Composites Science and Technology 36 (1989) 63-76

Load and Failure Analyses of CFRP Lamk'tates by

Means of Electrical Resistivity Measurements

K. Schulte
DFVLR, Institut fiir Werkstoff-Forschung, 5000 K~ln 90, FRG

&

Ch. Baron
FEA 4, Hills AG, 4370 Marl, FRG

(Received 20 December 1988; accepted 7 February 1989)

ABSTRA CT

The variation of the mechanical stiffness during fatigue loading has been
shown to be an important qualitative and quantitative damage analogue. A
correlation between stiffness reduction and the development of the various
types of damage, such as transverse and longitudinal intraply cracking or
delamination growth, has been made by various authors. However, the in-situ
monitoring offibre fracture occurring during static or fatigue loading in cfrp
laminates has not yet been made except by indirect methods. The
measurement of the variation of the electrical resistivity during loading
promises to be a valuable technique for this purpose.
In the case of a conventional metal sample, the conductivity is essentially
the samefor any direction of current flow through the sample. In cfrp samples,
however, the conductivity is not isotropic and depends on the orientation and
on the conductivity of the carbon fibres. Changes in the conductivity can
therefore be related to fibre fracture. The resistivity also varies with
temperature, and it is therefore necessary to correlate resistivity changes with
the temperature changes observed during a fatigue test. The resistivity of cfrp
laminates is further dependent on the applied load, the fibre volume fraction,
the laminate stacking sequence and the fibre type. Knowing all of these
factors, it is possible to correlate the change in the electrical resistivity with
damage and failure of the load-bearing 0 °fibres. However, the technique can
also be used to monitor continuously the actual load situation in a composite
component.
63
Composites Science and Technology 0266-3538/89/$03.50 © 1989 Elsevier Science Publishers
Ltd, England. Printed in Great Britain
64 K. Schulte, Ch. Baron

1 INTRODUCTION

Carbon fibre reinforced polymers (cfrp) have the advantage of high specific
properties (high elastic modulus and high strength combined with low
density, and they are therefore attractive for aircraft applications. With
increasing use in primary parts, it becomes urgently necessary to detect,
evaluate and understand the development of damage in cfrp components in
order to guarantee safe use over their entire lifetime.
In spite of the fact that a large body of literature already exists on the
mechanical behaviour ofcfrp materials, their fatigue and fracture behaviour
has not yet been sufficiently well evaluated. The fatigue behaviour of fibre
composites is superior to that of metals, despite the fact that failure under
fatigue loading is observed in composites. In metallic materials, the
endurance limit is reached when the maximum load in each fatigue cycle
exceeds some 30% of the static fracture stress. In fibre composites, the
effective endurance limit often reaches about 60-80% of the fracture stress.
As carbon fibres themselves are often said to be insensitive to fatigue
loading, the question remains why damage and failure in fibre reinibrced
materials can be observed under tension-tension loading.
The various damage mechanisms occurring in cfrp laminates under static
and fatigue loading have been described in Refs 1-3. Special attention has
been paid to describing the failure of the load-bearing 0 ° fibres as their
rupture initiates final failure of the test coupon. However, detecting fibre
fracture seems only to be possible by indirect measurements, e.g. stiffness
reduction during loading or directly by using destructive testing techniques,
such as matrix pyrolizing. 1"4
Carbon fibres are electrically conductive. It seems, therefore, to be
advantageous to take the variation of the electrical conductivity (the
increase in the electrical resistance) as an indicator of rupture of fibres in a
cfrp laminate.
In the case of metallic materials, their conductivity is largely independent
of the direction of the electric flux within a specimen. However, in cfrp
laminates the conductivity is not isotropic but depends mainly on fibre
orientation. The specific resistance of carbon fibres is essentially higher than
in metals (i.e. aluminium 2.5 × 10-8[2m). The epoxy matrix, with its high
electrical resistivity ( ~ 1013 f~m), can be taken as an insulator. Fibre fracture
in a composite with continuous fibres will therefore lead to a sudden and
stepwise increase in the electrical resistance, The method itself is, in general,
similar to the potential drop technique used in metals to detect crack
propagation. 5 In a cfrp laminate the electrical resistance is dependent mainly
on the volume fraction, the load level and length of the carbon fibres and the
laminate stacking sequence. Fibre fracture leads to a significant, stepwise,
 Load and failure analyses of cfrp laminates 65

increase in electrical resistivity of the laminate because of the reduction in

the number of electrically conductive fibres.
In carbon fibres, the electrical resistivity decreases with increasing
temperature. 6 Under fatigue loading an increase in the laminate tempera-
ture is observed which is dependent on frequency, load level and laminate
stacking sequence. 7 Under fatigue loading conditions, it is therefore
necessary to detect the temperature variation and correlate it with changes
in resistivity.
If this is taken into consideration, then the variation of the electrical
conductivity can be taken as an indicator of the fracture of load-bearing 0 °
fibres, a technique which was first proposed for static tensile testing and
tension-tension fatigue a and for three-point bend testing. 9

2 E X P E R I M E N T A L DETAILS

Figure 1 shows schematically the arrangement used for the measurement of

electrical resistivity. A constant current of about 50 mA is introduced into
the end faces of a cfrp test specimen. The value of 50mA was arbitrarily
chosen, as it was the lowest constant current that could be fixed at the
current source available. However, no increase in specimen temperature due
to this relatively high electric current could be detected. Only at a current
flow higher than 1 A was any specimen heating observed. Figure 2 shows the
increase in specimen temperature with increasing current. The voltage,
dependent on the current flow of 50 mA, was detected with an accuracy of

Voltmeter
Sfraingauge
/ l
constanf
j couplFe-~
e ' -SOMA
current

/N

i
Fig. 1. Schematic illustration of the test configuration for monitoring the electrical
resistivity, specimen temperature and strain of cfrp specimen.
66 K. Schulte, Ch. Baron

90
*C Laminate [ 0 z,90 2,02,9021 s //
80 MY 720/LY 556/5T3
o/
/
70 /
/
/
~u 60

SO-
/o" Temperature
t,O-
/o /
30-

20 i F r I r p

1 2 3 ~ 5 6 A
Electric current

Fig. 2. Temperature increase of an cfrp cross-ply laminate due to electric current.

about 0.01 mV. With the aid of a reverse voltage source the total voltage was
potentiometrically suppressed to near zero so that the variation in voltage
due to loading of the specimen could be directly plotted and converted, by
Ohm's Law, to a resistivity change.
End tabs of glass fibre reinforced plastic were used to insulate the
specimen from the grips and for load transfer. The electric current was
introduced into the specimen with the aid of a copper plate which extended
beyond the end of the specimen (Fig. 3). Electrical contact between carbon
fibres and the copper plate was made with a colloidal silver compound which
covered both the carbon fibre ends and the copper plate. To increase contact
between the carbon fibres and the silver, the specimen ends were polished
with 3/~m aluminium oxide powder. After polishing, the fibres, which are
harder than the epoxy resin, are left proud of the matrix, allowing better
current transfer. The measurement cables are soldered to the copper plate.

measure
and
current-cat

cimen

J
copper shl

Fig. 3. Introduction of the electric current into the end surface of a cfrp specimen.
 Load and failure analyses of cfrp laminates 67

The temperature dependence of the electrical conductivity was first

measured on an unloaded specimen, together with its dependence on the
specimen cross section, the cross section being successively reduced by
increasing depths of saw cut.
The variation of electrical resistivity with fibre loading was measured on a
screw driven tensile test machine with a constant crosshead speed of
1 mm/min.
The fatigue tests were performed on a servohydraulic test machine, the
specimen temperature being simultaneously measured with a thermocouple
(Fe-Ni). To avoid a heat flux from the test machine (actuator) into the
specimen, the grips were cooled to a constant temperature of about 22°C.
Two extensometers were attached to either side of the specimen to pro-
vide a continuous measurement of the specimen strain in each load
cycle. The variation of the strain rate (in a load controlled test) can be used
directly to derive the variation of the secant modulus of the specimen. This
can be taken as a damage analogue and is often referred to as 'stiffness'
reduction. 1,2
The temperature increase measured due to fatigue loading can also be
correlated with damage development. ~ The fatigue tests were performed in
tension-tension at a constant R ratio of 0"1 (R -- ¢J¢o; o"u = minimum and
~ro = maximum stress in a load cycle) at a frequency of 10 Hz.
The cfrp test specimen had a width of 25 mm and a length of 190 ram. The
thickness was dependent on the laminate stacking sequence, being about
1 mm for the unidirectional laminate [08] and 2ram for the cross-ply
laminate [02,902, 02, 902]s.
The carbon fibres used were the HTA and ST3 fibres from the
Toho-Beslon Company, and two epoxy matrix systems from the
Ciba--Geigy company were used. Their mechanical and electrical properties
are summarized in Table 1.

TABLE 1
Mechanical and Electrical Properties of Fibres and Unrcinforced Resins

Matrix Fibre

M Y720 M Y720/L Y556 HT.4 ST3

Fracture stress era, MPa 61 93 3400 4 300

Elastic-Modulus, GPa 3-29 2-20 238 238
Fracture strain, % 1.87 4-25 1.40 1.80
Specific electrical
resistivity, ~ x lOs ~ 1021 ~ 1021 1"5 1"5
68 K. Schulte, Ch. Baron

3 RESULTS A N D DISCUSSION

3.1 Tests on unloaded specimens

In order to detect the influence of temperature on electrical resistivity, tests

were performed on unloaded specimens where the temperature was varied
between room temperature (~22°C) and a maximum of 72°C. With
increasing temperature, a decrease in the electrical resistivity was observed
(Fig. 4). This behaviour was dependent on the fibre type tested and was more
pronounced for the H T A fibres than for the ST3 fibres. This is probably due
to fibre processing differences. In the ST3 fibre an anisotropic graphite lattice
structure exists with fewer vacancies, which does not permit the transport of
ions and electrons in the same way as occurs in the HTA fibre.
A further interesting observation is that after return to room temperature,
the electrical resistivity remains lower than before. With each temperature
cycle the electrical resistivity decreases further but with a smaller increment,
asymptotically approaching a constant value. One possible reason for this
behaviour is that the heating could lead to an improvement of the current
flow into the specimen ends. But certain laminate and resin effects could also
favour this kind of behaviour, as was corroborated by measurements on
single fibre strands. A temperature increase led to a decrease in the electrical
resistivity. However, after cooling down again the increase in electrical
resistivity was markedly delayed. This leads to the speculative suggestion
that a temperature increase combined with an electric current flow could
cause a reversible change of the lattice structure of the carbon fibres.~°
The effect of the reduction in specimen cross section on the variation of the
electrical resistivity is shown in Fig. 5. A saw cut was machined into the test

100 ~ - T T
I Train= 2 2 ~ ~ ' - ~

~.~
\\ . / ~ L aminate I 0 6 i

~l. i/ ~ - ~L ~ in,ix " electr resistivity

ell ~ ~ ~ / Laminate [ 0 8 ]
.~_ 90 - ~.~.......__~__ M¥720/HTA
~u Tmax = 72 oE
Tmax: 72 °E
Tmin = 22 °E
i d ...... ----R
1000 2 00 rain 3000
Time

Fig. 4. Dependence of the electrical resistivity on time, under variation of specimen

temperature (Tm~x= 72"C, Train = 22"C).
 Load and failure analyses of cfrp laminates 69

MY720/LY5S6/HTA
m~ Lan~naht 108l
600
T=5OmA
Test (2 sew cut)
o/ J~ - ".+."

theoretical value .~ 7 "

3S t
500
.~.. (2.saw cut}
,.--:

~oo
i I I I
loo 90 BO 70 60
Specimen cross section, %

Fig. 5. Dependence of the electrical resistivity on specimen cross section.

piece of a unidirectional laminate [08] and the variation of the electrical

resistivity, dependent on the length of the saw cut (reduction in specimen
cross section) was measured. A saw cut in an unidirectional specimen means
that the electrically conductive fibres are cut through and only the remaining
fibres can carry current. The electrical resistivity must increase in proportion
to the number of severed fibres. This expected variation in the electric
resistivity is represented by the dotted line (theoretical value) in Fig, 4.
However, the test values are essentially lower than those calculated
theoretically.
The reason is that above and below the saw cut the carbon fibres touch
neighbouring fibres, which allows electrical contact and current flow
transverse to the fibre direction, so that cut fibres also contribute to the
current flow. This is schematically shown in Fig. 6. The dependence of the

////Z//I////

._~
"-
g DC pa
r-

resin matrix fibres

¢_

I ! I !

0 0"2 0"4 0"6 0"8 1-0

Fibre volume fraction
Fig. 6. Increase in electrical conductivity transverse to the fibre direction due to fibre
volume fraction. ~
70 K. Schuhe, Ch. Baron

conductivity transyerse to fibre direction has been studied by various

authors.l 1.12 With increasing fibre volume fraction the number of fibre
contacts also increases, resulting in an increased conductivity transverse to
the fibre direction. Because of this fibre contact broken or cut fibres also
contribute to the overall electrical conductivity of a cfrp specimen, with the
result that the decrease in conductivity does not fully reflect fibre fracture.
For the specimen shown in Fig. 5, a second saw cut was introduced about
20 mm above the first. The electrical resistivity increased after 30% of the
specimen cross section was cut through (dotted line in Fig. 5) but before the
second saw cut was equalin length to the first one. For saw cuts of identical
length a shift towards the theoretical value was observed, a further
indication that current flows into neighbouring fibres above and below
the saw cut as a result of fibre contact.
These tests performed on an unloaded specimen clearly show that a
method based on the measurement of the electrical resistivity could, in
principle, be used to monitor fibre fracture in a carbon/epoxy laminate.

3.2 Tests under static loading

Fracture of single carbon fibres is to be expected in cfrp laminates under

static loading. 3 The observation of fibre fracture is in general, not possible
during a load test except when fibres split off the specimen surface and can be
seen with the naked eye. However, the variation in electrical resistivity
during static loading is a clear indicator of fibre fracture, as seen from Fig. 7,
which shows the stress-strain curve of an unidirectional specimen [08]
together with the variation of electrical resistivity with laminate strain. The
electrical resistivity first increases approximately linearly with increasing

2000 - 350
TensiLe Test mC~
MPa
1600 MY720/LY556/ST3 ,,"" "/ -- 3~,2 ~'~~

1200 ..." lllll

i Stress .

800 i . .~ b/Breakage-338 ,~

I+00 - - 33Z,
.__.__..f--.~ ~ Elecfricat Rer;tstivity
0 ~
," First Fibre Fracture
'. . . . I . 330
• I ' i • i , i

0 O-2 0"~ 0-6 0"8 1"0 1"2 1'.~. °/o 1,6

Strain
Fig. 7. Tensile test o f a u n i d i r e c t i o n a l l a m i n a t e , s h o w i n g the stress/strain curve and
d e p e n d e n t electrical resistivity o f a cfrp specimen.
 Load and failure analyses of cfrp laminates 71

fibre strain. This increase in resistivity, R, is expected, because

R=P 1
A
with p=specific electrical resistivity, / = c u r r e n t path length and
A = conductor section. The length 1, increases under load and A decreases.
After about 0.7% strain the electrical resistivity no longer varies linearly,
which can be related to fibre fracture. Above 1"2% strain the electrical
resistivity rises in stepwise fashion to higher values.1"his can be directly
related to fibre fracture. The fibre fracture is partly indicated--but with a
much lower resolution--in the stress-strain curve.
Thus, it appears that, provided fibre fracture has not occurred, the method
of measuring electrical resistivity can be used to monitor the actual load
situation in a composite laminate, if the appropriate calibration:curve can be
obtained. However, once fibre fracture begins to occur, the resistivity reacts
in an extremely sensitive fashion by a stepwise increase.

3.3 Tests under fatigue loading

If cfrp laminates are cyclically loaded, a variety of damage mechanisms can

be observed.1'2 Fibre fracture can be observed comparatively early in a test.
It was shown in previous studies that the variation of specimen stiffness can
be taken as an indicator (analogue) which is highly sensitive to damage,
especially fibre fracture. However, identification of the various damage
mechanisms occurring is not possible with this method.
Under fatigue loading, the temperature in a test coupon can not be
regarded as constant; at test frequencies of 10 Hz, for example, significant
increases in temperature have been observed. 7 The variatioll of the specimen
temperature can also be used as a means of detecting damage in a composite
specimen. Increasing specimen temperature is a result of increasing (matrix)
damage. A good correlation between increase in specimen temperature and
stiffness reduction has been found. 7
In Fig. 8 is shown the result of a fatigue test on a [02,902, 0a, 902]s cross-
ply laminate. Specimen temperature, electrical resistivity and stiffness
reduction (normalized secant modulus) were simultaneously measured and
correlated with fatigue life. During the test, the maximum load in a fatigue
cycle was increased in stepwise fashion. Figure 8 suggests the following:
(1) During the test, a continuous increase in temperature, dependent on
frequency and maximum load level, can be seen.
(2) The electrical resistivity also increases, but at each load step the
resistivity jumps to a higher level.
72 K. Schulte, Ch. Baron

(3) The high stiffness reduction at the beginning of the fatigue test can be
related to transverse crack development in the 90 ° plies.
(4) At the end of the fatigue test, approaching final failure, a strong
stiffness reduction is again observed, accompanied by a rapid
temperature increase. The resistivity jumps in steps to higher levels,
which indicates thatload bearing 0 ° fibres have failed. The onset of
final failure is then initiated.
It was shown in previous papers 1'2 that the final reduction in stiffness can be
related to the failure of a single fibre or even of bundles of fibres. To verify
fibre fracture beyond doubt, the matrix de-ply method 4 (pyrolysis of a
matrix followed by careful scanning electron microscopy of the remaining
carbon fibres) had to be used. Figure 9 shows a scanning electron
micrograph of broken carbon fibres, which were oriented in the loading
direction, stemming from an inner ply of a fatigue specimen, where the test
was interrupted prior to final rupture. Fibre fracture occurred where a
transverse crack reached the load bearing 0 ° ply. 1'2
The method of monitoring the electrical resistivity is thus an appropriate
non-destructive in situ technique for the investigation of fibre fracture.
Figure 10 shows the temperature increase during the entire fatigue life.
The electrical resistivity increased with each step in the fatigue load level. At
the end of the fatigue life, during cycling at the maximum fatigue load level
of O'm,x = 720 MPa, the resistivity of the specimen decreases as a result of the
temperature increase. Erratic increases in resistivity occur only in
association with fibre fracture. The observed decrease in the electrical
resistivity can clearly lead to misinterpretation of the results.
A correction to the resistivity on the basis of temperature measurements is

2s-I MYVZOIHTA ] '

N
~.o- ¢K l
L(zminofe [07,90z,0 z,90]~ t
R=0"I ~ f=lOHz I /.--')
E
zo-( l
~, 320"9 ",, Seco ~f- Modum.us )
0¢-
>_ -./

0 " ~o- /
I

0,1
y>-------
t~
O'm(zx= 680MPo 720MPo
o 1'o 2~ 3'0 ~o s~ 6~ 7~ 80 9~O/o~OO
Fatigue L i f e - ~
Fig. 8. Variation of secant modulus, temperature and electrical resistivity during cyclic
loading (% of fatigue life) of a cross-ply laminate.
 Load and failure analyses of cfrp laminates 73

Fig. 9. SEM-micrograph ofa 0 ° fibre fracture in the vicinity of the tip of a transverse crack.
Matrix pyrolized.

therefore necessary. The variation of the resistivity due tothe temperature

increase of a cross-ply test piece has to be identified (Fig. 10). From this curve
the equivalent resistivity values can be taken to correct the actual resistivity
data of the fatigue test. The corrected values for the electrical resistivity for
another fatigue test are shown in Fig. 11. A continuous increase, or at least a
constancy in the electrical resistivity can be observed, but there is no longer
any decrease.
25~
mQ
.~ 252

c~ 250

~- 2z,a
LaminateWz.902.02, 902) s
MY720/HTA
2A.6 t , I I I

20 30 40 50 °C 60
Temperature
Fig. 1O. Variation of the electrical resistivity due to temperature change in an unloaded
specimen. Curve used for the correction of the temperature dependent electrical resistivity in
Fig. 9.
74 K. Schuhe, Ch. Baron

70
MPa
68
25-
K
250 i .
m~ -~fSecant.Modulus(CorrectedI ~
Electrical Resistivity :1 ......
20- 2/.6-]" .............................................
- ..................." . - L . . ~ . f ' " - "-- /
66 ==
i- /J Temperature Difference ~ L~." ~--J-~
/ !i
o
x 6z, ~ 15- 2Z,2-~ ~,J ' /." ,--_ ",~
J -..~'~_ r - - - l - - ' .............................
6z I - -".I-." ........
10- " Electrical Resistivity
I 6o
238-1 r" MYV20/HTA
~_ -4 i Laminate[0 2. 902. 0 2.90 21 s
58- ~ 5- 23t, 1~ R= *°min
max=0.1. f = 10Hz
oral x =731I"lPa
s6 ~ N8 = 136,740Cyc
0- 230J' ' ~ ~ , i i ~ ~ ~ i
0 10 20 30 z,O SO 60 70 SO 90 % 100
Fatigue Life
Fig. I !. Variation of secant modulus temperature and electrical resistivity during cycling
loading (% of fatigue life). Electrical resistivity corrected for temperature variation.

4 CONCLUSIONS

--Measurement of the electrical resistivity can be used in situ to detect

damage in a carbon/epoxy specimen or component under load. This
has successfully been demonstrated in Figs 5, 7 and 8.
--Dependent on the load applied to a carbon/epoxy part or specimen, the
electrical resistivity varies directly as a result of the actual strain or load
(compare Fig. 7).
- - T h e conclusion to be drawn is therefore that the measurement of
resistivity in a cfrp component is a non-destructive evaluation (NDE)
technique, with which composite parts can be inspected even when they
are in use. The inspection can be continuous or intermittent. Figure 12
gives an example of the application of this technique, e.g. in an aircraft
wing. An electric current is applied to the composite wing through the
load bearing 0 ° fibres. The actual loading of the wing during flight can
be continuously measured and indicated in the cockpit. Permanent
deviations of the resistivity would thus indicate composite damage.
- - I f the electric current is sufficiently high, then the carbon fibres are
resistively heated (Fig. 2). This can be used for processing composite
materials, since the heat can be directly controlled via the electric
current. Composite parts can therefore be internally heated, and heat
flux from outside, which is disturbed and influenced by parts such as
foils, plates, etc., can thus be avoided. It is conceivable that, with this
technique, the composite wing surface can be heated for deicing.
--Resistive heating of carbon fibres can also be used for processing other
 Load and.failure analyses of tfrp laminates 75

W/ing Fibre-

II I
Notification in [ockpif
Fig. 12. Schematic of a system for permanently controlling load and/or failure in a
composite wing structure. A similar set-up can also be used for heating the wing surface.

composite materials, e.g. melt infiltration of carbon fibre laminates,

fabrics or 3-D weaves with thermoplastic matrix materials. The current
heats the fibres and the matrix material melts, wets the fibres and
penetrates the gaps between the fibres.

5 SUMMARY

It has been shown that the variation of the electrical resistivity of cfrp
specimens during tensile and fatigue loading can be taken as a damage
analogue, especially for failure of the load-bearing 0 ° fibres. The method can
be used in situ, as a non-destructive evaluation technique. It is an
appropriate method for continuously monitoring the condition of a cfrp
component.

ACKNOWLEDGEMENT

The authors thank Dipl.-Ing. V. Bachmann and Cand.-Ing. M. Dusy for

valuable discussions and for performing some of the tests. The support by
the Deutsche Forschungsgemeinschaft (DFG) is gratefully acknowledged.

REFERENCES

1. Jamison, R. D., Schulte, K., Reifsnider, K. L. & Stinchcomb, W. W.,

Characterization and analysis of damage mechanisms in tension-tension
fatigue of graphite/epoxy laminates. A S T M - S T P 836, Philadelphia, PA,
1984, pp. 21-55.
76 K. Schulte, Ch. Baron

2. Schulte, K., Damage development under cyclic loading. In Proceedings of the

European Symposium on Damage Development and Failure Process in
Composite Materials. University of Leuven, Leuven, Belgium, 1987. pp. 39-54.
3. Schulte, K., Damage development in composite materials under static loading.
In Proceedings Advancing with Composites. International Conference on
Composite Materials. Centro Materiali Compositi, Naples, Italy, 1988,
pp. 105-13.
4. Freeman, S. M., Characterization of laminar and interlaminar damage in
graphite/epoxy laminates by the deply technique. ASTM-STP 787, Philadel-
phia, PA, 1982, pp. 50-62.
5. Bachmann, V., Trautmann, K. H., Sengebusch, P. & Marissen, R.,
MeBmethoden for den Ril~fortschritt bei Schwingbelastung. In Proc. D VM-AK
"Betrieb,ffbstigkeit'. Deutsche Verband fiir Materialpriifung, Berlin, 1984, pp.
239 63.
6. Schruggs, L. A. & Gajda, W. J., Jr, Low frequency conductivity of unidirectional
graphite/epoxy composite samples. In Proc. IEEE EMC Symposium, Seattle,
WA, USA, 2-4 August 1977, 396-402.
7. Neubert, H., Harig, H. & Schulte, K., Monitoring fatigue induced damage
processes in CFRP by means of thermometric methods: ICCM-VI/ECCM-2.
Elsevier Applied Science, London, 1987, pp. 1.359-1.368.
8. Baron, Ch. & Schulte, K., Elektrische Widerstandsmessung: Eine Methode zur
Bestimmung des Faserversagens in CFK-Laminaten. Z. Materialpri~fung
30 (1988) 367-76.
9. Robinson, D. A., Failure modes in composite materials. USNA-Trident
Scholar project report, No. 148, United States Naval Academy, Annapolis, MD,
1987.
10. Owston, C. N., Electrical properties of single carbon fibres. Brit. J. Appl. Phys.
(3. Phys. D.), 3 (1970) 1615-26.
11. Gajda, W. J., A fundamental study of the electromagnetic properties of
advanced composite materials, RADC-TR-78-158 A059029, 1978.
12. Balta CaUeja, F. J., Bayer, R. K. & Ezquerra, T. A., Electrical conductivity of
polyethylene--carbon-fibrecomposites mixed with carbon black. J. Mater. Sei.,
23 (1988) 1411-15.