Developed a finite element analysis (FEA) procedure for calculating the electric field in the wind turbine blades due to a lightning stepped leader and estimate whether dielectric breakdown occurs
Developed a physics-based model describing thermal interaction between the lightning channel and composite structure
Applied the developed models for evaluation of thermal damage in the non-conductive All-glass Baseline Wind Turbine Blade (SNL 100-00) due to lightning strike
Applied the model to evaluate the thermal ablation damage in the electrically conductive CFRP composite subjected to the lightning strike
Validated the numerical model by comparing the simulation results with reported experimental results
Sheet Pile Wall Design and Construction: A Practical Guide for Civil Engineer...
Modeling of Lightning-induced Thermal Ablation Damage in Anisotropic Composite Materials and Its Application to Wind Turbine Blades
1. Slide 1 of 49
Modeling of Lightning-induced Thermal Ablation
Damage in Anisotropic Composite Materials and
Its Application to Wind Turbine Blades
Yeqing Wang
Thesis Advisor: Prof. Olesya I. Zhupanska
Mechanical Engineering
The University of Iowa
Final Thesis Defense
June 23rd, 2016
2. Slide 2 of 49
• Introduction
• Thesis objectives
• Prediction of lightning-induced electric field in wind turbine blades
• Characterization of lightning-current-induced heat flux
• Modeling of lightning-induced thermal ablation in composite structure
• Prediction of thermal response and ablation in GFRP composite blade
• Prediction of thermal response and ablation in CFRP composite blade and
validation of the model by comparisons with experimental results
• Conclusion
Outline
3. Slide 3 of 49
• Introduction
• Thesis objectives
• Prediction of lightning-induced electric field in wind turbine blades
• Characterization of lightning-current-induced heat flux
• Modeling of lightning-induced thermal ablation in composite structure
• Prediction of thermal response and ablation in GFRP composite blade
• Prediction of thermal response and ablation in CFRP composite blade and
validation of the model by comparisons with experimental results
• Conclusion
Outline
4. Slide 4 of 49
• Common features:
• Reinforcement: glass fibers, carbon fibers
• Matrix: polymer (e.g., epoxy, vinyl ester resin)
• Composite type: unidirectional and woven
laminated fabrics
• Light and strong
Composite Materials for Wind Turbine Blades and
Lightning-induced Damage in Wind Turbine Blade
• Lightning strike accounts for 23.4% wind turbine insurance claims [1]
• Melting, burning, mechanical impact damage
• Dielectric breakdown (non-conductive glass fiber composites)
5. Slide 5 of 49
Lightning Strike Formation and Lightning Strike
Protections for Wind Turbine Blades
• Common lightning strike protections [2]:
• Receptors and down conductors
• Metallic mesh film
• Lightning strike formation:
Lightning
stepped leader
Answering
leader
First return
stroke
receptor
6. Slide 6 of 49
Lightning Current Conduction on Non-conductive
Wind Turbine Blade
(b)
Lightning
leaders
Answering leaders
(a)
Lightning
leaders
GFRP
composites
Down conductor
Receptor
Answering leader
(c)
Lightning
leaders
Answering leaders
Lightning current
(d)
GFRP composites
breakdown
Lightning
leaders
Lightning current
• Non-conductive GFRP wind
blades with the receptor and down
conductor system are still
subjected to lightning strike
damage
• Need to determine whether
dielectric breakdown occurs:
Dielectric breakdown occurs?
Yes No
Puncture and
catastrophic
damage
Direct heat
injection (thermal
damage)
7. Slide 7 of 49
• Introduction
• Thesis objectives
• Prediction of lightning-induced electric field in wind turbine blades
• Characterization of lightning-current-induced heat flux
• Modeling of lightning-induced thermal ablation in composite structure
• Prediction of thermal response and ablation in GFRP composite blade
• Prediction of thermal response and ablation in CFRP composite blade and
validation of the model by comparisons with experimental results
• Conclusion
Outline
8. Slide 8 of 49
Thesis Objectives
1. Predict whether immediate dielectric breakdown occurs in the non-conductive
GFRP composite blades due to lightning strike with FEA. (Results Updated!)
2. Develop a physics-based model and a corresponding computational procedure
with FEA to predict thermal response and thermal ablation damage in the
composite structure due to lightning strike. (Presented at Comp.)
3. Use the developed computational procedure to predict thermal response and
thermal ablation damage in non-conductive GFRP composite blade due to
lightning strike with FEA. (Presented at Comp.)
4. Use the developed computational procedure to predict thermal response and
thermal ablation damage in electrically conductive CFRP composite blade due
to lightning strike with FEA. Prove the effectiveness of the developed numerical
model by comparing the simulation results with reported experimental data.
(New Results!)
9. Slide 9 of 49
• Introduction
• Thesis objectives
• Prediction of lightning-induced electric field in wind turbine blades
• Characterization of lightning-current-induced heat flux
• Modeling of lightning-induced thermal ablation in composite structure
• Prediction of thermal response and ablation in GFRP composite blade
• Prediction of thermal response and ablation in CFRP composite blade and
validation of the model by comparisons with experimental results
• Conclusion
Outline
10. Slide 10 of 49
0 0
0
0 02
( ) 1 ( )
( )
( ), 10,
1
peak
peak
a G z I
H z
I a b
H z z
c d
• Lightning stepped leader: modeled as a vertical line charge [4]
• Non-uniform charge density [4]:
Problem Formulation: Prediction of Electric Fields along
Wind Turbine Blades
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
01000200030004000
Chargedensity(·10-3C/m)
Distancefromtheground(m)
Leaderlength,3750m
d
O
A
B C
4000 m
Not to scale
Cloud voltage V
Ground
Chargedensity,λ
Distancefromleadertip
toground,z0
Structureheight,250m
• Lightning striking distance
according to IEC 61400-24:
0.90
peak1.9 ,d I
11. Slide 11 of 49
FEA in COMSOL
• FEA in COMSOL Multiphysics®
Section n-n:
2.5m
2.5 m
Cylindrical
lightning stepped
leader channel
O
A
B C
n n
Air domain
• Boundary conditions:
• Cloud voltage, V=40 MV, is applied
to the top surface, V=0 is applied to
the bottom surface
• Non-uniform line charge density is
applied to the leader channel
• Open BCs are applied to side faces
FEA Case 1 FEA Case 2
Effects of Receptors
and Down Conductors
Not considered Considered
<Two FEA cases>
12. Slide 12 of 49
Predicted Electric Field Distribution
• Electric field distribution contour plot considering the effects of receptors and down
conductors:
<Electric field distribution: considering receptors and down conductors effects>
13. Slide 13 of 49
Predicted Electric Fields along Wind Turbine Blades
• Excluding the effects of receptors and down conductors will lead to significant
under-prediction of electric fields along wind turbine blades
<Predicted electric field strength: not
including the effects of receptors and
down conductors>
0
0.05
0.1
0.15
0.2
0.25
0 20 40 60 80 100
Magnitudeofelectricfield(·106V/m)
Distance from the root to the tip (m)
OA, LPL I
OB, LPL I
OC, LPL I
O
A
B C
Ipeak=200 kA
<Predicted electric field strength:
including the effects of receptors and
down conductors>
0
2
4
6
8
10
12
0 20 40 60 80 100
Magnitudeofelectricfield(·106V/m)
Distance from the root to the tip (m)
OA, LPL I
OB, LPL I
OC, LPL I
O
A
B CIpeak=200 kA
14. Slide 14 of 49
Dielectric Breakdown Strength of GFRP Composite Wind
Turbine Blade
• Estimate dielectric breakdown strength
of GFRP composites [3]:
Root buildup and spar cap
of the blade are considered
0
20
40
60
80
100
120
0 0.002 0.004 0.006
Breakdownstrength(·106V/m)
Thickness (m)
Experimental data
Fitted curve
1 2
1
,bE c c
t
<Dielectric breakdown strength of the
GFRP composites>
<Sandia 100-meter all-glass baseline
SNL 100-00 wind turbine blade [5]>
15. Slide 15 of 49
Safety Factor
• Excluding the effects of receptors and down conductors will lead to overestimation
of the safety factor
Dielectric breakdown strength
Predicted electric field strength from FEA
• Safety factor =
<Safety factor: not considering the effects
of receptors and down conductors>
< Safety factor: considering the effects of
receptors and down conductors>
0
10
20
30
40
50
60
70
80
0 20 40 60 80 100
Safetyfactor
Distance from the root to the tip (m)
Root buildup
Spar cap
55.39
0
2
4
6
8
10
12
0 20 40 60 80 100
Safetyfactor
Distance from the root to the tip (m)
Root buildup
Spar cap
1.52
16. Slide 16 of 49
Effects of Humidity/Moisture and Service Time
• Humidity/moisture is likely to decrease the dielectric breakdown strength of
GFRP composite by 3 times [6]
• GFRP composite structure becomes instantly conductive after dielectric
breakdown and will lead to substantial Joule heating generation
• The material behavior and damage after dielectric breakdown is not well studied.
In this thesis, we assume dielectric breakdown does not occur in GFRP
composite blade
Not Consider the Effects of
Humidity/Moisture and
Service Time
Consider the Effects of
Humidity/Moisture and
Service Time
Minimum Safety Factor 1.52 0.51
Dielectric Breakdown No Yes
Damage
Direct heat injection
(thermal damage)
Puncture and catastrophic
damage
17. Slide 17 of 49
• Introduction
• Thesis objectives
• Prediction of lightning-induced electric field in wind turbine blades
• Characterization of lightning-current-induced heat flux
• Modeling of lightning-induced thermal ablation in composite structure
• Prediction of thermal response and ablation in GFRP composite blade
• Prediction of thermal response and ablation in CFRP composite blade and
validation of the model by comparisons with experimental results
• Conclusion
Outline
18. Slide 18 of 49
Lightning Channel Radius Expansion
• Lightning channel expansion during initial discharge stage
(Component A pulsed lightning current):
h
a
bz
rO
R(t)
Q(r, t)
Lightning Current
Injected Heat Flux
1/3 1/2
( ) 0.097 , 50 μspeakR t I t t
<Lightning channel radius expansion>
19. Slide 19 of 49
Lightning Current Density
• Experimental measurements indicate that electric arc current density J(r) has
a non-uniform Gaussian-like spatial profile:
2
max ( )( , ) , ( ).cr
J r t e rJ t R t
0
1
2
3
4
5
6
7
0 0.002 0.004 0.006 0.008 0.01
CurrentDensity(·106A/m2)
Radial Distance (m)
Nestor, 1962
Tsai & Eagar, 1985
Lowke & Tanaka, 2006
Chemartin, et al., 2011
Our Model
‒ c is determined by fitting
experimental data
‒ Jmax is determined by the
relationship: the integral of current
density equals to the total current
2 2
2
ln(0.1)/(0.55 ( ))
2 ( )
0 0
( )
( , ) , ( ).r R t
R t
cr
I t
J r t e r R t
re drd
<Current density profile on composite surface>
20. Slide 20 of 49
Lightning-Current-Induced Heat Flux
• Lightning-current-induced heat flux is approximately linear to the lightning
current density:
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
0 0.002 0.004 0.006 0.008 0.01
TotalHeatFlux(107·W/m2)
Radial Distance (m)
Tanaka et al., 2002
Nestor, 1962
Tsai & Eagar, 1985
Lowke & Tanaka, 2006
Our Model
I=150 A
0
2
4
6
8
10
12
14
0 0.002 0.004 0.006 0.008 0.01
TotalHeatFlux(107·W/m2)
Radial Distance (m)
Nestor, 1962
Lowke et al., 2007
Gonzalez et al., 2005
Lago et al., 2004
Our Model
I=200 A
2 2
2
ln(0.1)/(0.55 ( ))
2 ( )
0 0
10 ( )
10 , ( ).r R t
R t
cr
I t
Q J e r R t
re drd
<Amplitude of heat flux due to lightning on the composite material surface>
5
2
10
b
a mat arc anode
k
Q J U T T
J
21. Slide 21 of 49
• Introduction
• Thesis objectives
• Prediction of lightning-induced electric field in wind turbine blades
• Characterization of lightning-current-induced heat flux
• Modeling of lightning-induced thermal ablation in composite structure
• Prediction of thermal response and ablation in GFRP composite blade
• Prediction of thermal response and ablation in CFRP composite blade and
validation of the model by comparisons with experimental results
• Conclusion
Outline
22. Slide 22 of 49
Electric-Thermal Problem Formulation
( ) ( )
T
T T C T
t
k J E
3
0
3
0
( ) ( , ), ( ),
( ) ( , ), ( ),
z
z
V
T J r t r R t
z
T
k T Q r t r R t
z
0. J
, J σ E
h
2a
2b
x
y
z 3
2
1
θO
Layer i
R(t)
I(r, t)
Q(r, t)
4 4
0
( ) , ( )z
z
T
k T T T r R t
z
• Nonlinear electric-thermal governing equations:
• Radiation boundary condition:
• Loading condition:
Glass Fiber
Composite
Carbon Fiber
Composite
Problem Heat transfer
Coupled
electric-thermal
23. Slide 23 of 49
Cause of Thermal Ablation in Composites
Lightning-
induced heat
flux:
High-intensity
Short-duration
Rapid
temperature rise
in composites
Resin starts to
decompose:
~330 °C
Fully
decomposed:
~800 °C
Fiber rapid
sublimation:
thermal ablation
• Assumption: ablation occurs immediately at a fixed temperature
• Fiber sublimation can be extremely fast due to the rapid and substantial
heat generation due to lightning strike
• Many materials (such as ablative heat shield materials) have approximate
fixed ablation temperature [7]
24. Slide 24 of 49
Modeling Thermal Ablation in ABAQUS
• Umeshmotion+ALE Method: Abaqus subroutine Umeshmotion+ALE
Receding surface, T=Tabl
N1 N2
N3N4
x
y
yabl
N1’ N2’
(a)
ξ
η
(-1,-1) (1,-1)
(-1,1) (1,1)
ηabl
N1 N2
N4 N3
N1’ N2’
(b)
Heat Flux
Heat Flux
T=sublimation
temperature
T>sublimation
temperature
T<sublimation
temperature
• Element Deletion Method: Matlab-Abaqus integrated computational procedure
Receding Surface, T=Tabl
Volume of Material Needs to be Removed Theoretically
Element Numerically Removed by ABAQUS
N1 N2
N3N6
yabl
Heat Flux
N4N5
t>t1
(a)
1, 2
1
N N
t ablT T
3, 6
1
N N
t ablT T
1
2
(b)
N1 N2
N3N6
yabl
N4N5
2 1
1, 2 1, 2N N N N
t t ablT T T
2
3, 6N N
t ablT T
t=t2 (t2>t1)
1
2
(c)
N3N6
yabl
N4N5
2
3, 6N N
t ablT T
t=t2 (t2>t1)
2
Ply 1
Ply 2
Interface
25. Slide 25 of 49
Modeling Thermal Ablation in ABAQUS
• Comparison of Plain Heat Transfer method, Umeshmotion+ALE method, and
Element Deletion method:
Umeshmotion+ALE
method
Plain Heat Transfer
method
Element Deletion
method
26. Slide 26 of 49
Summary
Plain Heat
Transfer method
Umeshmotion+ALE
method
Element Deletion
method
Ablation Zone
Tracking
No Yes Yes
Advantage Simple and fast
Fast computation when
ablation is within single
material domain
Applicable for
multiple material
domains and when
Joule heating is
involved
Disadvantage Not accurate
Not applicable for
multiple material
domains and when
Joule heating is
involved
Expensive
computational time
Novelty N/A
Developed user-
subroutine
Umeshmotion
Developed the
entire
computational
procedure
27. Slide 27 of 49
• Introduction
• Thesis objectives
• Prediction of lightning-induced electric field in wind turbine blades
• Characterization of lightning-current-induced heat flux
• Modeling of lightning-induced thermal ablation in composite structure
• Prediction of thermal response and ablation in GFRP composite blade
• Prediction of thermal response and ablation in CFRP composite blade and
validation of the model by comparisons with experimental results
• Conclusion
Outline
29. Slide 29 of 49
Temperature-dependent Material Properties
• Temperature–dependent material properties of the GFRP composites have
been determined using micromechanics considerations based on the
reported experimental data for glass and resin [20]
900
950
1000
1050
1100
1150
1200
1250
25 150 275 400 525 650 775 900
SpecificHeat(J/kg·°C)
Temperature (°C)
VectorPly Unidirectional
Knytex DBM 1708 Biaxial
SNL Triaxial
0.2
0.25
0.3
0.35
0.4
0.45
0.5
0.55
25 150 275 400 525 650 775 900
ThermalConductivity(W/m·°C)
Temperature (°C)
Vectorply Unidirectional
Knytex DBM 1708 Biaxial
SNL Triaxial
Resin starts to
decompose
330 °C
Resin starts to
decompose
330 °C
<In-depth Thermal conductivity> <Specific heat>
30. Slide 30 of 49
Lightning Protection Levels Used in Simulation
LightningCurrent(Nottoscale)
Component C
(Continuing Current)
A C
Component A
(First Return Stroke)
50μs 1 s
Time (Not to scale)
tm
• Lightning Protection Level (IEC 61400):
Pulsed Lightning Current (Component A) Continuing Lightning Current (Component C)
Case
Peak
Current, Ipeak
Rise
Time, tm
Duration
Decay
Constant, k
Action
Integral
Constant
Current
Duration
Charge
Transfer
(kA) (μs) (μs) (1/μs)
(·106
A/m2)
(kA) (s) (Coulombs)
LPL I 200 4 50 0.07 0.339 2 0.6 1200
LPL II 150 4 50 0.07 0.191 1.5 0.6 900
LPL III 100 4 50 0.07 0.085 1 0.6 600
LPL I LPL II LPL III
Peak
Current
200 kA 150 kA 100 kA
Probability
current is
greater
1% 2% 3%
• Assumption: Lightning channel expands
until the end of component A
31. Slide 31 of 49
Thermal Ablation FEA Results in GFRP Composite Blade
Due to Pulsed Lightning Current
LightningCurrent(Nottoscale)
Component C
(Continuing Current)
A C
Component A
(First Return Stroke)
50μs 1 s
Time (Not to scale)
tm
<Ablation depth vs. Time> <Ablation profile>
-0.09
-0.08
-0.07
-0.06
-0.05
-0.04
-0.03
-0.02
-0.01
0
0 10 20 30 40 50
DepthofAblation(·10-3m)
Time (·10-6 s)
LPL I
LPL II
LPL III
1/20 of layer 1
thickness
-0.09
-0.08
-0.07
-0.06
-0.05
-0.04
-0.03
-0.02
-0.01
0
0 5 10 15 20
DepthofAblation(·10-3m)
Radial Distance (·10-3 m)
t=4.0E-6 s, LPL I
t=4.0E-6 s, LPL II
t=4.0E-6 s, LPL III
t=5.0E-5 s, LPL I
t=5.0E-5 s, LPL II
t=5.0E-5 s, LPL III
1/20 of layer 1
thickness
Max. Ablation
Depth (mm)
Max. Surface
Ablation Radius
(mm)
Method
0.08 14 Umeshmotion+ALE
32. Slide 32 of 49
Thermal Ablation FEA Results in GFRP Composite
Blade Due to Continuing Current
LightningCurrent(Nottoscale)
Component C
(Continuing Current)
A C
Component A
(First Return Stroke)
50μs 1 s
Time (Not to scale)
tm
<Ablation depth vs. Time at LPL III>
-4.5
-4
-3.5
-3
-2.5
-2
-1.5
-1
-0.5
0
0 0.1 0.2 0.3 0.4 0.5 0.6
DepthofAblation(·10-3m)
Time (s)
Element Deletion Method
Umeshmotion+ALE Method
Layer 1, [±45]2
Layer 2, [0]2
Material Interface,
z=-1.8 mm
Method
Max. Ablation
Depth (mm)
Max. Surface
Ablation Radius (mm)
Case 1
Umeshmotion+A
LE
Analysis
aborted
Analysis aborted
Case 2 Element Deletion 5.13 27
<Ablation profile at various times, LPL I,
Case 2>
0.03 m
0.014m
Layer 1, [±45]2
Layer 2, [0]2
Layer 3, [0]4
Layer 4, [±45]2
Layer 5, [0]2
Layer 1, [±45]2
33. Slide 33 of 49
• Introduction
• Thesis objectives
• Prediction of lightning-induced electric field in wind turbine blades
• Characterization of lightning-current-induced heat flux
• Modeling of lightning-induced thermal ablation in composite structure
• Prediction of thermal response and ablation in GFRP composite blade
• Prediction of thermal response and ablation in CFRP composite blade and
validation of the model by comparisons with experimental results
• Conclusion
Outline
34. Slide 34 of 49
Carbon Fiber Composite Wind Turbine Blade
• Offshore wind turbines are equipped with long blades for large power generation
• Glass fiber composite cannot meet the requirement of strength at a critical blade
length
<Blade length vs. Blade mass> <Offshore wind turbines>
35. Slide 35 of 49
Challenges: Problem Formulation and FEA Implementation
( ) ( ) J
T
T T C T Q
t
k
0 J Maxwell's equation of
conservation of charge
JQ J E
• Coupling arises from:
JQ J E
( )T
1. Source 1:
2. Source 2:
• Electric-thermal coupling:
3
0
3
0
( ) ( , )
( ) ( , )
z
z
V
T J r t
z
T
k T Q r t
z
4 4
0
( )z
z
T
k T T T
z
• Radiation boundary condition:
• Loading condition:
Glass Fiber
Composite
Carbon Fiber
Composite
FEA step Heat transfer
Coupled electric-
thermal
Loading condition
update after
elements are
deleted
Only heat flux
Both heat flux and
current density
Non-uniform
loading condition
Subroutine DFLUX
No available
subroutine.
Discretized loading
condition used
Computational
time
Fast slow
36. Slide 36 of 49
Temperature-dependent Thermal-physical Properties
• Thermal conductivities of CFRP reported in the existing literature are not
consistent
Temperature,
T (°C)
Density,
ρ
(kg/m3
)
Thermal Conductivity, k (W/m·°C) Specific
Heat,
c
(J/kg·°C)
Longitudinal Transverse
Through-
the-
thickness
25 1.597 46.863 0.682 0.682 1225
330 1.597 28.790 0.407 0.407 2057
360 1.597 27.103 0.378 0.378 3178
375 1.597 26.400 0.368 0.368 5032
500 1.597 18.800 0.241 0.241 4910
525 1.150 17.400 0.223 0.223 3645
573 1.150 14.800 0.176 0.176 1646
815 1.150 12.810 0.155 0.155 1720
1168 1.150 10.790 0.125 0.125 1825
3316 1.150 10.603 0.125 0.125 2510
Temperature, T
(°C)
Density,
ρ
(kg/m3
)
Thermal Conductivity, k (W/m·°C) Specific
Heat,
c
(J/kg·°C)
Longitud
inal Transverse
Through-
the-
thickness
25 1.52 8 0.67 0.67 1065
343 1.52 2.608 0.18 0.18 2100
500 1.10 1.736 0.10 0.10 2100
510 1.10 1.736 0.10 0.10 1700
1000 1.10 1.736 0.10 0.10 1900
3316 1.10 1.736 0.10 0.10 2509
>3316 (load
elements – gas)
1.10 1.015 1.015 0.10 5875
>3316 (unload
elements – gas) 1.10 1.015 1.015 0.10 5875
<Thermal-physical properties reported
by Griffis et al.>
<Thermal-physical properties reported by
Abdelal & Murphy>
37. Slide 37 of 49
Temperature-dependent Electrical Conductivity
• The temperature-dependency can be characterized using the Arrhenius
equation:
1
ln ln ,
2
f
b
E
C
k T
Model Temperature
Activation Energy, ΔE (ev)
Longitudinal Transverse
Through-
the-
thickness
Sauder et al.
25 °C – 330 °C 0.0024 0.0024 0.0024
330 °C – 3316 °C 0.12 0.12 0.12
Takahashi &
Hahn
25 °C – 3316 °C 4.48×10-3 1.19×10-2 8.15×10-3
• The activation energy values reported in the existing literature are not consistent:
<Activation energy values of CFRP reported by Sauder et al. and
Takahashi & Hahn>
38. Slide 38 of 49
Thermal Ablation in Carbon Fiber Composite Blade:
Composite Materials
• Consists of 16 unidirectional Hexcel 8552/AS4 composite lamina layers and 2 woven
fabric Hexcel 8552/AS4 composite lamina layers [21]
• Laminate schedule: [(0/90F)/45/90/-45/0/45/90/-45/0/0/-45/90/45/0/-45/90/45/(0/90F)]
• Total thickness: 2.47 mm
h
2a
2b
x
y
z 3
2
1
θO
Layer i
R(t)
I(r, t)
Q(r, t)
CFRP composite
panel
<NASA CFRP panel installed in test bed [21]>
Laboratory
artificial spark
generator
CFRP Panel
39. Slide 39 of 49
Lightning Current Reported in Experimental study and
Used in Simulation
<Lightning current waveform used in NASA
experiments >
Component D Component B Component C
Experiment
kA ×106 A2·s
Not Applied
A ms Coulombs
20.8 0.220 368 468 172
Simulation Not Applied Not Applied 368 468 172
<Lightning current waveform used in FEA >
Time
(not to scale)
Current
(not to scale)
D B C*
Current Component
Waveforms
Time
(not to scale)
Current
(not to scale)
D B C*
Current Component
Waveforms
40. Slide 40 of 49
Modeling Thermal Ablation in CFRP using ABAQUS:
Literature Review
• These models lack the treatment of material phase transition
• Plain Heat Transfer method:
• temperature is allowed to keep increasing beyond
the ablation temperature during the analysis
• The area where the temperature is above ablation
temperature is considered to be ablated
• Ogasawara method:
• virtual latent heat (1×1011 J/kg) is added to limit the
maximum temperature to ablation temperature
• Abdelal method:
• In-depth electrical conductivity of the CFRP composite laminated panel was
assumed to increase from 2 S/mm to 1×106 S/mm when the temperature >
3316 °C
• prevents the system from absorbing further energy from the lightning current
• Same method used in Muñoz et al.
41. Slide 41 of 49
Modeling Thermal Ablation in CFRP using ABAQUS:
Demonstration of Element Deletion Method
42. Slide 42 of 49
Effects of Electrical and Thermal Conductivity
• In-depth electrical conductivity controls the extending of surface ablation area
• In-depth thermal conductivity controls the extending of the ablation depth
• The effects of in-plane electrical and thermal conductivity are insignificant
FEA Simulation
#
Method
Thermal
Conductivity
(W/mm2)
Max. Surface
Ablation
Radius (mm)
Max. Ablation Depth
(mm)
1 Plain Heat Transfer k3=0.000125 19.86 0.4875
2 Plain Heat Transfer k3=0.000682 17.37 0.91
Difference -12.50% +86.7%
<Effects of in-depth thermal conductivity>
<Effects of in-depth electrical conductivity>
FEA Simulation
#
Method
Electrical
Conductivity
(S/mm)
Max. Surface
Ablation
Radius (mm)
Max. Ablation Depth
(mm)
1 Plain Heat Transfer σ3=8.68×10-5 20.46 0.5525
2 Plain Heat Transfer σ3=8×10-7 38.26 0.6175
Difference +87.0% +11.76%
43. Slide 43 of 49
Thermal Ablation FEA Results in CFRP Composite Panel Due
to Continuing Current
<Ablation depth vs. Time from FEA > <Ablation profile from FEA>
-1.2
-1
-0.8
-0.6
-0.4
-0.2
0
0 0.1 0.2 0.3 0.4 0.5
DepthofAblation(·10-3m)
Time (s)
Layer 1, [(0/90F)]
Layer 2, [45]
Layer 3, [90]
Layer 4, [-45]
Layer 5, [0]
Layer 6, [45]
Layer 7, [90]
Layer 8, [-45]
Layer 9, [0] -1.2
-1
-0.8
-0.6
-0.4
-0.2
0
-50 -30 -10 10 30 50
DepthofAblation(·10-3m)
Position in the x Direction (·10-3 m)
t = 0.480 s
t = 0.150 s
t = 0.300 s
Layer 1,
[(0/90F)]
Layer 2, [45]
Layer 3, [90]
Layer 4, [-45]
Layer 5, [0]
Layer 6, [45]
Layer 7, [90]
Layer 8, [-45]
Layer 9, [0]
Method
Temp-dependent
Electrical
Conductivity
Temp-dependent
Thermal
Conductivity
Max. Ablation
Depth (mm)
Max. Surface
Ablation Radius
(mm)
FEA
Element
Deletion
Takahashi & Hahn Abdelal & Murphy 0.97 21
Experiments N/A 1.016 23
44. Slide 44 of 49
500 mm
200 mm
Validation by Comparison with Experimental Results
<Surface damage: comparison between experimental results (measured
by Ultrasonic C-scan) and FEA results>
Ablated
Zone
Charre
d Zone
Resin
Decomposed
Zone
Temperature
>3316 °C
(elements
are
deleted)
1800~
3316
°C
300~1800 °C
Radius 21 mm 26 mm 40 mm
• Radius of lightning
channel: 40 mm
45. Slide 45 of 49
FEA Case 2
<Ablation depth: comparison between experimental results (measured
by Pulse Echo Unit) and FEA results>
Validation by Comparison with Experimental Results
46. Slide 46 of 49
Conclusion
1. Developed a finite element analysis (FEA) procedure for calculating the
electric field in the wind turbine blades due to a lightning stepped leader and
estimate whether dielectric breakdown occurs
2. Developed a physics-based model describing thermal interaction between the
lightning channel and composite structure
3. Applied the developed models for evaluation of thermal damage in the non-
conductive All-glass Baseline Wind Turbine Blade (SNL 100-00) due to
lightning strike
4. Applied the model to evaluate the thermal ablation damage in the electrically
conductive CFRP composite subjected to the lightning strike
5. Validated the numerical model by comparing the simulation results with
reported experimental results
47. Slide 47 of 49
• Professor Olesya Zhupanska
• Professors Asghar Bhatti, Hongtao Ding, Shaoping Xiao, Kyung K. Choi
• Dr. Crystal Pasiliao, Air Force Research Lab
• Provost Butler
A special thanks to:
Acknowledgements
This work is supported by the National Science Foundation under Grant Number
EPS-1101284. Any opinions, findings, conclusions, or recommendations
expressed in this work are those of the author and do not necessarily reflect the
views of the National Science Foundation
48. Slide 48 of 49
• Publications
‒ Wang, Y., & Zhupanska, O. I. (2015). Lightning Strike Thermal Damage Model for Glass Fiber Reinforced Polymer Matrix
Composites and Its Application to Wind Turbine Blades. Composite Structures, 132, 1182-1191.
‒ Wang Y., Zhupanska O. I. Estimation of the Electric Fields and Dielectric Breakdown in Non-Conductive Wind Turbine Blades
Subjected to a Lightning Stepped Leader. Submitted to Wind Energy, revision submitted.
‒ Wang Y., Zhupanska O. I. On The Modeling of the Thermal Damage in Anisotropic Composite Materials Subjected to
Lightning Strike. in preparation.
‒ Wang Y., Zhupanska O. I. (2016). “Thermal Ablation in Fiber-Reinforced Composite Laminates Subjected to Continuing
Lightning Current”, 2016 SCITECH/AIAA/ASCE/AHS/ASC Structures, Structural Dynamics & Materials Conference, San
Diego, CA. January 4-8, 2016.
‒ Wang Y., Zhupanska O. I. (2015). Lightning-Strike-Induced Heat Transfer in Glass Fiber Polymer Matrix Composite Blades,
American Society for Composites 30th Annual Technical Conference, Michigan State University, East Lansing, Michigan,
September 28-30, 2015.
‒ Wang Y., Zhupanska O. I. (2014). Evaluation of the Thermal Damage in Glass Fiber Polymer-Matrix Composites in Wind
Turbine Blades Subjected to Lightning Strike, American Society for Composites 29th Annual Technical Conference, San
Diego CA, 2014.
• Awards
‒ GPSG Travel Award. University of Iowa, Febuary, 2016.
‒ First Place in Iowa EPSCoR Annual All-Hands Meeting Poster Competition, University of Northern Iowa, July 23, 2013
‒ Second Place, 2014 IWEA (Iowa Wind Energy Association) Conference Research Poster Competition. Des Moines Area
Community College, Ankeny, Iowa, March 12, 2014
Accomplishments
49. Slide 49 of 49
References
[1] GCube Insurance Services, Inc. (2012) Top 5 Wind Energy Claims.
[2] Lewke et al. (2007). 2007 EWEC Europe’s premier wind energy event. Retrieved from:
http://www.ewea.org/ewec2007/allfiles/41_Ewec2007presentation.ppt.
[3] Madsen S. PhD thesis, The Technical University of Denmark; 2006.
[4] Cooray V, Rakov V, Theethayi N. 2007. Journal of Electrostatics 65 (2007) 296–306.
[5] Griffith, D. T., & Ashwill, T. D. (2011). SAND2011-3779.
[6] Hong, T. P., Lesaint, O., & Gonon, P. (2009). IEEE Transactions on, 16(1), 1-10.
[7] Chen, Y. K., Milos, F. S. (2005). AIAA paper, 5064.
[8] Dec, J. A., Braun, R. D., Lamb, B. (2012). Journal of Thermophysics and Heat Transfer, 26(2), 201-212.
[9] Ogasawara, T., Hirano, Y., & Yoshimura, A. (2010). Composites Part A: Applied Science and Manufacturing, 41(8), 973-981.
[10] Muñoz, R. et al. (2014). Applied Composite Materials, 21(1), 149-164.
[11] Abdelal, G., & Murphy, A. (2014). Composite Structures, 109, 268-278.
[12] Braginskii, S. I. (1958).. SOVIET PHYSICS JETP-USSR, 7(6), 1068-1074.
[13] Plooster, M. N. (1971). Physics of Fluids (1958-1988), 14(10), 2124-2133.
[14] Paxton, A. H., Gardner, R. L., & Baker, L. (1986). Physics of Fluids (1958-1988), 29(8), 2736-2741.
[15] Hill, R. D. (1990). Physics of Fluids B: Plasma Physics (1989-1993), 2(12), 3209-3211.
[16] Nestor, O. H. (1962). Journal of applied physics, 33(5), 1638-1648.
[17] Tsai, N. S., & Eagar, T. W. (1985) Metallurgical Transactions B, 16(4), 841-846.
[18] Lowke, J. J., & Tanaka, M. (2006). Journal of Physics D: Applied Physics, 39(16), 3634.
[19] Chemartin, L., Lalande, P., Delalondre, C., Cheron, B., & Lago, F. (2011). Journal of Physics D: Applied Physics, 44(19), 194003.
[20] Lua, J., O'Brien, J., Key, C. T., Wu, Y., & Lattimer, B. Y. (2006). Composites Part A: Applied Science and Manufacturing, 37(7), 1024-1039.
[21] Griffis, C. A., Nemes, J. A., Stonesifer, F. R., & Chang, C. I., Journal of Composite Materials, Vol. 20, No. 3, 1986, pp. 216-235.
[22] Sauder, C., Lamon, J., and Pailler, R., Composites Science and Technology, Vol. 62, No. 4, 2002, pp. 499-504.
[23] Takahashi, K., & Hahn, H. T. (2011). Journal of Composite Materials, 0021998311416683.
[24] Salah, L., Kuruppuarachchige, C., and Salagame, R, National Institute for Aviation Research, Wichita State University, June 2013.
51. Slide 51 of 49
APPENDIX
<Distribution of the temperature and pressure at the attachment
point during the A/2 waveform [25]>
[25] Chemartin, L., Lalande, P., Peyrou, B., Chazottes, A., Elias, P. Q., Delalondre, C., ... & Lago, F. (2012). AerospaceLab, (5), p-1.
• For a typical pulsed lightning current, at t=200 μs, the pressure inside the
lightning channel becomes equal to the ambient pressure. When t=200 μs, r=40
mm.