You are on page 1of 8

NEW CARBON MATERIALS

Volume 35, Issue 1, Oct 2020


Online English edition of the Chinese language journal

Cite this article as: New Carbon Materials, 2020, 35(1): 42-49 RESEARCH PAPER

The relationship between the mechanical properties and


microstructures of carbon fibers
Mei-ling Wang 1, Wen-feng Bian 1,2
1
Harbin Institute of Technology, Harbin 150001, China;
2
National Engineering Laboratory for Carbon Fiber Preparation and Engineering, Weihai 264200, China

Abstract: The difference of microstructure of carbon fibers is subtle for the carbon fibers with similar mechanical properties. The mechanical
properties and microstructures of six carbon fibers were investigated by a universal material testing machine, X-ray diffraction, small angle X-ray
scattering and Raman spectroscopy to reveal the microstructure-dependent tensile strength of carbon fibers. Results indicate that the tensile strength
increases with decreasing d002 and ID/IG and with increasing Lc for the five of the six carbon fibers. The dispersion of the tensile strength characterized
by the Weibull modulus increases with increasing micropore radius. The Griffith theory over-estimates the tensile strength of the six carbon fibers.
A more accurate formula was proposed based on the experimental tensile strength and microstructure of carbon fibers
using Mathematics & MATLAB as the simulation tools, which takes into account of the tensile strength loss due to imperfections in the outer-
surface and in the inner-surface, defects and the dispersion of fibers, and is validated by T300 data from other researchers.

Key Words: Carbon fiber; XRD; Weibull; Element; Microstructure

1 Introduction

In the engineering applications of carbon fiber resin matrix that researchers have spent much of their time on different
composites (CFRPs), one or more carbon fibers are likely categoiesy of carbon fibers and ignored the existence of subtle
considered to design composites to satisfy the usage performance differences in the same type of a carbon fiber from
requirements. However, the prediction of the composite different manufacturers. Another is that users often need to
mechanical properties remains a challenging issue because of judge and select carbon fibers with similar properties or similar
the complex microstructures of carbon fibers [1–3]. Conventional products from different manufacturers. The influence of the
theories and methods such as the Griffith’s fracture theory [4], microcosmic structure in the use of the macroscopic properties
Peirce’s weak section theory [5] and Weibull’s distribution of carbon fibers can be the basis of judgment.
theory [6], have been proposed to explain and analyze the
mechanical properties of carbon fibers. These theories are
concentrated on the study of the defects or weak sections of
fibers, assuming that the factors affecting the mechanical
properties of carbon fibers are defects or weak sections that
exist in carbon fibers and the tensile strength depends linearly
on the number of defects. Weibull’s theory also assumes that
defects are the main cause of the discrete fiber strength.
The carbon fiber is generally recognized as a biphasic
system composed of turbostratic graphite crystallites and Fig. 1 The carbon fiber crystal structure model and tensile fracture
micropores. The faulty packing of the crystallite bundle creates mechanism. (a) The turbostratic graphite microstructure of the carbon
spaces between the microfiber bundles in the form of pores or fiber, (b) tensile stress causes a basal rupture along the La┴ direction
holes. Fig. 1 shows a crystal structure model of a carbon fiber and cracks to propagate in the La┴ and Lc directions and (c) as the
and the tensile fracture mechanism, the process of damaging is stress continues to increase, a damaging fracture occurs when the
clearly demonstrated as the stress continues to increase. Based fracture size exceeds the critical value [7].
on the above theory, six PAN-based carbon fibers with similar
properties were studied. We chose six PAN-based carbon fibers
with similar properties as the samples for two reasons. One is

Received date: 31 Dec 2019; Revised date: 23 Jan 2020


*Corresponding author. E-mail: bianwf@163.com
Copyright©2019, Institute of Coal Chemistry, Chinese Academy of Sciences. Published by Elsevier Limited. All rights reserved.
DOI: 1016/ S1872-5805(20)60474-7
Mei-ling Wang et al. / New Carbon Materials, 2020, 35(1): 42-49

2 Experimental when m is large. All data of the Weibull modulus is processed


with EXCEL and the relevance of data is analyzed with the
The following six carbon fibers were used in this study. origin 8 software.
They are (i) high-strength PAN-based SY300 carbon fibers
The section morphology of carbon fibers was obtained
prepared in our laboratory and (ii) high-strength PAN-based
with a JSM-6390 scanning electron microscope. The
commercially available T300, HTS40, A42, TC35-3K and
microcrystalline size and micropore radius parameters were
TC35-24K carbon fibers. Among them, the production process
analyzed using a D/max2500PC X-ray diffractometer (Rigaku
of the precursor fibers for T300 and SY300 was the DMSO–
Co.) using Cu Kα radiation (λ = 0.15418 nm) with a 40 kV tube
H2O method, that for HTS40 and A42 was the ZnCl2–H2O
voltage and a 200 mA tube current. The crystallite interlayer
method, and that for TC35-3K and TC35-24K was the DMF–
spacing, grain size, and porosity were detected using a wide-
H2O method.
angle X-ray detector (XRD). The samples were fixed on a
The diameter of the single carbon fibers, df was measured
fibrous attachment to perform equator scanning and the
using a laser scanning microscope (Aolong, A-5XC) before
micropore radius was detected using a small-angle X-ray
testing.
detector (SAXD). The samples were cut into pieces, pulverized
The density was measured using a density gradient tube.
in a mortar, and placed in a glass tank for scanning. The graphite
Carbon tetrachloride (1.596 g/cm3) was used as the light liquid
crystallite interlayer spacing d002 and crystallite thickness Lc
and dibromoethane (2.179 g/cm3) as the heavy liquid.
were calculated by the following equation (4) and (5) [8–9]:
The mechanical properties of monofilaments were tested
using a universal testing machine (Instron, 1122) with a gauge
of 25 mm, in accordance with the ASTM-D3379 standard.
There are no less than 30 valid samples and the invalid samples
are determined based on (1)abnormal curve, (2)breaks at where θ is the scattering angle, λ is the wavelength of the
the end not in the middle of the sample and(3)abnormal data. X-ray and K is the shape factor [10].
For example, two fibers are superimposed to make the data of
The micropore radius (r) was obtained by the Guinier
the sample too huge.
principle. The InI(h)–h2 curve is a straight line when h is small
According to the Weibull weakest connection theory and [11–12]
. The slope −Rg2/3 is obtained by fitting the linear portion
datum statistical methods, a carbon fiber is assumed to be ideal
with the Guinier formula [Fo. (6)], assuming that the micropores
single fracture morphology that has only one type of defect. The
in the carbon fiber were considered spherical. r was calculated
carbon fiber diameter is uniform and has a circular cross-
using the Equation (7). All data of XRD and SAXD were
section. When the stress field of carbon fiber is uniform, the
processed with the jade 5 software.
one-dimensional double parameter Weibull calculation
equations (1) and (2) are as follows [14–16]:

The degree of graphitization was estimated with a HR800


spectrometer (HORIBA Co., France) at room temperature in a
nitrogen atmosphere using a 457 nm Ar laser as a laser source.
R (=ID/IG) was used to evaluate the graphitization of materials.
where P is the survival probability of fibers, F is the
ID is the D band peak intensity of the Raman spectrum at
fracture probability of fibers, L is the effective length of the
1 360 cm−1 and IG is the G band peak intensity of the Raman
standard test paper frame, σ is the fiber strength, σ0 is the
spectrum at 1 580 cm−1 [13]. All Raman data were processed with
normalization factor, m is the form factor (also known as the the Labspec5 software.
Weibull modulus), N is the total number of measured carbon
fiber monofilaments, and R is the number of fragmented 3 Results and discussion
monofilaments at or below a stress level σR.
The logarithm of both sides of Eq. (1) leads to the 3.1 Mechanical properties
following linear equation (3) :
Table 1 lists the tensile strength at a gauge length 25 mm
of six carbon fibers. A42 has the highest tensile strength and
T300 has the lowest tensile strength. In the datum sheet
Thus, m, the discrete strength of carbon fibers, can be provided by the producer, the six carbon fibers have the same
obtained through linear regression from the Weibull plot of average diameter of 7.0 μm. However, the actual diameter
Equation (3). The strength dispersion of carbon fibers is small deviation is very large and the maximum diameter of TC35-3K
Mei-ling Wang et al. / New Carbon Materials, 2020, 35(1): 42-49

is 7.69 μm, and the minimum diameter of TC35-24K is 7.13 μm. smaller the diameter of a carbon fiber, the greater its tensile
Meanwhile, the density is dissimilar. TC35-3K and TC35-24K strength, the smaller the dispersion. As seen from Table 1 and
difference only in terms of the bundles of filaments, but the Fig. 2, TC35-24K has the smallest diameter, but the dispersion
density has a difference of 0.019 g/cm3. The results shown in coefficient of tensile strength is the largest, the Weibull
Table 1 also clearly indicate that the tensile strength of the modulus is the smallest,and the tensile strength ranks the fourth.
carbon fiber exhibits appreciable scattering. The statistical It is completely inconsistent with the theory. To get full
distribution of the fiber strengths can be described by the appreciation of data in Table 1, there is no correlation between
Weibull modulus. Fig. 2 shows the Weibull plots of the six the Weibull modulus, diameter and the tensile strength for
carbon fibers. The Weibull modulus m for SY300, T300, carbon fibers with similar properties. In fact, the diameter of the
HTS40, A42, TC35-3K, and TC35-24K fibers are 6.394 7, carbon fiber cannot be constant when the Equation (1) is used
4.559 7, 6.296, 7.566 5, 5.865 7, and 4.4675, respectively. for calculating Weibull modulus. From this view, the diameter
A42 has the highest Weibull modulus and TC35-24K has the should be related to the dispersion of carbon fibers. But for
lowest. carbon fibers with
The volume effect suggests that the smaller the diameter
of the fiber, the less defects in the fiber. That is to say, the

Table 1 Physical and mechanical properties of carbon fibers.

a
: Producer’s data sheet, the number in () indicates the standard deviations

Fig. 2 Weibull plots of PAN-based carbon fibers at a 25-gauge length: (a) SY300, (b) T300, (c) HTS40, (d) A42,
(e) TC35-3k, and (f) TC35-24K.
reflected in the nanometric, the volume effect is not obvious
similar properties, the difference in diameter is only
Mei-ling Wang et al. / New Carbon Materials, 2020, 35(1): 42-49

and there is no dominant rule. The production process of 3.2 Cross-section morphology
carbon fibers is long, each step can lead to defects, in
As can be seen in Fig. 3, the cross-sections of (a)SY300
addition to the difference in macroscopic physical properties are mainly ellipses, (b) that of T300 include nearly round,
and microstructure. We believe that the tensile strength and elliptic and kidney shapes, (f) that of HTS40 are nearly round,
dispersion of carbon fibers with similar properties are and that of (c)TC35-3k, (d)TC35-24k and (e)A42 are mainly
caused by crack propagation and the differences in defects kidney-shaped. The irregular cross-section makes the diameter
of carbon fiber discrete. But as shown in Table 1, the standard
are related to the differences in microstructures and
deviation of the diameter of carbon fibers with a non-circular
morphologies. cross-section morphology is not necessarily large. The standard
deviation of diameter of A42 is smallest and its cross-section
morphology is uniform.

Fig. 3 The section morphology of carbon fibers.

with the same precursor fiber production process that has a


3.3 X-ray diffraction analysis
higher density and tensile strength has a more perfect
Fig. 4 shows the XRD patterns of six carbon fibers that are microcrystalline structure, e.g. HTS40 and A42, TC35-3K, and
observed via the equatorial scan, from which the crystallite TC35-24K. Although the microcrystalline structure of T300 is
thickness Lc and the crystallite interlayer spacing d002 are more perfect, it is less dense than SY300, which means that the
calculated according to Equations (4) and (5), respectively. The compactness of SY300 is slightly worse, and there are more
literature [8] revealed that Lc increased and d002 decreased with micropores in the microstructure, which reduces its strength.
the increment of the tensile strength, this graphitization
3.4 Raman spectroscopy analysis
behavior is due to the increasing degree of graphitization of the
carbon fibers caused by the molecular mobility and structural Fig. 6 shows the Raman spectra from the surface of the
rearrangement. The main factors that influence the Lc and d002 carbon fibers. The degree of structural disorder (also called the
are the heat treatment temperature. The relationship of the degree of graphitization) of fibers can be characterized by the
tensile strength with Lc and d002 are plotted in Fig. 5 and the ratio of the integrated intensity of the disorder-induced band
results show a slow increase in Lc and decrease in d002 as the (ID(1360 cm-1)) to the Raman-allowed band (IG(1 580 cm-1)),
tensile strength increases except for SY300. The order of this is which describes the degree of disorganized material in the
T 300, TC35-24K, TC35-3K, HTS40, A42. The most obvious carbon fibers, and the formula is as follows:
difference between SY300 and T300 is the difference in density.
Compared with the four other carbon fibers, the carbon fiber
Mei-ling Wang et al. / New Carbon Materials, 2020, 35(1): 42-49

as that in the inner-surface and core.

Fig. 4 X-ray diffraction patterns of carbon


fibers in the equatorial scan.

Fig. 7 Relationship between the tensile strength and the R values of


carbon fibers.

Fig. 5 Relationship of tensile strength with d002 and the


Lc of carbon fibers. Fig. 8 Schematic illustration of the three-layer structure of a carbon
fiber consisting of a core, inner-surface, and outer-surface [18].

3.5 Small-angle X-ray diffraction analysis

According to the Equation (6), the InI(h)–h2 curve is


shown in Fig. 9 and the micropore radius of carbon fibers can
be calculated by the Equation (7). As shown in Fig. 10, for
carbon fibers with the same production process for
precursor fibers, the micropore radius decreases as the tensile
strength increases in a manner that is consistent with the
density changing trend. This can also be used to explain why
the graphitization structure of SY300 is worse than that of T300
Fig. 6 Raman spectra taken from the surface of carbon fibers. yet the tensile strength of SY300 is higher than T300 .
The carbon fibers with a larger micropore size is more
As shown in Fig. 7, there is a decrease in R with increasing likely to form stress concentrations under the action of tensile
the tensile strength except for SY300, the order is T300, TC35- stress,which results in the reduction of the tensile strength. As
24K, TC35-3K, HTS40, and A42, which has the same variation shown in Fig. 11, the micropore radius decreases as the Weibull
trend as d002. The R values of T300 and SY300 are obviously modulus increases and the order is TC35-24K, T300,
much larger than the other four carbon fibers and the tensile TC35-3K, HTS40, SY300, and A42. Thus, it can be concluded
strengths of T300 and SY300 are lower than the other four that the micropore radius is one of the main reasons for the
fibers. As shown in Fig. 8, the measurement of the X-ray difference in the Weibull modulus of carbon fibers with similar
diffraction is the inner-surface and core part of the carbon fibers properties. At the same time, comparing the fiber diameter with
and the Raman spectra is quite limited in the outer-surface part. the micropore diameter, the difference in fiber diameter do not
For the six specimens with similar properties, the degree of directly affect the size of micropores, which is consistent with
graphitization in the outer-surface of carbon fibers is the same the result of 3.1.
Mei-ling Wang et al. / New Carbon Materials, 2020, 35(1): 42-49

suitable method and theory. At present, the most common


theory is the Griffith (Irwin) formula, which gives the
relationship between the tensile strength and crack tip radius:

Where σu is the estimated tensile strength, KIC is the


fracture toughness and taken as the same value 1.25
MPa·m1/2[22], Y is geometric constant for a crack and taken as
0.75 [23] and C is the radius of a sharp-edged crack.
The Griffith,s theory is used to calculate the strength based
on the following two assumptions. (1) Carbon fibers can be
Fig. 9 The relationship between lnI(h) and h2.
modeled as a composite structure composed of microfibers,
microcrystals and micropores [24], so carbon fiber can be
considered to be isotropic. (2) Hemispherical defects are used
in sample making with the assumption that the crack of carbon
fiber is ellipsoidal.
It’s worth noting that carbon fiber is distinctly anisotropic.
Moreover, the diameter of micropores is not the direction of
fracture. As seen from Table 2, the difference between the
tensile strength calculated by the Griffith theory and the actual
tested is huge, the σu-σ of T300 is 2 345 MPa. But, the tensile
strength calculated by the Griffith theory of six carbon fibers
are similar and the difference between the highest A 42 and the
smallest TC35-24K is only 276 MPa. It is considered that the
Fig. 10 The relationship between the tensile strength and the
defect is the most important factor that affects the mechanical
micropore radius of carbon fibers.
properties of carbon fibers. However, defects are not the only
factor, they are also related to the degree of the graphitization
of carbon fibers, its imperfect structure, and its own coefficient
of variation. This is the main reason why the measured data are
much smaller than the theoretical calculated ones.
In summary, the over estimation of the tensile strength of
carbon fibers by the Griffith and Weibull theory can be caused
by the loss of tensile strength by microstructural factors.
According to the experimental results in this paper, we propose
an expression to describe the tensile strength of the carbon
fibers by the equation (9):
σ = 5679 - σ R - σ d - σ r - σ m - χ (9)
Where σ is the actual strength of carbon fibers, 5,679 is
Fig.11 The relationship between the Weibull modulus and the the average tensile strength calculated by Griffith theory of the
micropore radius of carbon fibers. six specimens, σR is the strength loss due to imperfections in the
outer-surface, σd is the strength loss due to imperfections in the
inner-surface, σr is the strength loss due to defects, σm is the
3.6 Presentation of estimation formula of the tensile
strength loss due to the dispersion of fibers and x is other factors
strength of carbon fibers that affect the tensile strength.

It is well known that the tensile strength of carbon fibers Based on the Griffith and Weibull theory and with the
is greatly influenced by the presence of flaws and defects [18-21]. n u m e r i c a l s i m u l a t i o n a n d a n a l y s i s u s i n g Mathemat
In the process of studying the mechanical properties of carbon ics & MATLAB, the equation (9) can be re-expressed as the
fibers, scientists have been trying to find a way to estimate the equation (10):
12.76
tensile strength. However, it has not been found the most 78716
σ = 5679 - 14107 R + 222072d 002 - - 11.48 m - 45579 (10)
r
Mei-ling Wang et al. / New Carbon Materials, 2020, 35(1): 42-49

It can be seen that the equation (10) is a systematic [4] Griffith A A. The phenomena of rupture and flow in solids[J].
estimation model and is not composed of the proportional Philosophical Transactions of the Royal Society of London, 1920,
distribution of each microstructure factor. According to the data 211: 163-198.
of T 300 carbon fibers given by other researchers, R is 1.08 [25], [5] Peirce F T. Tensile tests for cotton yarns, “The weakest link”
d002 is 0.354 5 nm [26], r is 15.4nm [27] and m is 6.13 [14], the Theorems on the strength of long and of composite specimens[J].
estimated tensile strength is 3 369 MPa, which is close to 3 530 Journal of the Textile Institute, 1926, 17(7) : 355-368.
MPa given by Toray and also consistent with the T300 tensile
[6] Weibull W. A statistical theory of the strength of materials[J]. The
strength tested in this paper. It is worth mentioning that 3 530
Royal Swedish Institute for Engineering Research Preceeding,
MPa is usually the tensile strength of multifilaments in carbon
1939, 151: 1-45.
fibers, while the tensile strength of carbon fiber monofilament
[7] Fu H. Carbon Fiber and Graphite Fiber[M]. Chemical Industry
is generally small, which is also consistent with the estimated
Press, 2010: 369.
results. The equation (10) can be used to determine the
correlation between mechanical properties and [8] Fujie L, Haojing W, Linbing X, et al. Effect of microstructure on
microstructure of T300 level carbon fibers.Table 2 Tensile the mechanical properties of PAN-based carbon fihers during
strength estimation for carbon fibers. high-temperature graphitization[J]. Journal of Materials Science,
2008, (43): 4316-4322.
4 Conclusions [9] Cédric S, Jacques L, René P. The tensile behavior of carbon fibers
at high temperatures up to 2400 ℃[J]. Carbon, 2004, 42(4): 715-
For carbon fibers with similar properties, the degree
725.
ofFig. graphitization on their outer-surface is different from that
[10] Ogalev A A, Lin C, Anderson D P, et al. Orientation and
in the inner-surface and the central part, but the graphitization
dimensional changes in mesophase pitch-based carbon fibers[J].
degree on the outer-surface and central part present the same
Carbon, 2002, 40(8): 1309-1319.
change trend with the tensile strength. The higher the degree of
graphitization (the higher the Lc, the smaller the d002 and R), the [11] Guinier A, Foumet G. Small-angle scattering of X-rays[C]. New
higher the tensile strength of carbon fibers. York: Wiely, 1955: 17-65.

The correlation law between the density and tensile [12] Bale H D, Schmidt P W. Small-angle X-ray-scattering
strength is the same as the micropores and tensile strength in investigation of submicroscopic porosity with fractal properties[J].
that that the density and micropore radius decrease as the tensile Physical Review letters, 1984, 53(53): 596-599.
strength increases when carbon fibers are prepared from the [13] Tuinstra F, Koening J L. Raman spectrum of graphite[J]. Journal
same precursor with the same method, such as T300 and SY300, of Chemical Physics, 1970, 53(3): 1126-1130.
TC35-24K and TC35-3K, and HTS40 and A42. [14] Weidong Y. Jiangwei Y. Tensile strength and its variation of PAN-
There is a significant linear correlation between the based carbon fibers. I. Statistical distribution and volume
micropore radius and the Weibull modulus of carbon fibers. The dependence[J]. Journal of Applied Polymer Science, 2007, 101(5):
micropore radius decreases as the Weibull modulus increases. 3175-3182.
Based on the experimental data of the mechanical [15] Natio K. Tensile properties and Weibull modulus of some high-
properties and microcosmic structure, the tensile strength of performance polymeric fibers[J]. Journal of Applied Polymer
carbon fibers can be simulated and analyzed numerically Science, 2013, 128(2): 1185-1192.
using Mathematics & MATLAB to obtain a formula that [16] Natio K, Yang J, Tanaka Y, et al. The effect of gauge length on
matches the real condition through the comparison of T300 tensile strength and weibull modulus of polyacrylonitrile (PAN)-
data from other researchers. and pitch-based carbon fibers[J]. Journal of Material Science,
2012, 47: 632-642.
References
[17] Johnson D J, Tyson C N. Low-angle X-ray diffraction and
[1] Tsai S W, Wu E W. A general theory of strength for anisotropic physical properties of carbon fibres[J]. Journal of Physics d-
materials[J]. Journal of Composite Materials, 1971, 5: 58-80. applied Physics, 1970, 3: 526.
[2] Harlow D G, Phoenix S L. Bounds on the probability of failure of [18] Kobayashi T, Sumiya K, Fukuba Y, et al. Structure heterogeneity
composites[J]. International Journal of Fracture, 1979, 15(4): and stress distribution in carbon fiber monifilament as revealed by
321-336. synchrotron microbeam X-ray scattering and micro-Raman
[3] Jamal E, Franclois T, Raymond G. Review of failure criteria of spectral measurements[J]. Carbon, 2011, 49: 1646-1652.
fibrous composite materials[J]. Polymer Composites, 1996, 56(6): [19] Johnson D J. Structure-property relationships in carbon fibers[J].
183-196. Journal of Physics applied Physics, 1987, 20: 287.
Mei-ling Wang et al. / New Carbon Materials, 2020, 35(1): 42-49

[20] Moreton R, Watt W, Johnson W. Carbon fibres of high strength temperature treatment process[J]. Journal of Materials Science
and high breaking strain[J]. Nature, 1967, 213: 690-691. and Engineering, 2013, 31(6): 803-806.
[21] Cooper G A, Mayer R M, The strength of carbon fibres[J]. Journal [26] Kuniaki H. Fracture toughness of PAN-based carbon fibers
of Material Science, 1971, 6: 60-67. estimated from strength-mirror size relation[J]. Carbon, 2003,
[22] Honjo K. Fracture toughness of PAN-based carbon fibers 41(5): 979-984.
estimates from strength-mirror size relation[J]. Carbon, 41(5): [27] Zetian C. Studies on the structure and micro-mechnical properties
979-984. of carbon fiber by Raman spectrum[D]. Shanghai: Dong Hua
[23] Bansal G K. Effect of flaw shape on strength of ceramics[J]. University, 2011.
Journal of the American Ceramic Society, 1976, 59(1/2): 87-88. [28] Dongfeng L, Haojing W, Fu H. Structure and properties of T300
[24] Wenfeng B, Shuxiang L, Qingtian L. Micro compound method of and T700 carbon fibers[J]. New Carbon Matericals, 2007, 22(1):
elasticity coefficient of carbon fiber[J]. Acta Material Composite 59-64.
Sinica, 2012, 29(1): 212-215. [29] Yi S, Caihong Z, Yao X, et al. Investigation of PAN-based carbon
[25] Chen W, Mengfan W, Daxin L, et al. Relationship between the fiber microstructure by 2D-SAXS[J]. New Carbon Matericals,
density and graphite structure of carbon fiber during high 2009, 24(3): 270-276.

You might also like