightning strike.
order to clarify the influence of lightning parameters and specimen Impulse electrical current test is performed by implementing
size, artificial lightning testing was performed on a series of com- current waveforms with different peak values with regard to dif-
posite laminates. Damage state was assessed via visual inspection, ferent lightning zonings. According to SAE ARP5414 standard about
ultrasonic scanning, micro X-ray inspection and sectional observa- aircraft lightning zoning [19], three lightning zonings such as IA, IB
tion. At the same time, the coupled thermal/electrical analysis of and IIB are selected in accordance with three test specimens for
carbon fiber reinforced polymer composites exposed to the simu- above each type of specimens, respectively. The corresponding
lated lightning current has been gradually conducted by several lightning current waveforms such as A, A + D and D wave are
scholars. Ogasawara et al. [10] developed a coupled thermal/ loaded to IA, IB and IIB zonings respectively. Taking into account
electrical model to analyze the angle-ply composite laminates, that aircraft test waveforms in SAE ARP5412 standard [20] and
which takes into account laminate lay-up and anisotropic lightning test methods in SAE ARP 5416 standard [21] are mainly
thermal/electrical behavior of unidirectional composite layers. How- applicable to lightning strike tests of aircraft metal structures
ever, their work ignored temperature dependency of electrical/ther- and fuselage, the actual specific criteria have not really been
mal material properties and electrical arc load applied in numerical formed to test direct effects of composite structures by artificial
simulation may result in inaccurate results. Abdelal and Murphy [11] simulated lightning currents in a laboratory environment. There-
also presented a physics based modeling procedure to predict ther- fore, above standards can only be used as a reference for artificial
mal damage of composite materials suffered from lightning strike, in lightning tests of composite structures, although these standards
which finite element method with non-linear material models was are not always suitable.
used to represent the extreme thermal behaviors of composite mate-
rial and the embedded copper mesh protection system. Muñoz et al. 2. Experimental description and numerical analysis method
[12] presented a finite element model to consider damage source
observed in lightning strike such as thermal damage caused by Joule 2.1. Experimental description
overheating and electromagnetic/acoustic pressure induced by
electrical arc around the attachment point. Composite specimens used for lightning strike test are carbon
Experimental and numerical simulation results from previous fiber/epoxy laminates T700/3234 with a 16-ply lay-up [45/
45/0
literatures can be concluded that composite materials will appear 2/45/90/
45/0]s, which size is 500 mm  250 mm  2 mm and
severe thermal ablation damage under lightning strike effects. thickness of each ply is 0.125 mm. For all the specimens with
Several scholars [13,14] used theoretical and experimental meth- spraying aluminum coating, both kinds of aluminum layer thick-
ods to study thermal ablation microscopic mechanism and thermal ness such as 0.1 and 0.2 mm is considered. For each type of the
shock effect on composite materials subjected to laser heat source. unprotected and protected specimens, 3 test specimens are pre-
Griffis et al. [15,16] established a model with respect to tempera- pared and so all 21 test specimens are used. Test specimen with
ture field and thermal ablation of carbon/epoxy composite local spraying aluminum coating is shown in Fig. 1. With regard
laminate subjected to laser irradiation. Thermal decomposition of to test specimens with spraying aluminum coating on glass cloth
organic matrix material such as epoxy resin was approximated pasted to fastener head, metal fastener is inserted in geometrical
by bulk heat capacity elevation over an appropriate temperature center of composite laminate panel.
range and the better results were achieved by this model. Then
thermal stress was analyzed by finite element method. Negarestani
et al. [17] and Zhou et al. [18] simulated ablation process of com-
posite laminate and low carbon steel subjected to laser irradiation,
respectively. The continuous variation of ablation boundary in
analysis model was achieved and ablation size was described
quantitatively.
A lot of problems are still worth studying although several
scholars have also used finite element method to analyze damage
mechanism of composite materials exposed to the simulated light-
ning electrical current. Based on previous work that have already
been done, further study on ablation damage mechanism of com-
posite material and its protection systems will be carried out when
exposed to the simulated lightning current. Four different types of
composite laminates are selected, which are without protection,
with full spraying aluminum coating, with local spraying alu-
minum coating and with spraying aluminum coating on glass cloth
pasted to fastener head, respectively. Three-dimensional finite ele-
ment model is developed to predict transient temperature field
together with the subsequent material removal when lightning
electrical current inflicts the composite laminate specimens. The
commercial finite element software ANSYS is applied for this pur-
pose by utilizing its parametric design language. Firstly, tempera-
ture dependency of electrical/thermal material properties is
defined. Then change conditions of boundary and load as well as
the excited phase during lightning strike process are incorporated
into the model combing with element birth and death technique.
Taking into account fiber damage and resin deterioration, transient Fig. 1. Test specimen with local aluminum coating shown in the shaded area (Unit:
heat transfer and heat decomposition mechanism of composite mm).
228 F.S. Wang et al. / Composite Structures 145 (2016) 226–241
Lightning current generator and its impulse current waveform
are shown in Fig. 2. The impulse current generator shown in
Fig. 2(a) is used to simulate natural lightning strike, which is com-
posed of the main loop of the shock, capacitor constant voltage
charging device, the decoupling network box, control system,
waveform acquisition and data microcomputer processing system.
A, B, C and D wave presented in Fig. 2(b) is defined in SAE ARP5412
standard. The double exponential current waveform shown in
Fig. 2(c) and generated by impulse current generator can be
expressed as:
iðtÞ ¼ I0 ðe
at
e
bt Þ
where I0 is current constant, a is reciprocal value of wave tail
time constant, b is reciprocal value of wave front time constant
and t is the time. It also can be described as t1/t2 impulse electrical
current by introducing two parameters t1 and t2 presented in Fig. 2
(c), in which t1 is front rise time, t2 is the time to 50% of current
peak and Ip is current peak. When time values of tA, tB and tD cor-
responding to A, B and D points of impulse current waveform curve
are calculated by three known parameters t1, t2 and tF respectively,
three simultaneous equations corresponding to A, B and D points
can be combined to identify I0, a and b by four parameters tA, tB,
tD and Ip as described in our previous published paper [22].
The 10/350 ls impulse electrical current waveform with double
exponential type and test jig for lightning strike test are shown in
Fig. 3. The impulse current generator in this test can generate a
standard 10/350 ls impulse electrical current shown in Fig. 3(a)
with a 5–100 kA range of current peak. In order to prevent test
specimens from going flying which affects on the accuracy of test
Fig. 2. Lightning current generator and impulse current waveform.
results and load impulse electrical current easily, a reliable and
convenient test jig for composite specimens shown in Fig. 3(b)
should be adopted because impulse electrical current generator
will generate large current peaks in the test. At the same time,
the left and right sides of test specimens are clamped as well as
the top and bottom sides of test specimens are pressed through
the protected covers and bayonet forceps in order to ensure test
specimens fixing well. Both of copper strips are placed on the left
and right sides of test specimens respectively, which ends are
linked with ground wire so as to ensure a good ground connection.
After test specimens were completely fixed, a thick copper probe is
ð1Þ
placed on the slip crossbar above the center of test specimens and
distance of test specimens from copper probe is set to 3 mm. Cop-
per probe is connected with a discharge copper ball through cop-
per sheet, which is fixed on upside copper probe. Impulse
electrical current generated by gap breakdown of copper ball con-
ducts to copper probe via copper sheet and then injects onto the
surface of test specimens.
Because composite specimen as a small part of the whole air-
craft structure is selected for lightning strike test, so double expo-
nential current waveforms such as above 10/350 ls impulse
electrical current can meet the test requirement, which is varying
with similar rules from A and D wave. Composite specimen is
added in the closed test system. Due to nonlinear factors of com-
posite specimen such as resistance and inductance and so on, elec-
trical parameters of circuit system will vary and result in the
difference between theoretical and final actual electrical current
waveform attached to surface of test specimen. As the most impor-
tant factor affecting damage extent of composite structures
 F.S. Wang et al. / Composite Structures 145 (2016) 226–241
(a) The 10/350 µs impulse current waveform
(b) Test jig
Fig. 3. The 10/350 ls impulse electrical current waveform and test jig for lightning
strike test.
suffered from lightning strike, integral action of the selected
impulse electrical current waveform should be the same as that
of the corresponding theoretical current waveform required to be
loaded in this test [9,10], which thus will not affect the validity
of final test results within the allowable error range. According to
the equivalent transformation of integral action, test waveforms
collected by oscilloscope are shown in Fig. 4 and current peaks
are given in Table 1 for different lightning zonings. The
10/350 ls impulse current waveform with 31.3 kA current peak
is used and equivalent test waveform collected by oscilloscope is
shown in Fig. 4(a), which corresponds to IIB lightning zoning and
lightning current waveform D. The 10/350 ls impulse current
waveform with 88.4kA current peak is used and equivalent test
waveform collected by oscilloscope is shown in Fig. 4(b), which
corresponds to IA lightning zoning and lightning current waveform
A. The 10/350 ls impulse current waveform with 93.7 kA current
peak is used and equivalent test waveform collected by oscillo-
scope is shown in Fig. 4(c), which corresponds to IB lightning zon-
ing and lightning current waveforms A + D. Because the pre-
acquisition time and sampling length by automatically setting is
employed for impulse electrical current waveform, wave length
of effective electrical current is almost constant and about
1000 ls, although there is some difference for pre-acquisition time
and sampling length shown in Fig. 4(a–c), respectively.
2.2. Numerical analysis method
Electrical arc is a breakdown response due to excessive voltage
applied to a gap between two electrodes. Once breakdown hap-
pens, an ionized channel is formed in the gas and electrical current
flows through the ionized channel. Arc velocity of lightning strike
is around 2  105 m/s and carries about an order of 2–20  10–4
229
Coulombs charge [11]. Thermogravimetric analysis results under
the heating rate range from 2.5 to 20 °C/min show that thermal
decomposition temperature of resin matrix for carbon fiber/epoxy
composites starts at about 300 °C and terminates at about 500 °C,
which is estimated to have a temperature increase range from
100 to 200 °C under the 1000 °C/min heating rate [10]. Here, it is
supposed that resin matrix will decompose between 300 and
500 °C when subjected to the increased temperature considering
the typical thermosetting behavior. Electrical arc will continue to
contact with the conductive carbon fibers when temperature is
above 500 °C. And carbon fibers will be charred because a large
of lightning electrical current carried by electrical arc can quickly
heat up carbon fibers. Sublimation of carbon fibers will occur once
the temperature exceeds 3000 °C. Joule heat will always be gener-
ated until carbon fibers appear breakage and thus ablation temper-
ature of composite material will be controlled by sublimation
temperature of fiber breakage. So, the final ablation temperature
of carbon fiber/epoxy composite is defined as 3316 °C [15].
A possible way to provide enough electrical conductivity for
lightning strike protection is by changing a non-conductive mate-
rial into a conductive one. For example, conductive coating with
some thickness can be used to provide enough conductivity [23].
Metal aluminum is a well-known material with low volume resis-
tivity that is useful for preventing aircraft damage from lightning
strike. When carbon fibers are embedded into resin matrix to make
up of composite laminate, electrical resistance will be 2000 times
larger than that of aluminum material. Lightning strike with
enough high current density may be dissipated through heat
energy and results in carbon fiber and resin matrix to be degraded.
Aluminum coating is generally sprayed on surface of carbon fiber/
epoxy composite laminate when suffered from lightning strike
[24]. The melting temperature is 933 °C, the boiling temperature
is 2793 °C and critical temperature is 8860 °C for aluminum coat-
ing. Complex coupled physical process of the protected composite
laminate will happen under lightning electrical current and alu-
minum coating may heat up, melt and vaporize. Explosive boiling
will occur once surface temperature reaches approximately 90% of
critical temperature, which results in a mixture ejection of vapor
and liquid droplets. Therefore, upper ablation limit of aluminum
coating is assumed as explosive boiling in analysis model [25].
Three-dimensional finite element models of benchmark speci-
men, the specimen with full spraying aluminum coating, the spec-
imen with local spraying aluminum coating and the specimen with
spraying aluminum coating on glass cloth are built by thermal/
electrical solid elements numbering SOLID 69 in ANSYS element
library in order to simulate ablation process of these specimens
suffered from lightning strike shown in Fig. 5. Lightning electrical
current is applied to central region of simulation model. Transient
temperature field of composite laminate and aluminum coating is
solved through the nonlinear thermal analysis procedure in which
automatic time-step option is used to determine the time-step size
in response to electrical current state under consideration. Abla-
tion criterion of finite elements is defined once numerical solution
is obtained, which specifies that the final ablation temperature of
composite laminate is 3316 °C and that of aluminum coating is
7974 °C, respectively. Combing with comparison results of average
temperature and the corresponding ablation temperature, birth
and death condition of all elements is defined. According to instan-
taneous temperature field, lightning electrical current load will
continue to be applied to finite elements if their average tempera-
ture is less than ablation temperature. While the elements will be
ablated and considered removing if average temperature of ele-
ments is greater than ablation temperature and the deleted ele-
ments boundary will be used as a new boundary to apply
lightning electrical current load for next time step. At this time,
the rest elements and nodes are renumbered to generate a new
230 F.S. Wang et al. / Composite Structures 145 (2016) 226–241

(a) IIB lightning zoning (b) IA lightning zoning

(c) IB lightning zoning

Fig. 4. Equivalent test waveforms collected by oscilloscope for different lightning zonings.

Table 1 straint and new boundary condition are specified as the initial con-
Current peaks at different lightning zonings for the 10/350 ls impulse electrical ditions for numerical calculation of next time step. The iterative
current waveform. computation is repeated until the final results are obtained. Above
Integral action (A2s) Lightning Impulse Current element deletion program is compiled by APDL parametric design
zoning current waveform language of ANSYS software.
peak (kA) number Thermal and electrical properties of composite laminate [16,28]
2  106 (A waveform) IA 88.4 10/350-1 and aluminum material [25,29] are defined in Table 2. Boundary
2.25  106 (A + D waveform) IB 93.7 10/350-2 condition is simulated to match with above experimental setup
0.25  106 (D waveform) IIB 31.3 10/350-3
as far as possible. Thermal radiation is given for the upper and side
surfaces of test specimen and bottom surface is adiabatic. Thermal
emission ratio is 0.9 and air temperature is 25 °C. Electrical poten-
finite element model. Different electrical and thermal properties of tial at the bottom and side surfaces of a specimen are assumed to
the elements that not ablated in new model are defined depending be zero. Ogasawara et al. [10] applied electrical charge at center
on different temperature condition. It is different from the only node of composite laminate plate using time profile of a concen-
definition of thermal properties which is published in our previous trated electrical charge, in which lightning bolt size is ignored that
paper [26,27]. New boundary surfaces are the initial internal sur- significantly influence on the predicted numerical results. Thus a
faces of elements adjacent to the deleted elements. Moreover, conductive electrical channel with 10 mm diameter is herein
the unloaded surfaces may be loaded once again as a movement assumed for numerical simulation [11]. Finite element models of
of lightning electrical current. The applied lightning current, con- composite laminate and its protection structures are shown in
 F.S. Wang et al. / Composite Structures 145 (2016) 226–241 231

metal fastener is 1Cr18Ni9Ti and its properties are shown in

Table 3. In Fig. 6(c), actual size of metal fastener is shown with full
line and cylindrical metal fastener is replaced and shown with dot-
ting line for simplicity, which diameter and length of is 8 and
2 mm, respectively. Material of glass cloth is EW100A and its prop-
erties are also shown in Table 3. Geometrical size of glass cloth is
30 mm  30 mm  0.1 mm. Because electrical resistivity of glass
cloth is so large, so only a small amount of electrical current may
conduct through the glass cloth to the fastener. Here, thermal
and electrical coupling elements are used to simulate glass cloth
and fastener but lightning ablation criterion is not considered.
Finite element models of metal fastener, glass cloth and composite
laminate are shown in Fig. 6(c–e), respectively.

3. Results and discussion

3.1. Numerical method validity

As a validity example in reference [9], carbon fiber/epoxy com-

Fig. 5. Ablation simulation process suffered from lightning strike.
Fig. 6. Total element number of benchmark specimen is 40,000.
Finite element model of benchmark specimen is shown in Fig. 6
(a) when suffered from impulse electrical current. Node distribu-
tion is shown in Fig. 6(b) to apply on lightning electrical current
load. Total element number of the specimen with full spraying alu-
minum coating is 42,500 in which aluminum element number is
2500. Total element number of the specimen with local spraying
aluminum coating is 82,280 in which aluminum element number
is 2280.
For fastening parts installed in composite structures such as air-
craft fuel tank, some protection methods such as sealing and insu-
lating layers are usually used to prevent metal fasteners from
generating electrical spark and avoid accidental explosion because
of contact between fuel vapor and electrical arc. Total element
number of the specimen with spraying aluminum coating on glass
cloth is 90,043, in which number of metal fastener, glass cloth and
aluminum elements is 513, 265 and 5000, respectively. Material of
posite laminate IM600/133 is used for ablation damage analysis
when suffered from lightning strike. Its size is 150 mm 
100 mm  4.7 mm and stacking sequence is [45/0/
45/90]s. Its
ply number is 32 and thickness of each ply is 0.147 mm. Because
IM600/133 and T700/3234 composite laminates are both made
up of carbon fiber and epoxy resin in which volume contents of car-
bon fiber such as 60% are almost same, so here thermal and electri-
cal properties of both kinds of composite laminates are
approximately regarded as same values and defined in Table 2.
Above numerical simulation method is used to assess lightning
ablation damage consistent with experimental setup in reference
[9], in which element type selection, loading way of impulse electri-
cal current, constraint and boundary condition definition is same as
that of carbon fiber/epoxy composite laminate T700/3234. The
2.6/10.5 ls impulse electrical current is used and current parame-
ters are given in Table 4 when current peaks are 20, 30 and 40 kA
respectively, in which accurate values of I0, a and b are obtained
by an optimization identification method combing global Multi-
Island Genetic Algorithm with local Sequential Quadratic Program-
ming (NLPQL). In order to verify reliability of the calculated results,
the calculated values of electrical charge and action integral
obtained according to current waveform curve are given and com-
pared with those in reference [9] by calculating their errors. At the
same time, approximate values of electrical charge and action inte-
gral in our previous published paper [26] as preliminary study are
also given and compared with those in references [9] by calculating

Table 2
Thermal and electrical properties of composite laminate and aluminum material.

Material Temperature Density Specific heat Thermal conductivity Electrical conductivity (1/Omm)
(°C) (kg/mm3) (J/kg °C) (1/Omm)
Longitudinal Transverse Longitudinal Transverse Through-thickness
Composite laminate 25 1.52  10
6 1065 0.008 0.00067 35.97 0.001145 3.876  10
6
343 1.52  10
6 2100 0.002608 0.00018 35.97 0.001145 3.876  10
6
500 1.1  10
6 2100 0.001736 0.0001 35.97 2 2
510 1.1  10
6 1700 0.001736 0.0001 35.97 2 2
1000 1.1  10
6 1900 0.001736 0.0001 35.97 2 2
3316 1.1  10
6 2509 0.001736 0.0001 35.97 2 2
Aluminum 25 2.7  10
6 940 0.270 3.69  104
311 2.7  10
6 1013 0.274 3.71  104
526 2.7  10
6 1013 0.231 1.77  104
1351 2.7  10
6 1013 0.107 3.62  103
2727 2.256  10
6 1082 0.148 1.99  103
3576 2.157  10
6 1082 0.151 1.83  103
5538 1.893  10
6 1082 0.163 1.17  103
6292 1.836  10
6 1138 0.168 1.06  103
7974 1.4  10
6 1138 0.0044 25.2
232 F.S. Wang et al. / Composite Structures 145 (2016) 226–241

(a) Benchmark specimen suffered from (b) Node distribution to apply on lightning
impulse electrical current electrical current load

(c) Metal fastener with above geometrical model for the specimen with spraying
aluminum coating on glass cloth (Unit: mm)

(e) Composite laminate for the specimen with

spraying aluminum coating on glass cloth

(d) Glass cloth for the specimen with

spraying aluminum coating on it
Fig. 6. Finite element models of composite laminate and its protection structures.

their errors, in which t1 is denoted as the time to current peak in culated values by optimization identification parameters are more
order to avoid complex mathematical calculation of real front time. close to those in reference [9] than approximate values as a whole,
For electrical charge and action integral, it can be seen that the cal- although two errors of electrical charge value is more bigger.
 F.S. Wang et al. / Composite Structures 145 (2016) 226–241 233

Table 3
Material properties of metal fastener and glass cloth.

Material Elastic modulus Specific heat Density Poisson Shear modulus Resistivity Thermal conductivity
(GPa) (J/kg°C) (kg/m3) ratio (GPa) (Xm) (W/m °C)
Metal fastener 206 450 2164 0.3 59 5  10
7 50
Glass cloth 85 800 2550 0.21 36 1  1014 1.2

Table 4
Performance parameters of 2.6/10.5 ls impulse current waveform.

t1/t2 (ls) a (s
1) b (s
1) Current peak Ip (kA) I0 (A) Electrical charge (C) and its error Action integral (A2s) and its error Duration time (ls)
2.6/10.5 174,777 380,346 40 141,837 0.4386/10.5% 11,519/0.2% 200
[0.49] [11,500]
(0.514/4.9%) (12,083/5.1%)
181,155 365,911 30 117,557 0.3277/11.4% 6504/3.2%
[0.37] [6300]
(0.386/4.3%) (6797/7.9%)
191,437 344,767 20 93,425 0.2170/8.5% 2899/0.0%
[0.2] [2900]
(0.257/28.5%) (3021/4.2%)

Note: [] represents the values in reference [9]. () represents the approximate values in our previous published paper [26] and their errors compared with the values in
references [9].

Transient temperature field of composite laminate obtained by
numerical calculation is compared with damage pattern of exper-
imental results as shown in Fig. 7 when subjected to 20, 30 and
40 kA electrical current peaks, respectively. It can be seen that
numerical simulation results can be well matched with the corre-
sponding experimental results. Ablation area, ablation depth and
ablation width vertical to fiber direction will increase with the
increase of electrical current peak. Additionally, the deleted ele-
ment number will also increase with the increase of electrical cur-
rent peak.
Fiber damaged elements of composite laminate are those
which ablation temperature reaches fracture and sublimation
temperature of carbon fiber such as 3316 °C at attachment point
of impulse electrical current, which area can be referred as fiber
damaged area. Relationship of fiber damaged area and electrical
current peak by numerical calculation and comparison with
experimental results is shown in Fig. 8. It can be seen that numer-
ical simulation results can be well matched with experimental
results and fiber damaged area will increase with the increase
of electrical current peak. Fiber damaged area will be further
expanded because anisotropic electrical conductivity of composite
laminate increases the possibility of internal explosion. As inter-
nal explosion in numerical analysis is not considered, it may be
a reason that numerical results are slightly smaller than experi-
mental results. Epoxy resin may be decomposed at ambient
regions of ablation elements due to high temperature although
Joule heat generated by impulse electrical current is not sufficient
to make carbon fibers be failure. Two narrow belts of high tem-
perature can be observed from both sides of ablation elements
along 45 degree direction at the top lamina, which are also sym-
metrical with respect to the attachment point of impulse electri-
cal current. Though the temperature does not exceed fracture and
sublimation temperature of carbon fibers, it is over decomposition
temperature of epoxy resin. Resin deterioration elements of com-
posite laminate are those which ablation temperature is between
decomposition temperature of epoxy resin and fracture and sub-
limation temperature of carbon fiber, which area can be referred
as resin deterioration area.

Fig. 7. Transient temperature field of composite laminate obtained by numerical calculation and comparison with experimental results when subjected to different electrical
current peaks.
234 F.S. Wang et al. / Composite Structures 145 (2016) 226–241
Fig. 8. Relationship of fiber damaged area and electrical current peak by numerical
calculation and comparison with experimental results.
Ablation damage will be generated at surface and internal com-
posite laminate due to the powerful electrical current, which is
mainly caused by carbon fiber breakage and thermal decomposi-
tion of resin matrix. So, here the whole damaged area is assumed
to be that greater than decomposition temperature of resin matrix.
At the same time, the maximum damaged thickness is also
assumed to be that greater than decomposition temperature of
resin matrix along thickness direction [10]. When decomposition
temperature of resin matrix is 300, 400 and 500 °C respectively,
relationship of the damaged area and action integral by numerical
calculation and comparison with experimental results is shown in
Fig. 9. Relationship of the maximum damaged thickness and elec-
trical current peak by numerical calculation and comparison with
experimental results is shown in Fig. 10. It can be seen that the
damaged area and the maximum damaged thickness estimated
from the calculation curves shows good agreement with experi-
mental results when decomposition temperature of resin matrix
is supposed as 400 °C. The damaged area and the maximum dam-
aged thickness will increase with the increase of electrical current
peak. So, lightning ablation model established in this paper can be
used to effectively simulate damage behavior of composite lami-
nate suffered from lightning strike. For the presented failure crite-
ria such as fiber damaged and resin deterioration, failure elements
Fig. 9. Relationship of the damaged area and action integral by numerical
calculation and comparison with experimental results.
Fig. 10. Relationship of the maximum damaged thickness and electrical current
peak by numerical calculation and comparison with experimental results.
of composite laminate are shown in Fig. 11 for four selected tem-
perature when subjected to 40 kA electrical current peak, which
will be deleted when the temperature is greater than the selected
temperature.
However, internal explosion will occur when internal heat
quantity of composite laminate continues to accumulate to a cer-
tain degree because surface layer will hinder instantaneous release
of internal ablation heat. It makes the damaged area and the max-
imum damaged thickness increase due to internal explosion. Addi-
tionally, temperature field distribution of composite laminate
along the thickness direction is non-uniform at the attachment
regions of impulse electrical current, which will inevitably result
in a large of residual thermal stress along thickness direction.
Residual thermal stress can also make the damaged area and the
maximum damaged thickness of composite laminate increase
when subjected to lightning electrical current. As residual thermal
stress in numerical analysis is also not considered, so it is another
reason that numerical results will be slightly smaller than those of
the actual experiment results.
3.2. Experimental and numerical results
3.2.1. Benchmark specimens
Clear ablation damage such as fiber breakage and tilting
appears at attachment region of the unprotected composite lami-
nate when impulse electrical current reaches specimen surface
through copper probe. At the same time, the broken fibers begin
to continuously burn within a few seconds because thermal effect
of impulse electrical current results in rapid temperature increase
of specimen surface. Temperature at the bottom of test specimen is
so high and there is also accompanied with an irritating odor,
which may be produced by the decomposed resin molecules. Black
matter below copper probe is attached to specimen surface due to
resin matrix decomposition and fiber carbonization.
Transient temperature field of composite laminate obtained by
numerical calculation is compared with damage pattern of experi-
mental results as shown in Fig. 12 when subjected to different elec-
trical current peaks. Because both sides of benchmark specimen is
placed upside down in test due to incorrect operation, so the stack-
ing sequence of composite laminate is changed and becomes
[
45/45/02/
45/90/45/0]s. As the stacking sequence transforma-
tion of benchmark specimen does not affect the accurate evaluation
of lightning ablation damage, here stacking sequence of finite
element model is also set as [
45/45/02/
45/90/45/0]s in order
 F.S. Wang et al. / Composite Structures 145 (2016) 226–241 235

ć
(a) Resin deterioration greater than 400ć (b) Resin deterioration greater than 1000ć

(c) Resin deterioration greater than 2000ć (d) Fiber damaged at 3316ć
Fig. 11. Failure elements of composite laminate for four selected temperature when subjected to 40 kA electrical current peak.

Fig. 12. Transient temperature field of composite laminate obtained by numerical calculation and comparison with experimental results when subjected to different
electrical current peaks.

to compare with test results. Impulse electrical current with high
peak and relatively long action time conducts mainly along carbon
fiber direction with
45 degree direction on top of composite lam-
inate and instantly generates a large amount of Joule heat when it is
attached to composite plate. Because sharp rise temperature makes
carbon fiber occur fracture and electrical current conduction along
carbon fiber direction is interrupted, so a large part of impulse elec-
trical current is too late to release out from edge of benchmark
specimen along carbon fiber direction and only conducts along
other both of directions, which has relatively small electrical con-
ductivity and little effect on electrical current conduction. And
because electrical resistance between composite layers is too large,
so only a part of Joule heat generating instantly will transfer rapidly
from the first layer to other layers when impulse electrical current
spreads out on top of composite laminate with extremely short
action time. Thus, a small area of ablation damage will occur on
other layers and ablation pit will occur at lightning attachment
region along thickness direction. Finally, ablation damage area of
benchmark specimen appears ellipse.
For the IIB lightning zoning with 31.3 kA current peak, temper-
ature of composite laminate rises rapidly under direct effect of
impulse electrical current and that at current attachment region
236 F.S. Wang et al. / Composite Structures 145 (2016) 226–241

(a) The specimen obtained by numerical calculation and comparison with

experimental results for 0.1mm aluminum thickness

(b) The specimen obtained by numerical calculation and comparison with

experimental results for 0.2mm aluminum thickness

(c) Composite laminate for 0.1mm aluminum thickness

(d) Composite laminate for 0.2mm aluminum thickness

Fig. 13. Transient temperature field of the specimen and composite laminate with full spraying aluminum when subjected to different electrical current peaks.

is even up to 3000 °C. Heat decomposition arises for epoxy resin of composite layer is mainly along carbon fiber direction and there
matrix as well as sublimation and fracture arises for carbon fiber has ablation damage and delamination along thickness direction.
instantaneously. It can be clearly observed that surface damage At the same time, there is a small area of fiber fracture and swell
 F.S. Wang et al. / Composite Structures 145 (2016) 226–241 237

(a) The specimen obtained by numerical calculation and comparison with

experimental results for 0.1mm aluminum thickness

(b) The specimen obtained by numerical calculation and comparison with

experimental results for 0.2mm aluminum thickness

(c) Composite laminate for 0.1mm aluminum thickness

(d) Composite laminate for 0.2mm aluminum thickness

Fig. 14. Transient temperature field of the specimen and composite laminate with local spraying aluminum when subjected to different electrical current peaks.

with discrete distribution around the current attachment region.
Surface layer resin at these regions has been disappeared and car-
bon fiber is exposed.
For the IA lightning zoning with 88.4 kA current peak, ablation
damage is more serious and ablation mark is more obvious because
integral action of impulse electrical current waveform loaded in IA
lightning zoning is about 8 times of that in IIB lightning zoning.
With the peak of impulse electrical current increasing, a part of
carbon fibers directly fly out in addition to tilt of carbon fibers,
fracture and combustion for benchmark specimen compared with
that in IIB lightning zoning. When the large and transient electrical
current attaches to composite panel, the maximum electrical
238 F.S. Wang et al. / Composite Structures 145 (2016) 226–241
current results in the highest temperature of composite laminate
at attachment region because of its relatively poor electrical con-
ductivity and then the sharp high temperature is acted on internal
composite plate. Because sublimation temperature of carbon fiber
is much higher than thermal decomposition temperature of resin
matrix, so resin matrix will firstly occur thermal decomposition
and release a large amount of pyrolysis gas, which may be wrapped
in composite plate. Expansion of internal gas between each layer
will make composite plate bulge under external action of air pres-
sure and thus internal explosion comes into being. Internal explo-
sion has a huge shock and finally make composite panel produce
more severe damage such as delamination damage. Damage area,
damage depth and tilt height of carbon fiber and resin matrix will
increase with peak increase of impulse current.
For the IB lightning zoning with 93.7 kA current peak, ablation
damage is the most serious and ablation trace is the most obvious
for test specimens when compared with the previous two experi-
ments. Because integral action of impulse electrical current wave-
form in IB lightning zoning is about 9 times of that in IIB lightning
zoning, damage area, damage depth and tilt height of carbon fiber
and resin matrix all increases to the maximum value when peak
value of impulse current tends to the maximum.
3.2.2. The specimens with full spraying aluminum coating
When thickness of the spraying aluminum is 0.1 and 0.2 mm,
transient temperature field of the specimen with full spraying alu-
minum coating is obtained by numerical calculation and compared
with experimental results as shown in Fig. 13(a) and (b), respec-
tively. It can be seen that ablation damage area of test specimen
is mainly presented as a circular area shape, in which attachment
point is the center of circle. Aluminum layer is uniformly dis-
tributed and electrical current conducts outward to specimen
boundary along aluminum layer surface. When impulse electrical
current attaches to the full spraying aluminum layer, substantial
Joule heat is instantly generated and makes temperature of alu-
minum layer rise rapidly. Intense vaporization of aluminum layer
will occur when temperature reaches the boiling temperature
2793 °C.
In order to analyze lightning protection effect of the specimen
with full spraying aluminum coating and compare with benchmark
specimen, temperature distribution of composite laminate with 0.1
and 0.2 mm aluminum layer thickness is shown in Fig. 13
(c) and (d), respectively. It can be seen that Joule heat generated
by electrical current conducts to composite plate via aluminum
layer so rapidly that circular ablation area appears on top 45
degree layer of composite laminate as the attachment condition
of aluminum layer, which neglects anisotropic effect of composite
laminate itself. When impulse electrical current with short dura-
tion is applied to central region of the specimen, a part of Joule heat
generated by electrical current with high peak will conduct along
thickness direction of composite laminate plate and so different
sizes of ablation pit at attachment region is produced.
For the IIB lightning zoning with 31.3 kA current peak, surface
damage area of the protected test specimen and composite damage
area under the protected layer is relatively smaller than that of
benchmark specimen. Because aluminum layer as lightning protec-
tion method is an isotropic material and its electrical conductivity
is much larger than composite material, so impulse electrical cur-
rent will quickly conduct to outward boundary along surface of
aluminum layer from the attachment region and the ablation dam-
age shape of test specimen appears mainly as a circular area.
Although a part of impulse electrical current conducts to compos-
ite plate along thickness direction and makes temperature of com-
posite plate rise, it does not cause serious ablation damage of
composite laminate.
For the IA lightning zoning with 88.4 kA current peak and the IB
lightning zoning with 93.7 kA current peak, similar damage can be
observed. Because peak value of the applied impulse current
increases especially when peak value of impulse current becomes
the maximum, composite material will appear larger damage such
as fiber fracture and resin decomposition although aluminum layer
of test specimen is in favor of current conduction.
3.2.3. The specimens with local spraying aluminum coating
When thickness of the spraying aluminum is 0.1 and 0.2 mm,
transient temperature field of the specimen with local spraying
aluminum coating is obtained by numerical calculation and com-
pared with experimental results as shown in Fig. 14(a) and (b),
respectively. It can be seen that ablation damage of test specimen
mainly distributes along width direction of aluminum layer.
Impulse electrical current attached to test specimen will be prior-
ity to conduct rapidly from attachment point to outward bound-
ary along width direction of aluminum layer, in which aluminum
layer appears serious damage. Only a small part of electrical
current conducts to outward boundary along length direction of
aluminum layer, in which aluminum layer appears minor damage
area.
In order to analyze lightning protection effect of the specimen
with local spraying aluminum coating and compare with that with
full spraying aluminum coating, temperature distribution of com-
posite laminate with 0.1 and 0.2 mm aluminum layer thickness is
shown in Fig. 14(c) and (d), respectively. It can be seen that Joule
heat generated by impulse lightning current conducts rapidly to
composite laminate through aluminum layer. Elliptical ablation
area appears on composite plate around the attachment point of
lightning electrical current. Although Joule heat does not make
composite laminate under aluminum layer along width direction
appear a large area of ablation damage, a large area of resin matrix
decomposition occurs on composite material because ablation
temperature exceeds thermal resin decomposition temperature
400 °C. In this study, distance of discharge electrode from test
specimen is set to 3 mm in lightning strike test, which can basically
guarantee impulse electrical current attaching to the geometrical
center of test specimen. But there is still a part of electrical arc
attaching to composite laminate without aluminum coating when
peak current is much larger. If lightning current in reality attaches
to composite laminate rather than aluminum layer, damage degree
of the specimen with local spraying aluminum coating and bench-
mark specimen would be same and aluminum coating will lose its
due lightning protection effect.
For the IIB lightning zoning with 31.3 kA current peak, test spec-
imen with local spraying aluminum protection has larger surface
ablation damage than that with all spraying aluminum protection.
Aluminum layer of test specimen with full spraying aluminum is
uniformly distributed and electrical current mainly conducts out-
ward boundary with circular area shape along the surface of alu-
minum layer. For most of electrical current, it’s too late to release
out from outer boundary and only conduct downwards along thick-
ness direction of composite plate through the aluminum layer.
Therefore, surface damage area of composite plate is smaller for test
specimens with full spraying aluminum protection.
For the IA lightning zoning with 88.4 kA current peak and the IB
lightning zoning with 93.7 kA current peak, it can be seen that
ablation damage of test specimen with local spraying aluminum
still mainly distributes along width direction of aluminum layer,
in which a part of damage appears on composite plate where no
aluminum layer covering exists. Only a small part of electrical cur-
rent conducts downwards through the aluminum layer along
thickness direction and damage area of composite plate is rela-
tively smaller.
 F.S. Wang et al. / Composite Structures 145 (2016) 226–241 239

(a) The specimen obtained by numerical calculation and comparison with

experimental results for 0.1mm aluminum thickness

(b) The specimen obtained by numerical calculation and comparison with

experimental results for 0.2mm aluminum thickness

(c) Composite laminate for 0.1mm aluminum thickness

(d) Composite laminate for 0.2mm aluminum thickness

Fig. 15. Transient temperature field of the specimen and composite laminate with spraying aluminum on glass cloth when subjected to different electrical current peaks.

3.2.4. The specimens with spraying aluminum coating on glass cloth and compared with experimental results as shown in Fig. 15
When thickness of the spraying aluminum is 0.1 and 0.2 mm, (a) and (b), respectively. It can be seen that ablation damage region
transient temperature field of the specimen with spraying alu- is more less and distributes mainly as circular area shape, in which
minum coating on glass cloth is obtained by numerical calculation attachment point of impulse electrical current is the center. For the
240 F.S. Wang et al. / Composite Structures 145 (2016) 226–241

Table 5
Damage quantities of benchmark specimens and their protection structures.

Damage type Current peak (kA) Local spraying Full spraying aluminum Spraying aluminum on Benchmark
aluminum glass cloth
0.1 (mm) 0.2 (mm) 0.1 (mm) 0.2 (mm) 0.1 (mm) 0.2 (mm)
Fiber damaged area (mm2) 31.3 332 226 463 239 0 0 1383
88.4 1375 794 3365 852 683 192 7239
93.7 1493 912 3443 1021 731 257 7920
The damaged area (mm2) 31.3 1637 783 1193 482 263 152 3813
88.4 9837 5932 7583 2684 3821 1823 16,372
93.7 11,647 6529 8028 2903 4583 2264 19,738
The maximum damaged thickness (mm) 31.3 0.14 0.12 0.18 0.14 0.08 0.03 0.58
88.4 0.27 0.19 0.35 0.26 0.2 0.12 0.74
93.7 0.28 0.2 0.36 0.29 0.21 0.13 0.75

reason that the insulated glass cloth is pasted at metal fastener cap
and then aluminum layer with a certain thickness is sprayed on
glass cloth surface, electrical current attached to test specimen will
conduct quickly from the attachment region to outward boundary
along aluminum layer surface. When there is a part of electrical
current conducting downward along thickness direction of alu-
minum layer, the pasted glass cloth can play a good role in isolat-
ing electrical current and prevent electrical current from
conducting through glass cloth to metal fastener. Most electrical
charge gathers at glass cloth boundary and makes a certain extent
of ablation damage appear.
In order to analyze lightning protection effect of the specimen
with spraying aluminum coating on glass cloth and compare with
benchmark specimen, temperature distribution of composite lam-
inate with 0.1 and 0.2 mm aluminum layer thickness is shown in
Fig. 15(c) and (d), respectively. It can be seen that most of electrical
current gathers at glass cloth boundary because of high electrical
conductivity for aluminum layer and good insulation performance
for glass cloth. Joule heat generated by electrical current conducts
rapidly to composite plate through aluminum layer and high tem-
perature distribution appears at glass cloth boundary. More electri-
cal current will gather and a little ablation area will appear at glass
cloth boundary when electrical current peak applied is larger. Joule
heat at glass cloth boundary conducts along composite plate sur-
face and finally makes composite plate appear thermal damage
in shape of circular area, although the top fiber direction of com-
posite laminate is 45 degree.
For the IIB lightning zoning with 31.3 kA current peak, Surface
damage area of test specimen with spraying aluminum coating
on glass cloth is smallest, in which no damage of the pasted glass
cloth can be observed at attachment region of impulse electrical
current. It can be seen from test results that glass cloth not only
can prevent metal fastener from contacting with electrical current,
but also can make surface damage area of aluminum layer occur
smaller.
For the IA lightning zoning with 88.4 kA current peak and the IB
lightning zoning with 93.7 kA current peak, a great quantity of
electrical charge aggregates and serious damage appears at glass
cloth boundary due to its insulation with peak increase of impulse
electrical current applied. Ablation damage area of aluminum layer
surface appears still very small although peak value of impulse
electrical current increases. When peak value of impulse electrical
current becomes the maximum, more and more electrical charge
gathers at glass cloth boundary and test specimen appears rela-
tively serious damage at this place.
3.2.5. Damage qualities of benchmark specimens and their protection
structures
When thickness of the spraying aluminum is 0.1 and 0.2 mm,
fiber damaged area, the damaged area and the maximum dam-
aged thickness is given in Table 5 under different electrical cur-
rent peaks for benchmark specimen, the specimen with full
spraying aluminum coating, the specimen with local spraying alu-
minum coating and the specimen with spraying aluminum coat-
ing on glass cloth. It can be seen that three kinds of damage
quantities for benchmark specimens are largest when compared
with their protection structures. Fiber damaged area and the max-
imum damaged thickness for the specimen with local spraying
aluminum are both smaller while the damaged area is larger than
the specimen with full spraying aluminum coating. Because dis-
tance of center point from lateral boundary of specimen length
is nearer and the lateral boundary connects with ground and its
potential is 0, most of electrical current attached to test specimen
with the local spraying aluminum coating will first conduct from
the attachment point to outward boundary along width direction
of aluminum layer rapidly and then release out. Only a small part
of electrical current conducts downwards along thickness direc-
tion of composite plate through the aluminum layer. So, fiber
damaged area and the maximum damaged thickness of composite
plate is smaller for test specimen with local spraying aluminum
protection. On the whole, the specimen with full spraying alu-
minum coating is more excellent than that with local spraying
aluminum coating to anti-lightning strike. Damage qualities for
the specimen with spraying aluminum coating on glass cloth
are smallest when compared with benchmark specimen, the spec-
imen with full spraying aluminum coating and the specimen with
local spraying aluminum coating, which therefore is the most
excellent protection effect on anti-lightning strike. Additionally,
three kinds of damage quantities will increase with the increase
of electrical current peaks.
For above four different types of carbon fiber/epoxy composite
laminates, it can be seen that simulation results are good agree-
ment with experimental results. For specimens with spraying alu-
minum coating with regard to relative electrical current peak, it
also can be seen that aluminum coating has a good effect on
anti-lightning strike. Damage quantities such as fiber damaged
area, the damaged area and the maximum damaged thickness for
0.2 mm aluminum thickness protection are much smaller than
those for 0.1 mm aluminum thickness protection, which matches
with analysis results obtained by Wang et al. in reference [30] that
0.2 mm aluminum thickness protection is most suitable to anti-
lightning strike. So, the greater aluminum layer thickness and more
easily electrical current to conduct along aluminum layer to out-
ward boundary, which makes electrical current release in a short
time through the left and right sides of copper strips placed in test
specimens. At the same time, the greater aluminum layer thickness
and more effective to resist direct effects of impulse current and
prevent composite structure producing serious ablation damage.
But the greater aluminum layer thickness and more heavy aircraft
composite structures.
 F.S. Wang et al. / Composite Structures 145 (2016) 226–241
4. Conclusions
Against the anti-lightning strike background of carbon fiber/
epoxy composite laminates widely used in aircraft structures,
damage research and protection design of composite laminates
suffered from lightning strike was carried out based on the com-
bined methods of test and numerical simulation. The following
main conclusion can be obtained.
1. Ablation damage model proposed in this paper can be accu-
rately used to simulate and assess damage extent of composite
laminate and its protection structures with surface aluminum
layer when subjected to impulse electrical current.
2. Lightning electrical current conducts mainly along top carbon
fiber direction with the maximum electrical conductivity and
finally releases from the edges of benchmark specimen. Fiber
damaged area, the damaged area and the maximum damaged
thickness increases with the increase of current peak value.
3. Specimens with lightning protection methods have the better
anti-lightning effect and can significantly reduce ablation dam-
age of composite material compared with benchmark speci-
mens. The thicker aluminum layer and the better to anti-
lightning strike. Ablation damage area of test specimen with full
spraying aluminum is presented as circular shape and that of
test specimen with local spraying aluminum mainly distributes
along the width direction of aluminum layer.
4. Compared with test specimen with full spraying aluminum,
fiber damaged area and the maximum damaged thickness is
both smaller while the damaged area is larger for test specimen
with local spraying aluminum. On the whole, specimens with
full spraying aluminum protection are more excellent than
those with local spraying aluminum protection to anti-
lightning strike.
5. Fiber damaged area, the damaged area and the maximum dam-
aged thickness of the specimen with spraying aluminum on
glass cloth is the smallest compared with the specimens with
full and local spraying aluminum. The pasted glass cloth can
play a good role in isolating electrical current and prevent elec-
trical current from conducting through glass cloth to metal
fastener.
Acknowledgement
This study is supported by the National Natural Science Founda-
tion (No: 51475369).
References
[1] Deierling PE, Zhupanska OI. Experimental study of high electric current effects
in carbon/epoxy composites [J]. Compos Sci Technol 2011;71(14):1659–64.
[2] Larsson A. The interaction between a lightning flash and an aircraft in flight [J].
CR Phys 2002;3:1423–44.
[3] Mall S, Ouper BL, Fielding JC. Compression strength degradation of
nanocomposites after lightning strike [J]. J Compos Mater 2009;43
(24):2987–3001.
241
[4] Gou JH, Tang Y, Liang F, Zhao ZF, Firsich D, Fielding J. Carbon nanofiber paper
for lightning strike protection of composite materials [J]. Compos B 2010;41
(2):192–8.
[5] Chemartin L, Lalande P, Peyrou B, Chazottes A, Elias PQ, et al. Direct effects of
lightning on aircraft structure: analysis of the thermal, electrical and
mechanical constraints [J]. J Aerosp Lab 2012;5:1–15.
[6] Metwally IA, A-Rahim AA, Heidler F, Zischank W. Computation of transient-
temperature profiles in objects exposed to simulate lightning currents [J]. Int J
Therm Sci 2006;45(7):691–6.
[7] Kawakami H, Feraboli P. Lightning strike damage resistance and tolerance of
scarf-repaired mesh-protected carbon fiber composites [J]. Compos A 2011;42
(9):1247–62.
[8] Feraboli P, Miller M. Damage resistance and tolerance of carbon/epoxy
composite coupons subjected to simulated lightning strike [J]. Compos A
2009;40(6–7):954–67.
[9] Hirano Y, Katsumata S, Iwahori Y, Todoroki A. Artificial lightning testing on
graphite/epoxy composite laminate [J]. Compos A 2010;41(10):1461–70.
[10] Ogasawara T, Hirano Y, Yoshimura A. Coupled thermal–electrical analysis for
carbon fiber/epoxy composites exposed to simulated lightning current [J].
Compos A 2010;41(8):973–81.
[11] Abdelal G, Murphy A. Nonlinear numerical modelling of lightning strike effect
on composite panels with temperature dependent material properties [J].
Compos Struct 2014;109:268–78.
[12] Muñoz R, Delgado S, González C, López-Romano B, Wang DY, et al. Modeling
lightning impact thermo-mechanical damage on composite materials [J]. Appl
Compos Mater 2014;21(1):149–64.
[13] Zhang XY. Coupled simulation of heat transfer and temperature of the
composite rocket nozzle wall [J]. Aerosp Sci Technol 2011;15(5):402–8.
[14] Ogasawara T, Aoki T, Hassan MSA, Mizokami Y, Watanabe N. Ablation behavior
of SiC fiber/carbon matrix composites under simulated atmospheric reentry
conditions [J]. Compos A 2011;42(3):221–8.
[15] Griffis CA, Masumura RA, Chang CI. Response of graphite epoxy composite
subjected to rapid heating [J]. J Compos Mater 1981;15(5):427–42.
[16] Griffis CA, Nemes JA, Stonesifer FR, Chang CI. Degradation in strength of
laminated composites subjected to intense heating and mechanical loading [J].
J Compos Mater 1986;20(3):216–35.
[17] Negarestani R, Sundar M, Sheikh MA, Mativenga P, Li L, et al. Numerical
simulation of laser machining of carbon-fibre-reinforced composites [J]. Proc
Inst Mech Eng Part B: J Eng Manuf 2010;224(7):1017–27.
[18] Zhou B, Liu YJ, Tan SK. Efficient simulation of oxygen cutting using a composite
heat source model [J]. Int J Heat Mass Transfer 2013;57(1):304–11.
[19] SAE-APP-5414A, Aircraft lightning zoning; 2005.
[20] SAR-ARP-5412A, Aircraft lightning environment and related test waveforms;
2005.
[21] SAE-ARP-5416, Aircraft lightning test methods; 2005.
[22] Liu ZQ, Yue ZF, Wang FS, Ji YY. Combining analysis of coupled electrical–
thermal and BLOW-OFF impulse effects on composite laminate induced by
lightning strike [J]. Appl Compos Mater 2015;22:189–207.
[23] Gagné M, Therriault D. Lightning strike protection of composites [J]. Prog
Aerosp Sci 2014;64:1–16.
[24] Affi J, Okazaki H, Yamada M, Fukumoto M. Fabrication of aluminum coating
onto CFRP substrate by cold spray [J]. Mater Trans 2011;52(9):1759–63.
[25] Gragossian A, Tavassoli SH, Shokri B. Laser ablation of aluminum from normal
evaporation to phase explosion [J]. J Appl Phys 2009;105(10):103304.
[26] Ding N, Zhao B, Liu ZQ, Wang FS, Gan J. Simulation of ablation damage of
composite laminates subjected to lightning strike [J]. Acta Aeronaut Astronaut
Sin 2013;34(2):301–8 (in chinese).
[27] Wang FS, Ding N, Liu ZQ, Ji YY, Yue ZF. Ablation damage characteristic and
residual strength prediction of carbon fiber/epoxy composite suffered from
lightning strike [J]. Compos Struct 2014;117:222–33.
[28] Fanucci JP. Thermal response of radiantly heated kevlar and graphite/epoxy
composites [J]. J Compos Mater 1987;21(2):129–39.
[29] Hind S, Robitaille F. Measurement, modeling, and variability of thermal
conductivity for structural polymer composites [J]. Polym Compos 2010;31
(5):847–57.
[30] Wang ZP, Yu G, Hu YL, Li N. Studies on lightning simulation of surface
conductive layer of aircraft composite material [J]. J Civil Aviation Univ China
2011;29(6):22–6 (in Chinese).