Professional Documents
Culture Documents
Theory of
structure and mechanics
of fibrous assemblies
Bohuslav Neck
and
Dipayan Das
Cambridge
Oxford
Philadelphia
Preface
vi
Preface
(Therefore, the textile fibrous assemblies are used also as specific technical
materials). To understand the structure of textile fibrous assemblies, it
needs specific methods such as mathematical modeling. At the present
time, it appears that the exact knowledge about textile structure is widely
spreaded, thanks to the computer application and computer-aided design.
A lot of textbooks have dealt with the methods of manufacturing of
textile materials, but there are only a few books available on the exact
formulation of the internal structure of textile materials and the
mathematical modeling on the behavior of textiles. The main aim of this
book is to introduce the theory of structure and mechanics of fibrous
assemblies in order to partially fulfill the shortage of literature in this
field. It includes mainly the original results of the theoretical researches
carried out on the structure and mechanics of fibrous assemblies. We hope
that this book will be used as a textbook in universities and as a special
study material for scientific researchers. Each topic is therefore started
with very basic and simple discussions and gradually continued to the
more sophisticated formulations for specialists only. We have tried to keep
the continuity of logical way of thinking in deriving the relationships
needed for explanation without any discontinuation. The derivation of the
mathematical expressions is provided relatively in a detailed manner so
that the reader, who has less experience in formulation and manipulation
of mathematical expressions, can easily follow the text. (The authors do
not like the idiom The reader can himself easily derive, the so-called
easy derivation may represent a work of one month!). This results in
relatively large number of equations which may cause a repulsive view.
The dimensionless equations are valid in any coherent unit system (for
example, international SI system). To keep the logical continuity of the
text, some special mathematical formulations are given separately in
appendices.
This book should be useful for the university students as well as the
experienced researches. A few topics mentioned in this book can be used
for teaching of the undergraduate and postgraduate students and the other
special and sophisticated topics can be studied by the doctoral students
and the scientific researchers. We would like to remind our dear readers
that any topic of this book cannot be automatically studied and
mechanically processed. The topics of this book should be understood as
a road map only, which would guide us to create our own ideas and own
understanding of the structure and mechanics of fibrous assemblies using
our own mind. We hope that this book will prove to be very useful by the
readers.
We gratefully acknowledge the support received from the Grant Agency
of Czech Republic GAR under project number 106/09/1916 for carrying
Preface
vii
out some of the research themes reported in this book and processing of
the manuscript of this book. We are also thankful to our universities
Technical University of Liberec and Indian Institute of Technology Delhi
for supporting our research work.
Bohuslav Neck
Dipayan Das
A fiber is a coherent and slender entity with sufficiently high length-todiameter ratio, often known as aspect ratio or slenderness ratio. The high
aspect ratio is intrinsically associated with high specific surface area (ratio
of surface area to mass). The high aspect ratio and the associated high
specific surface area impart a unique advantage to the fibers over other
materials in many critical applications.
The fibers are generally procured in the form of large ensembles such
as bales, tows, etc. They are mechanically processed and arranged to
create fibrous assemblies of the first hierarchical level such as sliver,
roving, yarn, nonwoven, etc. and, when necessary, followed by more
complex hierarchical levels such as woven fabric, knitted fabric, braided
fabric, etc. The fibrous assemblies of the first hierarchical level are called
as simple or primary fibrous assemblies; they are produced directly
from fibers. The fibrous assemblies of the second hierarchical level are
known as composed or secondary fibrous assemblies; they are produced
from primary fibrous assemblies. Sometimes, from geometrical point
of view, the fibrous assemblies are divided into three categories: onedimensional or linear fibrous assembly, two-dimensional or planar fibrous
assembly and three-dimensional or spatial fibrous assembly.
This book is dealt with the fibrous assemblies of the first hierarchical
level, otherwise known as simple or primary fibrous assemblies. Each
simple or primary fibrous assembly has its own specific characteristics as
mentioned below.
(1) Constituents of fibrous assembly. The simple or primary fibrous
assembly consists of fibers, and the fibers are either of same type or
of different types.
(2) Geometrical arrangement of fibers. The geometrical arrangement of
fibers in the simple or primary fibrous assembly is characterized by
(a) Packing arrangement of fibers, i.e., how the fibers are packed
inside the fibrous assembly
(b) Directional arrangement of fibers, i.e., how the fibers are
oriented in the fibrous assembly
xi
xii
1
Basic properties of single fibers and fibrous
assemblies
1.1
L N l , m N mf , V N Vf , A N Af . (1.1)
Fiber density (). Using Eq. (1.1), the fiber density
can be expressed as follows
mf Vf m V .
mf
Vf
Af
(1.2)
Table 1.1 shows the density values of some commonly used fibers.
Table 1.1 Fiber density values (according to Goswami et al. [1]).
Fiber
[kgm3]
Cotton
Linen, jute
Wool
Natural silk
Viscose
Acetate
Polyester
Polyamide
Polyacrylonitrile
Polypropylene
1520
1520
1310
1340
1500
1320
1360
1140
1300
910
mf m
.
l
L
(1.3)
Fineness
Micro-fibers
Cotton and compatible manufactured fibers
Wool and compatible manufactured fibers
Carpet fibers, industrial fibers
< 1 dtex
about 1.6 dtex
about 3.5 dtex
> 7 dtex
is, 1 kg of this fiber has a total length L = 5852 km. A normal shirt of
200 g contains a total fiber length L = 1176 km.
The fineness of cotton fibers is usually determined by air flow method.
Using the well-known micronaire instrument, the rate of airflow through
a porous plug of cotton fibers is measured. The measurements from this
instrument are designated as micronaire values tmic in the unit of microgram
per inch. The fineness of cotton fibers typically ranges from 3 micronaire
to 5 micronaire (fine to coarse). The system in which the fineness of fibers
is expressed in terms of mass for a specified unit of length is called direct
system. The most commonly used unit for fiber fineness in direct system
is denier, where the mass is measured in gram and the specified length
is 9000 m. Thus, t[tex] = 0.111t[den].
Notes: Wool is the only fiber whose fineness, in practice, is specified
by its diameter in the unit of micrometer as shown in Table 1.3.
By applying Eqs. (1.2) and (1.3), we get
t
Vf
V
,
l
L
Vf
tl
,
V
tL
,
Vf V t
.
l
L
(1.4)
t s,
s
Vf V t
.
l
L
(1.5)
4s
4t
,
t
d 2
4
. (1.6)
a)
d
d
p
b)
s
d
d 2
.
(1.7)
4
By applying Eqs. (1.5) and (1.7), the fiber volume is expressed as follows
s
d 2
l.
4
Analogously, the volume of a fibrous assembly is given by
Vf sl
(1.8)
d 2
L.
(1.9)
4
Example 1.5: Referring to the values of fiber density given in Table 1.1
and by applying Eq. (1.6), we calculate the diameter of a polyester microfiber of 0.7 dtex fineness is d = 0.0081 mm, the diameter of a cotton fiber
of 1.7 dtex fineness is d = 0.0119 mm, the diameter of a wool fiber of 3.5
dtex fineness is d = 0.0184 mm, and the diameter of a polyester fiber of
7 dtex fineness is d = 0.0256 mm. Generally, the diameter of textile fibers
ranges from 0.005 mm to 0.035 mm.
The fineness of wool fibers is often expressed by the so-called Bradford
fiber fineness scale (for example 60s, 80s, etc.). Hladk [2] reported the
Bradford fiber fineness values and the corresponding equivalent fiber
diameter values and these are shown in Table 1.3. The relationship between
the Bradford fiber fineness value BBrad1 and the equivalent fiber diameter
d is obtained from Table 1.3, and is expressed by
V sL
BBrad
Average value
80s
70s
64s
60s
58s
56s
50s
48s
46s
44s
40s
36s
1.
18.8
19.7
20.7
23.3
24.9
26.4
30.5
32.6
34.0
36.2
38.7
39.7
Minimum value
19.25
20.20
22.00
24.12
25.65
28.45
31.55
33.30
35.10
37.45
39.20
Maximum value
19.25
20.20
22.00
24.12
25.65
28.45
31.55
33.30
35.10
37.45
39.20
The numerical count grade implies that a yarn can be spun using the indicated
grade fiber and has a weight of 453.69 (one pound) for a length of 512m (560 yd)
times the numerical count.
1.544u
0.722v
39.86
18.8
u
1 e
1 e v
d m
,
69.66 0.772 BBrad
u
BBrad 65.88 2 , v BBrad 38.6.
After calculating the value of d, Eq. (1.6) can be used to estimate fiber
fineness.
Fiber aspect ratio (). The fiber length and fiber diameter are frequently
used to characterize the geometry of the fibers. It is then reasonable to
introduce the expression of fiber aspect ratio, which is defined by the ratio
of fiber length l and fiber diameter d as shown below
= l/d.
(1.10)
Cotton
Wool
Flax
Ramie
1500
3000
1250
3000
q
p
1 0,
d
p d
1 q .
(1.11)
The values of q, given by Malinowska [3], depend on fiber crosssectional shape, in other words, on shape factor (q = 0 for cylindrical
fibers only). The value of shape factor q becomes higher when the shape
of fiber cross-section is irregular, i.e., far away from circular shape. Some
typical shape factor values are given in Table 1.5. Note that Morton and
Hearle [4] used another definition of shape factor and the fiber with circular
cross-section showed the shape factor of zero.
Table 1.5 Cross-sectional shape and shape factor
Shape of fiber cross-section
Circle ideal ()
Circle real fiber
Triangle ideal ()
Triangle real fiber
Mature cotton
Irregular saw
q [1]
0
0 to 0.07
0.29
0.09 to 0.12
0.20 to 0.35
> 0.60
(1.12)
Analogously, by using Eqs. (1.11) and (1.12), the surface area of fibers
in a fibrous assembly A can be expressed as follows
A NAf d
1 q Nl d
1 q L.
(1.13)
mf Vf d 2 4 l.
m V d 2 4 L.
(1.14)
(1.15)
Fiber specific surface area is expressed by surface area per unit mass of
fiber. By applying Eqs. (1.12) and (1.14) or (1.13) and (1.15), we get the
following expression for fiber specific surface area
a
Af
A d
1 q l d
1 q L 4
1 q
.
mf m
d
d 2 4 l
d 2 4 L
(1.16)
a
4
1 q
1 q
2
.
4t
t
(1.17)
Af A d
1 q l d
1 q L 4
1 q
.
Vf V
d
d 2 4 l
d 2 4 L
(1.18)
4
1 q
2.
2
1 q
.
4t
t
(1.19)
For example, on the basis of B.E.T. method, the specific surface area of bleached
cotton fiber ranges from 6000 m2kg1 to 8000 m2kg1.
(1.20)
Example 1.7: Let us refer the previous example, where cotton fibers
are of 1.5 dtex fineness, shape factor of 0.28, and density of 1520 kg m3.
By applying Eq. (1.19), we find the fiber surface area per unit volume is
= 456.8 mm1.
The fiber surface area per unit volume is a general geometrical variable
and does not depend on fiber density and is more useful than the fiber
specific surface area a.
According to Eq. (1.18), the inverse of fiber surface area per unit volume
is directly proportional to the equivalent fiber diameter. The value 1/ has
the dimension of fiber length and to some extent, it is a measure of fiber
thickness. Later we show that this is a very useful variable for calculating
the size of pores among fibers.
Notes: Sometimes it is very useful to write the expressions describing
an object or a process without any dimension. Here we introduce a
useful new dimensionless variable A f3/2 /V f. According to Eqs. (1.12),
(1.8), and (1.10), it is true that Af3 2 Vf
d
1 q l
1 2
1 q
32
l d 1 2 1 2
1 q 3 2 1 2 .
32
d 2 4 l
F F
.
t s
(1.21)
The value of stress at which the fiber (or yarn) breaks is called tenacity.
Earlier, the so-called breaking length L = R was defined by the length
required to break the fiber under its own weight. According to Eq. (1.3), the
mass of a fiber of length R is Rt. By using the acceleration due to gravity
g = 9.81 m s2, the weight (gravitation force) is F = Rtg. According to Eq. (1.21),
it is valid that = F/t = Rg, where [cN tex1] = 0.981 R[km]. Thus, the tenacity
in cNtex1 is approximately equal to the breaking length in kilometers. Also,
the unit of force F was earlier expressed by the so-called pond [p], which
10
1.2
Packing density ( ) definition. The cotton wool is so fine and soft product
that it is used for surgical dressings, while during the middle age the wooden
stakes were used as an execution tool. Interestingly both of these materials
are composed of cellulose. This peculiar example is given to show that
the behavior of an ultimate material depends not only on the constituent
material, but also on the compactness of the final product.
Figure 1.3 illustrates a three-dimensional section of a fibrous assembly
of total volume Vc. The volume of the fibers occupied by this section is V,
thus V Vc. The difference between the volumes Vc V expresses the
volume of air present in the three-dimensional plane, i.e., the empty spaces
between fibers.
The fiber compactness is measured by the ratio of the volume occupied
by the fibers to the total volume of the fiber assembly as shown below.
V
,
Vc
0, 1 .
(1.22)
11
Fibrous assemblies
[1]
Linear textiles
Monofilament
Limit structure(*)
Hard twisted silk
Wet spun linen yarn
Combed cotton yarn
Carded cotton yarn
Worsted yarn
Woolen yarn
Cotton roving
Sliver
1
0.907
0.75 to 0.85
about 0.65
0.5 to 0.6
0.38 to 0.55
0.38 to 0.50
0.35 to 0.45
0.10 to 0.20
about 0.03
Other textiles
Woven fabric
Knitted fabric
Cotton wool(**)
Leather (textiles)(**)
0.15 to 0.30
0.10 to 0.20
0.02 to 0.04
0.005 to 0.02
Other materials
Earthenware(**)
Wood(**)
Animal leather(**)
0.20 to 0.23
0.3 to 0.7
0.33 to 0.66
(*) See later. (**) Piller and Trvniek [5] and Piller [6].
12
dVc ab dh Sc dh,
(1.23)
where
N
j 1
(1.24)
j 1
sj S is the total sectional area of all fibers that are present
in the upper wall section. Now we can express the sectional area packing
density using the general definition given in Eq. (1.22) as follows
dV
S dh
S
.
dVc S c dh Sc
(1.25)
It is obvious from the above expression that the packing density can be
expressed as a ratio of the sectional area of all fibers to the total area of
dh
sj
dh
a
(a)
(b)
13
the fibrous assembly including the empty spaces and fibers. The
expression stated in Eq. (1.25) can be considered as the areal interpretation
of packing density.
Such expression is called local packing density and may be used as a
measure of compactness of small areas around the sectional plane (the
upper wall section) of the fibrous assembly. If we assume that the observed
fibrous assembly has the same packing density in all sections, then we can
use Eq. (1.25) as the packing density of the whole assembly. This
assumption is very useful to express the packing density of linear textiles
(for example, yarns, etc.) as the ratio of sectional areas.
Packing density ( ) mass density interpretation. The fibrous assembly
shown in Fig. 1.3 has mass m and total volume Vc. The mass density * of
the fibrous assembly is then given by the fraction m/Vc as shown below
m Vc
m .
(1.26)
The mass m refers to the mass of fiber only (The mass of air and the
mass of adhesives imparted during finishing are not considered). The
volume of these fibers is V. According to Eq. (1.2), fiber mass density is
= m/V, V = m/. Applying Eqs. (1.26) and (1.2) into Eq. (1.22), we find
m
.
m
(1.27)
14
we can estimate the packing density (Earlier authors, e.g. Marschik [7],
used directly * in lieu of packing density).
Example 1.9: An American cotton bale of dimensions 1.6m0.8m0.6m
has a weight of 230 kg. Accordingly, the bale volume is Vc = 0.768 m3, and
the bale density is calculated from Eq. (1.26) as * = 299.5 kgm 3. By
applying the estimated bale density value and the given cotton fiber density
value according to Table 1.1 into Eq. (1.27), we find the packing density
of the bale is = 0.197.
Porosity (permeability)
. The measure of fiber compactness in a
fibrous assembly can also be characterized by the presence of relative
amount of air in the fibrous assembly. For example, the fibrous assembly
shown in Fig. 1.3 has a total volume of Vc including fiber volume V. Then
the volume of air (volume of pores) between fibers in that assembly is
Vp = Vc V.
(1.28)
Vp
Vc
Vc V
V
1 1 .
Vc
Vc
(1.29)
d
d
h
2
2
(a)
Figure 1.5 Hexagonal structure.
(b)
15
considered equal to the packing density of the whole structure. The length
of each side of the triangle is d+h and the height of the triangle is
(d+h)cos30. Hence, the area of the triangle is obtained as follows
Sc
d h
d h cos 30
2
d h 2 .
4
(1.30)
The area occupied by the fibers in the triangular section of the assembly
is equal to the summation of the areas of three equal sectors (shown by the
shaded color in Fig. 1.5b), each is making an angle of 60 to the vertex of
the triangle. This area is given by the following expression
S
d 2
.
8
(1.31)
By substituting Eqs. (1.30) and (1.31) into Eq. (1.25), we get the packing
density of the triangle and thus for the whole structure as shown below
d 2
8
3
2
d h
4
2 3
h
1
d
(1.32)
Limit structure. In the most compact type of fiber arrangement, all fibers
are in contact with each other. This structure is called limit structure.
Theoretically no more fibers can penetrate further into it. In such a limit
structure, the distance between fibers is h = 0 and by applying this into
Eq. (1.32), we obtain the limit of packing density as shown below
lim
2 3
! 0,907.
(1.33)
16
d
d
h
3h
2h
(a)
(b)
(c)
"
2 3
d 2
1
d
2 3
1.5
! 0.403 .
(1.34)
(a)
17
2h
(b)
2 3
d
1
d
1
! 0.227.
2
2
2 3
(1.35)
d
h
18
Limit
Compact or
tight
Intermediate
Loose
Packing density
0.907
from 0.403
to 0.907
from 0.227
to 0.403
less than
0.227
Images
of
border
structures
h=0
h = d/2
h=d
lim = 0.907
= 0.403
= 0.227
1.3
19
m
m,
mi m.
(1.36)
i 1
gi
mi
,
m
mi mgi .
(1.37)
g
1.
(1.38)
i 1
mi
g
m i,
i
i
(1.39)
20
i 1
i 1
V Vi m
gi
.
i
(1.40)
The mean fiber density of the blend is obtained from Eqs. (1.2) and
(1.40), and it is expressed by
m
V
1 n gi
.
1
n
gi
i 1 i
i 1
(1.41)
It is obvious from Eq. (1.41) that the mean value of fiber density of the
blend is the weighted harmonic (not arithmetic!) mean value of the fiber
density values of the individual components of the blend (Remember that
the harmonic mean value is lower than the arithmetic mean value).
Volume fraction (vi). The volume fraction vi of i-th component of the
blend is defined by the ratio between the volume of i-th component of the
blend and the total volume of the blend V. From Eqs. (1.39) to (1.41), we get
V
vi i
V
gi
i
gi
n
g
i
m i
i 1
i
m
n
vi 1 .
i 1
(1.42)
mi
g
m i.
ti
ti
(1.43)
i 1
i 1
L Li m
gi
.
ti
(1.44)
t
m
n
m
i 1
gi
ti
1
.
gi
i 1 ti
21
(1.45)
t
1
gi
i 1 si i
n
1
n
vi
i 1
s
.
vi
i 1 si
n
(1.46)
s
t
1
n
vi
i 1 si
.
(1.47)
22
d 2
4
1
,
4vi
2
i 1 d i
n
d
1
n
vi
2
i 1 d i
(1.48)
gi
L
t
t
% i i n i gi
gi
L
ti
m
i 1 ti
m
n
% i 1 .
i 1
(1.49)
Li
g
m %i
m i
.
li
li ti li t
(1.50)
From the above expression, we can express the total number of fibers
in the blend as follows
n
N Ni
i 1
m n %i
.
t i 1 li
(1.51)
23
The mean fiber length of the blend is obtained from Eqs. (1.1), (1.3)
and (1.51) as shown below
m
L
t
l
N m n %i
t i 1 li
1
.
%i
i 1 li
n
(1.52)
Here again we recognize that the mean fiber length of the blend is a
weighted harmonic mean of the mean fiber length of the individual
components.
Relative frequency (&i). The ratio between the number of fibers in
i-th component and the total number of fibers in the blend is known as the
relative frequency &i. By applying Eq. (1.50) into Eq. (1.52), we find the
following expression for the relative frequency of number of fibers of i-th
component of the blend
m %i
N
l t
l
&i i i n
%i
%i
N m
li
t i 1 li
n
&i 1 .
i 1
(1.53)
Mean fiber aspect ratio (). The mean fiber aspect ratio of the blend is
found directly from Eq. (1.10) by applying the mean fiber length and the
mean equivalent fiber diameter of the blend. It can also be derived from
the fiber aspect ratio of the individual components of the blend. According
to Eq. (1.10), fiber aspect ratio of i-th component of the blend is expressed
by i = li /di and from Eq. (1.53) we find li = %il/&i. By combining both of
these expressions, we find the following expression for the fiber diameter
of i-th component of the blend
di
li
%l
i .
i &i i
(1.54)
The mean fiber aspect ratio of the blend is found from Eq. (1.10) where
= l/d. By rearranging Eqs. (1.48) and (1.54), we find the following
expression for the mean fiber aspect ratio of the blend
l
l
d
vi
l
2
i 1 d i
vi &i2 i2
2 2
i 1 % i l
n
vi &i2 2
2 i .
i 1 % i
n
(1.55)
24
Mean specific surface area (a). From Eq. (1.16), it is valid that the
surface area of i-th component of the blend is Ai = miai. By substituting
this expression into Eq. (1.37), we find the surface area of i-th component
of the blend as follows
Ai = miai = mgiai .
(1.56)
i 1
i 1
A Ai m
gi ai .
(1.57)
a
A n
gi ai .
m i 1
(1.58)
The mean specific surface area of the blend is the summation of the
product of mass fraction and the mean specific surface area of the individual
components of the blend.
Mean specific surface area per unit volume (). The relationship between
the volume and the specific surface area of i-th component of the blend is
given by Eq. (1.20) as ai = i /i. Similarly, for the blend it is valid that
= a. Furthermore, according to Eq. (1.42), it is valid that gi = vii /. By
substituting these expressions into Eq. (1.58), we find
n
n
v n v
a i i i i i , a
vi i ,
i i 1
i 1
i 1
vi i .
i 1
(1.59)
The mean surface area per unit volume of the blend is the arithmetic
mean of the fiber surface area per unit volume of the blend components.
Mean shape factor of fiber cross-section (q). The mean shape factor of
fiber cross-section of the blend can be found directly from Eq. (1.18) by
applying the expressions for mean specific surface area per unit volume
and the mean equivalent fiber diameter of the blend. Of course, it is also
possible to derive this expression from the mean shape factor of fiber crosssection of the individual components of the blend. From Eq. (1.6), the
fineness of the blend is t = (d2/4) and the fineness of i-th component of
2
the blend is ti d i 4 i . The ratio between these two expressions is
obtained as follows
t
d2
2 .
ti di i
25
(1.60)
From Eq. (1.49), we find t/ti = %i /gi and from Eq. (1.42), we find /i =
v i /g i. By substituting both of these expressions into Eq. (1.60) and
rearranging, we get the following expression for the equivalent fiber
diameter of i-th component of the blend
% i d 2 vi
,
gi di2 gi
di2
vi 2
d ,
%i
di
vi
d.
%i
(1.61)
From Eq. (1.18), we obtain 1+q = d/4 for the blend and i = 4(1+qi)/di for
the i-th component of the blend. By substituting the above expression for
the equivalent fiber diameter of i-th component of the blend into Eq. (1.59),
we find the following expression for the mean shape factor of fiber crosssection of the blend
1 q
d n 4
1 qi
d d n
vi i 'vi
(
4
4 i 1
4 i 1
di
d n 4
1 qi n
vi % i
1 qi ,
'vi v % d ((
4 i 1 '
i 1
i
i
n
q vi % i
1 qi 1.
i 1
(1.62)
Fiber surface area fraction ()i). The fiber surface area fraction of i-th
component of the blend )i is defined by the ratio between the mean surface
area of the i-th component of the blend Ai and the surface area of all fibers
of the blend A. This is obtained by applying Eqs. (1.56), (1.57), and (1.58)
as follows
)i
Ai
A
mgi ai
n
m
gi ai
gi
ai
.
a
(1.63)
i 1
Mass fraction (gi). Sometimes, it is required to find out the mass fraction
of i-th component of the blend from the other known parameters.
If the volume fraction of vi of i-th component of the blend is known, we
can use Eqs. (1.42) and (1.38) to estimate the mass fraction of i-th
component of the blend in the following manner
26
v
vii gi,
i i
i 1
gi ,
gi
i 1
vii
n
v
i i
(1.64)
i 1
% i ti t g i t ,
% i ti gi t ,
i 1
gi
i 1
% i ti
% t
i i
(1.65)
i 1
&i li % i l
gi
gi
tl , &i li ti gitl ,
ti
&i li ti
n
&iliti
<
i i i
i 1
tl g i tl ,
i 1
(1.66)
i 1
)i 1 n
1
gi , gi
a
a
a
i 1
i 1
i
n
)i
ai
.
n
)i
i 1 ai
(1.67)
27
Table 1.8 Given and calculated values of the parameters for the individual
components and for the blend.
Parameters
Fineness t
Length l
Density
Fiber shape factor q
Cross-sectional area
s (1.5)
Equivalent diameter
d (1.6)
Aspect ratio (1.10)
Specific surface area
a (1.17)
Surface area/unit
volume (1.20)
Mass fraction g
Volume fraction v
Length fraction %
Frequency fraction &
Surface area fraction )
Dimension
Components
Blend
Cotton
Polyester
Value
Equation number
[dtex]
[mm]
[kg m3]
[1]
[ m2]
1.5
28
1520
0.30
98.7
1.7
40
1360
0.04
125
1.614
33.77
1420
0.143
114
(1.45)
(1.52)
(1.41)
(1.17), or (1.62)
(1.5), or (1.47)
[ m]
11.2
12.6
12.0
(1.6), or (1.48)
[1]
[m2kg1]
2500
305.2
3175
242.5
2814
267.6
(1.10), or (1.55)
(1.58)
[mm1]
463.9
329.7
379.9
(1.59)
[%]
[%]
[%]
[%]
[%]
40
37.4
43.0
51.9
45.6
60
62.6
57.0
48.1
54.4
(1.42)
(1.49)
(1.53)
(1.63)
References
1.
2.
3.
4.
5.
28
29
2
Pores in fibrous assemblies
2.1
Pores between fibers. The porosity expresses the relative volume of air
present in a fibrous assembly. It is determined as one minus packing density
of the fibrous assembly see Eq. (29). Nevertheless, a given volume of
air can be concentrated in a few relatively big spaces or the same volume
of air can be distributed in a relatively higher number of very small
channels. (One can imagine a homogenous block of polyester bored with
some big holes, although its porosity can be comparable with a polyester
fabric but its fluid flow behavior would be definitely different). The porosity
and the size of air gaps in a fibrous assembly are very important factors to
decide the fluid flow and filtration behaviors of the fibrous assembly.
It is not an easy task to determine the size and shape of pores
experimentally. It is possible to analyze pore characteristics using some
direct experimental methods. These methods are generally based on the
evaluation of microscopic sections of the fibrous assembly using the image
analysis technique. Some indirect methods are also frequently used. The
principle of these methods is based on capillarity and fluid flow phenomena
of fibrous porous materials. The so-called mercury porosimetry or liquid
porosimetry works on the principle of capillarity [13]. Other methods
are based on the application of Kozeny-Carmans flow equation [46].
(The well-known fiber micronaire measuring apparatus is based on the
principle of air-flow method and this apparatus is calibrated to directly
give fiber fineness expressed in g/in). The pore size can also be evaluated
indirectly from the size of particles passed through the fibrous porous
material. The experimental method of aerosol filtration is given in ISO
12956 [7]. In textile practice, it is easier to estimate the pore characteristics
from theoretical models rather than from experimental methods. A few
theoretical models [811] based on Poissionian polyhedra theory was
developed to estimate the pore characteristics in nonwoven fibrous
materials, however, a more generalized theory [12] to predict the pore
characteristics of fibrous materials was also reported.
29
30
sp
(a)
(b)
4 sp
d p2
s
p
.
4
(2.1)
The total perimeter of pore is composed of two parts: the real part, i.e.,
the common border between pore and real fiber, and the fictive part i.e.,
the distinguishing lines between pores. Let us now introduce a variable
called the perimeter of pore pp, which is defined in contrast to fiber by
the total length of real borders only, because the fictive borders do not
31
pp d p
1 qp .
(2.2)
The pore shape factor can take a negative value, as it is evident from
the definition of pore perimeter.
Although in a real fibrous assembly, the pore characteristics vary widely,
but the variability of pore characteristics is not considered in the following
discussion. We assume that all pores have the same characteristics. The
variables and expressions corresponding to all fibers in a fibrous assembly
as mentioned in Chapter 1 are valid in a similar way for all pores in a
fibrous assembly. These characteristics can be regarded as the average of
all pores in a fibrous assembly.
Using the expressions derived earlier for fibers we can obtain the
corresponding expressions for pores (air fibers). If we denote the total
pore length in a fibrous assembly by Lp, then with an analogy to Eq. (1.9),
we can find the following expression for the total pore volume in a fibrous
assembly
Vp sp Lp
d p2
4
Lp .
(2.3)
Similarly, using Eq. (1.13), the total pore surface area A p can be
expressed as follows
Ap pp Lp d p
1 qp Lp .
(2.4)
The pore surface area per unit volume p can be expressed with analogy
to Eq. (1.18) as follows
p
Ap
Vp
4
1 qp
dp
(2.5)
(The same expression can also be obtained by applying Eqs. (2.3) and
(2.4) into Eq. (2.5) that defines the pore surface area per unit volume as
= Ap/V p).
Relationship between fiber and pore characteristics. The area of real
borders in a fibrous assembly defines the surface area of pores. A major
32
(2.6)
The pore surface area per unit volume expressed in Eq. (2.5) can be
rearranged using Eqs. (2.6), (1.18), (1.22) and (1.29) as shown below
p
Ap
Vp
1
A A V Vc
.
Vp V Vc Vp
1
(2.7)
Substituting Eqs. (2.5) and (1.18) into Eq. (2.7), we can obtain the
following expression for equivalent pore diameter dp
4 1 qp
dp
4
1 q
d
1 qp 1
d p 1 qp 1
d,
, dp
. (2.8)
1
1 q
1 q
d
Further, substituting Eqs. (1.13) and (1.18) into Eq. (2.6) and using
Eq. (2.8) we can find the following expression for the total pore length Lp
d p 1 qp Lp d
1 q L,
2
1 q
Lp
L,
1 qp 1
1 qp 1
1 q
d 1 qp Lp d
1 q L,
L
1 q 2
p
.
L 1 qp 1
(2.9)
The equivalent pore diameter represents the (mean) size of air gaps
among fibers and the total length of pores represents the quantity of pores
(air tubes) in a fibrous assembly. Eqs. (2.8) and (2.9) express these
quantities as functions of parameters q, and d or L. Nevertheless, there
is also an unknown quantity, i.e., pore shape factor qp whose value depends
33
d p2
1 2
L
d L.
4
1
This expression shows that the total volume of pores Vp also depends
only on the fiber characteristics. Further, substituting Eq. (2.8) into Eq.
(2.5) and then applying Eq. (1.18) we obtain
p
4 1 qp
dp
4 1 q 1 q 1 4
1 q
1 q 1 d 1 d
p
(2.10)
p
.
,
1
1
It shows that the pore surface area per unit volume p too is a function
of the fiber quantities only.
34
2.2
qp qp 0.
(2.11)
Note: The quantities related to the conventional pore are displayed with
the superscripted character *.
d p
1 1
1 1
d,
1 q
d 1 q
(2.12)
(2.13)
35
q0
0.25
d p d
0.5
Lp L
p
0.2
0.4
0.6
0.8
2.4 Graph of conventional pore diameter, conventional pore length, and pore surface
area per unit volume as a function of packing density .
d p d p
1 qp .
(2.14)
The behavior of Eqs. (2.12), (2.13) and (2.10), are graphically shown
in Fig. 2.4 for different values of q. (The dotted curve of p/ is identical
to the curve of Lp L when q = 0. This can also be obtained from Eqs. (2.13)
and (2.10)).
The conventional pore diameter (and/or ratio d p d ), i.e., a measure of
size of gaps, is decreasing with the increase in packing density. On the
other hand, the total length of conventional pores (and/or ratio Lp L ) is
increasing with the increase in packing density. (The earlier big spaces
are converted into many small pores with the increase of packing density).
Pores with constant shape factor (variant I). Sometimes it is reasonable
to consider that the pore shape factor qp is independent of packing density
of fibrous assembly, i.e., it is a constant. Then we can write
1+qp = kconstant.
(2.15)
Substituting Eq. (2.15) into Eq. (2.8), the equivalent pore diameter can
be obtained as follows
36
dp
dp
k 1
k 1
d,
,
1 q
d 1 q
(2.16)
and from Eq. (2.9), the total pore length can be expressed as follows
Lp
1 q 2
k
Lp
1 q 2
L,
k2 1
1
L
.
(2.17)
1 qp k
where k L 1 q constant.
,
Lp
1
(2.18)
Note: According to the last expression the pore shape factor q p is not a
constant now, but it increases with the increase in packing density. That
means the shape of pore cross-section is more deviating from circularity.
Substituting Eq. (2.18) into Eq. (2.9), the total pore length is obtained
as follows
Lp
1 q
k2
L,
Lp
1 q 2
, constant.
k2
L
(2.19)
Substituting Eq. (2.18) into Eq. (2.8), the equivalent pore diameter is
obtained as follows
dp
dp
k
k
k
1
1
1
d
d,
. (2.20)
1 q 1
1 q
d 1 q
37
(2.21)
1 qp 1
1
k
1
1
d
d , 1 qp
k
,
1 q
1 q
1 a
(2.22)
qp k
1.
1
By applying Eq. (2.22) into Eq. (2.9), the following expression for the
total pore length is obtained
Lp
1 q 2
1 a
k
1
1 q
L
1
k 2 1
2
2 a 1
2
Lp L
1 q k 2
.
1
2 a 1
L,
(2.23)
38
dp 1 q
1 qp
a 0.5
a 0.75
a 1
0.5
a 1
a 0.75
a 0.5
0.5
(a)
(b)
2
Lp
k2
L
1 q 2
1
a 0.5
a 0.75
a 1
0
0.5
(c)
2.5 Plots of the effect of packing density on the values of d p, q p, and L p for different
values of parameter a . (a) Equation (2.21), (b) Equation (2.22), (c) Equation (2.23).
1 layer
0.00735
d p 0.3095mm
0.01471
d p 0.2338mm
0.02206
d p 0.1920mm
2.3
0.3
dp
[mm]
0.2
0.1
39
k 1.52,
a 0.43
(var. III)
k 1.12,
a 0.5
(var. II)
0.01 0.02
40
12
2
1
-
23
13
3
at the common point of all these three media. A liquid column will be pulled
up by the surface tension 12 .13, which is in equilibrium with the vertical
component of the surface tension 23 at the common point, i.e., 12 13 =
23 cos-. The value of all these surface tensions and the magnitude of angle
- are constant for the given three media. The entire (real) pore perimeter pp
(see Equation (2.2)) pulls up the liquid column due to the force
pp
12 13 pp 23 cos - d p
1 qp 23 cos - . This force must be in
sp h
d p2 4 h ; the mass of the liquid column is d p2 4 h 3 . From the
cos -
1 q 4
d p 1 qp 23 cos -
h
4 23
g3
dp
cos - 1
,
g3
d p
23
d p2
4
h3 g ,
h
(2.24)
4 23 cos - 1 q
.
g3
1 d
According to the above expression, we can state that the height of liquid
column is indirectly proportional to the magnitude of the conventional
pore diameter.
Example 2.2: We consider a shirt of 180g mass prepared by using cotton
fibers of 1.5 dtex fineness, 0.32 shape factor, and 1520 kg m3 density.
The equivalent fiber diameter according to Eq. (1.6) is d = 0.112 mm. Let
the packing density of yarn is = 0.42. The conventional pore diameter,
41
h
4 23 cos - 1
4 cos - 1 q
, h 23
.
g3
kdp
g3
1 kd
(2.25)
1 a
Then, we obtain the following expression for estimating the height of liquid
column
h
4 23 cos -
1
g3
1 a
1
4 cos -
1 q
, h 23
.
1 kd
kd p
g3
(2.26)
p1
b)
p2
42
Q1
d p4 /p
1280 H
(2.27)
(2.28)
The volume of fluid flowing through the porous material per unit time
is
Q Q1 np
G
1 d p2 /p
.
320
H
(2.29)
By substituting the expression for fiber surface area per unit volume
and applying Eq. (1.18) into Eq. (2.21) for equivalent pore diameter (variant
III), we obtain
a
dp
4k
1
.
(2.30)
By applying Eq. (2.30) into Eq. (2.29) and rearranging, we find the
following expression for Q
k 2 A /p
1
Q
2a
2 20 H
2 a 1
(2.31)
43
factor, and 1520 kg m3 density, is tested for its air flow behavior. The
pressure drop across the fabric is maintained at 125 Pa and the crosssectional area of the fabric under the test is 38.30 cm2. The equivalent
fiber diameter according to Eq. (1.6) is d = 0.112 mm. The packing density
of the fabric, according to Eq. (1.27), is = 0.1974. The conventional
pore diameter, according to Eq. (2.12), is d p 0.0370 mm . Let the dynamic
viscosity of air be 0 = 1.78 105 Pa s. Then, the volumetric rate of air
flow through the fabric is calculated by us ing Eq. (2.29) as
Q = 0.001847 m3 s1.
(a)
(b)
Eq. (2.20), is
dp
1 1
d.
2
(2.32)
44
2.4
1.
References
Jena, A., and Gupta, K. (2003). Liquid extrusion techniques for pore structure
evaluation of nonwovens, International Nonwovens Journal, 4553.
2. Jena, A., and Gupta, K. (2002). Measurement of pore volume and flow through
porous materials, Materialprfung, 44, 243.
3. Miller, B., and Tyomkin, I. (1994). Liquid porosimetry: New methodology and
applications, Journal of Colloid and Interface Science, 162, 163170.
4. Kozeny, J. (1927). Ueber kapillare Leitung des Wassers im Boden, Sitzungsber
Akad. Wiss., Wien, 136, 271306 (German).
5. Carman, P. C. (1937) Fluid flow through granular beds, Transactions, Institute
of Chemical Engineers, London, 15, 150166.
6. Carman, P. C. (2010) Flow of gasses through porous media, Butterworths,
London.
7. International Organization for Standardization (2010). Geotextiles and
geotextile-related products determination of the characteristic opening size,
ISO 12956.
8. Faure, Y. H., Gourc, J. P., Millot, F., and Sunjoto, S. (1986). Theoretical and
experimental determination of the filtration opening size, Proceedings of the
Third International Conference on Geotextiles, Vienna, 12751280.
9. Faure, Y. H., Gourc, J. P., and Gendrin, P. (1990). Structural study of porometry
and filtration opening size of geotextiles, In Geosynthetics: Microstructure and
Performance (ed. Peggs, I. D.), 1076, 102119.
10. Lombard, G., Rollin A., and Wolff, C. (1989). Theoretical and experimental
opening sizes of heat-bonded geotextiles, Textile Research Journal, 59, 208217.
12.
45
46
3
Arrangement of fibers fiber orientation
3.1
46
47
expensive technologies available today that are being used for layering of
textile layers in defined directions (for example, tri-axial fabric
manufacturing technology, which was developed during Apollo shuttle
program). Recent advances in electrospinning especially in the design of
web collection devices made it possible to create nano-fibrous scaffolds
with controlled fibre orientation [5].
The natural vegetable and animal tissues also utilize the advantage
of fiber orientation. The longitudinal orientation of grass or straw imparts
more mechanical resistance against the stresses created by wind. The
structure of bone is oriented in the most frequent stressing direction. From
mechanical point of view, the spider web is found to have almost optimum
network structure.
Concept of orientation vector. Figure 3.1 illustrates the curvature of a
single fiber, which varies widely in a fibrous assembly; thus the term
orientation is not clear, but it can be intuitively thought. Therefore, the
term orientation is required to be exactly defined.
B
i
A
AB
48
x3
r 1
2
x2
x1
-i
(3.1)
5 w
-, 2 d- d2 55
-
0, 2 ,
2
0, 2
w
-, 2 d- d2 1.
49
(3.2)
Let us remember the logical sense of the expression w(-, 2) d- d2. This
expresses the relative frequency and/or probability1 that an orientation
vector falls in an elementary class of vectors, where the first angle lies
in an interval ranging from - to - + d-, and the second angle lies in another
interval ranging from 2 to 2 + d2 see the shaded area EFGH shown in
Fig. 3.3.
x3
G
H
-
F
E
d-
x2
x1
2
C
B
d2
2
5 w
-, 2 d2.
(3.3)
1.
2
u
- d- 1 must be valid.
50
OA OD OH sin - sin -, 6
HG AD OA d2 sin - d2, 7
77
HE OH d- d-,
8.
and the elementary area 7
7
EFGH HG * HE sin - d2 d- 79
(3.4)
(3.5)
1
2
2
1
(3.6)
2
k
sin
d
d
k
sin
55
5
5 d2 d- 2k , k 2 .
- 0, 2 ,
0
0
2 0,2
sin .
2
(3.7)
51
The corresponding distribution (marginal distribution) of the angle can be obtained by using Eqs. (3.3) and (3.7) as follows
u
-
2
5
0
3.2
sin d2 sin -.
2
(3.8)
x3
c
c B
:l
;
A
2
e
b
x1
a
Modeling concept. Let us think about the distribution of very short fiber
segments, so that they have a straight shape as short abscissas. A prism of
dimensions a, b, c shown in Fig. 3.4 represents either a fibrous assembly or
a part of the fibrous assembly. The dotted sectional plane , perpendicular
to x3-axis2 , can cut a large number of fiber segments of very short constant
length. One such fiber segment of length :l is shown in Fig. 3.4. The
directions of the fiber segments are described by the angles - and 2.
2.
52
The fiber segment c has its lower end A just touching the sectional
plane and upper end B lying over the sectional plane. The distance
between the lower and upper ends of this fiber segment is given by
h :l cos -.
(3.9)
P-
abh h :l cos
.
abc c
c
(3.10)
N :l cos w
- , 2 d - d2 .
c
(3.11)
The number of all intersected fiber segments in the prism can be found
out by means of integration of the previous expression over all possible
directions. This is shown by using Eq. (3.3) as follows
n 5 dn
3
3.
N :l
cos - w
-, 2 d- d2 ,
c 53
(3.12a)
n
N :l
c
2
5
0
2
2
N :l
cos - ' 5 w
- , 2 d2 ( d-
cos - u
- d-.
c 50
0
53
(3.12b)
w
- , 2 d- d2
cos - w
- , 2 d - d 2
dn
,
n
5 cos - w
- , 2 d - d 2
3
w
- , 2
cos - w
- , 2
5 cos - w
- , 2 d - d 2
(3.13a)
w
- , 2 d- d2
dn cos - w
-, 2 d- d2
2
,
n
5 cos - u
- d-
(3.13b)
w
-, 2
cos - w
-, 2
2
5 cos - u
- d0
u
-
2
5 w
- , 2 d2
0
cos - 5 w
- , 2 d2
0
2
5 cos - u
- d0
2
cos - u
-
5 cos - u
- d0
. (3.14)
54
3.3
Areas of fiber sections. Figure 3.5 illustrates a very short fiber segment
of length :l. It makes an angle - to the normal x3 of the sectional plane.
The sectional area of the fiber is s * and the cross-sectional area,
perpendicular to the fiber axis is s. It is evident that s s*. The height of
the fiber segment perpendicular to the sectional plane is denoted by :y .:l.
The expression cos - :y :l can be found out from Fig. 3.5. The volume
of the observed (shaded) fiber section can be expressed in two ways: either
by the product s :l, or by the product s*:y. It is valid that s :l = s *:y.
Accordingly, the following expression can be obtained for the fiber
sectional area
x3
A-A :
s
:l A
A
:y
s
s
s
s
.
:y :l cos -
(3.15)
2
s u
- d- .
obtained for s
4.
s
s
5
0
cos - u
-
( dcos - u
- d- (
(
s
2
2
cos - u
- d-
2
55
2
u
- d-
(3.16)
5 cos - u
- d0
kn s s , kn 1
(3.17)
kn
5 cos - u
- d-.
(3.18)
w
-, 2
u
-
cos - w
-, 2
,
kn
cos - u
-
.
kn
(3.19)
(3.20)
56
kn
cos - u
- d-
2
5
0
2
sin 2 -
1
cos - sin - d-
.
2
2 0
(3.21)
Substituting Eqs. (3.7), (3.8), and (3.21) into Eqs. (3.19) and (3.20),
we obtain the following expressions
w
-, 2
u
-
sin 2 sin 2- ,
12
2
cos -
(3.22)
(3.23)
0.5
u
- sin 2u
- sin -
6
30
3
60
2
90
57
&S
n
k
k n kn .
n
Sc s
s
t
t
(3.24)
x3
S c , , k n
3.4
58
i2
- -
i1
x1
2 2)
x2
2 2)
6
8=
9
2 , 2 .
(3.25)
59
2
1
2
f0
f0
2
2
f0
,
f0
.
(3.26)
If we consider that the fiber curvature is a smooth one, then the short fibers can
be satisfactorily considered as fiber segment. (If a very short segment is cut
from a very crumpled wire, it will be straight).
60
>
y0 >
>
>
>
>
0
>
>
>0
>
>
>
>
>
>
>
>
>
> >p
> >
> >
> >
> >
> > x
x0
(a)
on the line p. Its initial point is located at the origin of the Cartesian
coordinates, the end point is located at (x0, y0) and a point below the end
point of fiber segment lying on the flexible belt is designated by
symbol. The direction of the fiber segment is defined by the orientation
angle
2 , 2 , and then we can write
tan
x0 y0 .
(3.27)
Let us now stretch the flexible belt in the direction of the shaded arrows
(direction ? of y-axis) as shown in Fig. 3.9b. The fiber segment slips
between the spikes of the belt, but it is still laying on the line p, which
passes through the origin and the point . The coordinate x0 of point
remains the same (The lateral contraction of the imaginary belt is not
considered), y0 is changed to a higher value y. The ratio C corresponds to
the drafting value of the imaginary flexible belt and it is expressed as
follows
C
7.
y
,
y0
C 1.
(3.28)
This imagination has no direct relation with the spiked lattice found in textile
machines. It is only an assisted imaginary representation, which simply
simulates the surrounding actions on the fiber segments (in other words, such
representations facilitate easy modeling). For example, in mechanical
engineering, the elastic deformation is represented by the compression of a
spring, which, in reality, is a completely different phenomenon.
61
Due to the stretching of the imaginary flexible belt, the spikes will try
to swing the fiber segment in such a manner that the segment will make an
angle (smaller) to y-axis. According to Fig. 3.9b and by using Eqs. (3.27)
and (3.28), we can write
tan
x0
x
tan
0
y C y0
C
(3.29)
d
cos 2
d
d
d
d
0
d 0
C cos 2
1 tan
C
C
2
1 C tan 2
cos
d
1 C
0
tan 2
(3.30)
C
C C 2 cos 2
(3.31)
C
cos
C 2 sin 2
cos 2
C
2
.
2
C C 1 cos 2
62
y
d
0
x
3.10 Frequency of fiber segments in elementary class.
f0
0
d
d
f0
0
d
d
C
1
2
1 C tan 2
C
1
C 2 C 2 1 cos 2
cos
(3.32)
(3.33)
C
(3.34)
Similarly, the minimum value of the probability density function fmin is
f max f
0
obtained at angle
1
f min f
2 C
(3.35)
(3.36)
63
C8
C4
f
f
1
0
2 3 6
C2
6
C 1
2
3
1
C
5
2
2
2 1 C tan
1
C tan
x C tan
C
2
1 x cos 2
; dx C cos 2
cos 2
cos 2
C
dx
;d
1
1
arctan x CAtan arctan
C tan
(3.37)
cos 2
C dx
1
.
2
t
A, A .
(3.38)
dt
1
d
cos 2
(3.39)
64
C
d
1
=
d t 1 C 2 tan 2
cos
cos 2
1 C
.
1 C 2t 2 (3.40)
y
A
yA
A
xA x
65
y x
1
.
1 2 t2
y
x
(3.41)
(3.42)
B
) .
(3.43)
preferential
direction
y
)
fiber
segment
x
3.13 Angles
and B.
66
1
C
1 C 2 tan 2 B cos 2 B
1
C
,
2
2
1 C tan
) cos 2
)
f
B
f
B
(3.44)
C
1
C 2 C 2 1 cos 2 B
(3.45)
C
1
.
2
2
C C 1 cos 2
)
-
(3.46)
(3.47)
u
-
1
C
2
2
1 C tan
- ) cos2
- )
1
C
,
2
2
1 C tan
- ) cos 2
- )
(3.48)
5 2 g
d . Because
U
- 5 g
d G
- G
-
G
u
-
C
1
2
2
C C 1 cos 2
- )
67
(3.49)
C
1
.
2
2
C C 1 cos 2
- )
68
t 3.5dtex
:l 0.1mm
t 1.7 dtex
:l 0.1mm
0.5
2
3
6
0.5
6
3
2
2
(a)
3
6
6
3
2
(b)
1.5
t 3.5dtex
:l 12.8mm
1.0
0.5
2
3
6
6
3
2
3.15 Orientation of viscose fiber segments of length :l = 12.8mm, taken from carding
web, [9]. Fibers are of 3.5 dtex, 60mm length; C = 4.28 and ) = 0.0401rad (2.3).
value of the parameter C from 1.84 to 4.28 is associated with the increase
in length :l of the fiber segments chosen from 0.1 mm to 12.8 mm. A fiber
shown in Fig. 3.16 is drawn by thin line. Its macro-trend follows the
preferential direction but its micro-shape is full of loops, waves, etc.
Therefore, the short fiber segments, whose orientation is illustrated by the
set of short arrows, are dispersed to many directions whereas the directional
orientation of longer fiber segments, marked by the longer arrows, strongly
69
Preferential
direction
follows the preferential direction. It means that the longer fiber segments
correspond to a higher value of C and vice-versa.
3.5
1
2
kn
5 cos - '' C
C
2
C
1
C
( d-
2
2
2
2
1 cos
- ) C C 1 cos
- ) (
2C cos )
arctan C 2 1 cos )
C 1
2
sin )
C2 1
ln
C 2 1 sin ) C
C 2 1 sin ) C
(3.50)9
70
1
1
C
C
cos - '
(
2
2
2
2
2
2
' C C 1 cos
- ) C C 1 cos
- ) (
.
u
-
2
)
)
2C cos )
sin
C
1
sin
C
arctan C 2 1 cos )
ln
C2 1
C 2 1 sin ) C
C2 1
(3.51)
f
g
1 .
(3.52)
u
- 2 .
(3.53)
kn
5 cos -
2 d- 2 .
(3.54)
u
-
cos -
2
cos -.
2
(3.55)
preferential direction makes angles ) 0 0 , 6 30 , 3 60 , 2 90
with the normal to the intersecting line (with y-axis). The probability density
function u(-) of non-oriented angles - can be calculated from Eq. (3.49).
The probability density function u*(-) of very short fiber segments in this
section can be calculated from Eq. (3.51). The behavior of the calculated
probability densities is shown in Figs. 3.17ad.
It is obvious from the previous example that the orientation of fiber
segments in the whole fibrous assembly differs significantly from the
orientation of fiber segments in a section of the fibrous assembly.
71
u
-
1
u
-
a)
)0
0
/2
D-
u
-
b)
u
-
) 6
c)
/2
D-
u
-
1 u
-
) 3
/2
D-
u
-
1 u
-
d)
) 2
0
D-
/2
3.17 Probability density function u (-) and u *(-) at preference level of C = 1.9.
(a) ) = 0, (b) ) 6
30 , (c) ) 3
60 , (d) ) 2
90 .
The section chooses mainly the fiber segments that are inclined at smaller
values of angle - and at these values it is always valid that u*(-) > u (-).
On the other hand, the section misses the fiber segments that are inclined
at higher values of angle - and at these values it is always valid that
u *(-) < u(-).
Note: The similar behavior can also be obtained for the negative values
of angle ).
Number of sectioned fibers per unit sectional length. A section of very
thin (planar) fibrous assembly with thickness :h, which is cut by a sectional
plane perpendicular to the y-axis is shown in Fig. 3.18. (Compare this
with Fig. 3.7). The thick lines determine an areal unit whose volume is
11:h. The mass per unit area (areal weight) G is expressed as follows
72
G
1*1* :h :h ,
G
,
:h
(3.56)
where denotes the packing density and indicates the mass density
of the fibers used.
The number of sectioned fibers per unit sectional length & is shown in
Fig. 3.18 as a set of shaded islands in the corresponding sectional area
1*:h. Using the number of the sectioned fibers per unit sectional area &S
as stated in Eq. (3.24) and the expression stated in Eq. (3.56), we can
write the following expression
&
1*:h & S :h
G
G
k n :h
kn kn .
t
t :h
t
(3.57)
y x3
:h
) preferential
direction
1
3.18 Section of a planar fiber assembly.
g1
C
1
2
2
C C 1 cos 2
)1
, )1 " 0.
73
(3.58)
longitudinal
fleece
direction
preferential
direction 1
)1
)2
y-axis
transversal
fleece
direction
preferential
direction 2
C
1
C 2 C 2 1 cos 2
)2
, ) 2 < 0.
(3.59)
C
1
2
2
C C 1 cos 2
)1 E
, E " )1.
(3.60)
Evidently, the one half of the webs in the fleece follows the preferential
direction 1 and the other half of the webs in the fleece follows the
74
gf
C
0.5
2
2
C C 1 cos 2
)1
C
0.5
, E " )1 ,
C 2 C 2 1 cos 2
)1 E
(3.61)
fleece
1
E
b
E 2
a
3.20 Scheme of angles in the web and fleece.
75
1
b
ma
c mw hw
mw bhw mf a,
f ,
.
c
c mw hw
b
mf a
(3.62)
Also, it is evident from Fig. 3.20 that c/b = sin (E/2), then the following
expression can be written
sin
E mw hw
.
2
mf a
(3.63)
76
3.6
77
P
R
Q
Step 1: Take a sufficiently long sliver. Twist one end of the sliver slightly
so as to mark the direction by which it was delivered by the machine, i.e.
forward or backward. Take out the top three plates P, Q, R (Figure 3.22)
and place the sliver onto the bottom plate D and clamp it by repositioning
the three top plates.
Step 2: Comb gently the sliver in the forward direction in order to remove
all the loose fibers that are not clamped by the plates Q and D. Discard the
combed-out fibers. Then cut the fibers using a sharp razor blade at the
right edge of the plate Q and weigh the cut fiber portion. Let this weight
be Wf. (The subscript f refers to the forward direction of the sliver).
Step 3: Remove the top plate Q. Then, comb the fibers held below it.
Retain the combed-out fiber portion and weigh it. Let this weight be Cf.
Step 4: Put the top plate Q back to its original position. Then, cut all the
fiber ends that are extending beyond the edge of the plate Q. Collect the
cut fiber portion and weigh it. Let this weight be Ef.
Step 5: Remove the plate Q again. Then, cut the fibers at the right edge
of the plate R. Collect the cut fiber portion and weight it. Let this weight
be Nf.
78
Step 6: Repeat the steps from 1 to 5 for the backward direction. Let the
corresponding weights of fibers portions be Wb, Cb, Eb, and Nb, respectively.
(The subscript b refers to the backward direction of the sliver).
Step 7: Repeat the steps from 1 to 6 for many samples of the sliver.
Then, calculate the average of the weights. Use these weights to determine
the index of fiber orientation by using the following formulas: index of
fiber orientation in forward direction Bf = 1 E f/Nf and index of fiber
orientation in backward direction Bb = 1 Eb/Nb.
Modeling scheme. Let us now consider the following assumptions: (1)
all fibers in the sliver have the same (straight) length l 10 and the same linear
density (fineness) t, (2) all fibers have the same waviness so that the shorter
crimped length a of each fiber is the same (see Figure 3.5 preliminarily), (3)
fibers are distributed randomly along the sliver, (4) the number of fibers is
very high in the cross-section of the sliver. The random organization of the
crimped fibers in the sliver is represented by the parallelograms as shown in
Figs. 3.23 a, 3.24a and 3.25a. Let us imagine that a part of such sliver is
firmly gripped by a bottom plate and two top plates R and Q. The vertically
shaded plate R permanently grips the sliver used, but the dotted plate Q is
removed two times during the process of measurement.
We must think about the following three cases.
Case (1): The crimped fiber length a is longer than the width d of the
top plate Q,
Case (2): The crimped fiber length a is shorter than the width d of the
top plate Q, but the straight fiber length l is longer than d,
Case (3): The crimped fiber length a as well as the straight fiber length
l is shorter than the width d of the top plate Q.
Figure 3.23 illustrates the first case (a > d) see sliver in Fig. 3.23a.
Consider that the fibers protruding from the right-hand edge of the top
(dotted) plate Q are combed as shown in Fig. 3.23a. As a result, the fibers
which were not gripped by the top and bottom plates are removed and the
fibers which were gripped by the plates are straightened. The straightened
fibers protruding from the right-hand edge of the top plate Q are then cut
and the first fringe of fibers of weight W is thus obtained. In the next step,
the top plate Q is removed and the rectangle of crimped fibers is seen to
be lying under it see Fig. 3.23b. Let us imagine that we straighten all
these fibers so that we obtain a wider rectangle of straight fibers as shown
in Fig. 3.23b. Of course, in reality, only the fibers protruding from the
right-hand edge of the top plate R are straightened by combing, whereas
the other fibers are combed out and contribute to the fringe weight C.
Now the plate Q is replaced back on the straight fibers gripped by the
10.
(a')
Q
weight W
(a)
R
(b')
R
weight C
(b)
Q
weight E
(c)
R
weight N
(d)
3.23 Scheme of modeling in accordance with Case (1) when a > d.
79
80
(a')
R
Q
weight W
(a)
R
(b')
R
weight C
(b)
R
Q
weight E
(c)
R
weight N
3.24 Scheme of modelling in accordance with Case (2) when a < d and l > d.
81
(a')
R
Q
weight W
(a)
R
(b')
R
weight C
(b)
R
weight
E=0
(c)
R
weight N
3.25 Scheme of modeling in accordance with Case (3) when a < d and l < d .
82
plate R. This is shown in Fig. 3.23c. The protruding fibers are then cut and
a small fringe of fibers of weight E is thus obtained. In the final step, the top
plate Q is once again removed and the fibers protruding from the right-hand
edge of the top plate R are observed as shown in Fig. 3.23d. The protruding
fibers are then cut and a fringe of fibers of weight N is obtained.
Figure 3.24 illustrates the second case (a < d and l > d). The commentary
of the used procedure can be quite the same, only the shapes of the fringes
are different (Compare the corresponding schemes displayed in Figs. 3.23
and 3.24).
Figure 3.25 illustrates the third case (a < d and l < d). The commentary
of the used procedure can also be quite the same. We only notice another
shape of the fringes and remark that the cut off fringe shown in Fig.
3.25c does not exist its weight E is equal to zero.
It can be noted that the aforesaid step-by-step procedure can also be
followed by putting another top plate P at the left-hand edge of the lefthand side top plate for backward direction of the sliver.
Fiber orientation. Each fiber is interpreted as a chain of small straight
segments of constant lengths :l as shown in Fig. 3.26. The projection of
a segment in the direction of the axis of the sliver is denoted by the crimped
length :a, and its non-oriented angle of inclination to the axis of the
sliver is indicated by -. They are related as follows
:a = :l cos-.
(3.64)
The quantities :a and cos- are random variables with mean at :a and
(3.65)
:a
:l
fiber
length l
sliver
axis
83
cos -
5 cos - u
- d- k
(3.66)
k n :a :l .
(3.67)
kn
2C arctan
C2 1
C2 1
(3.68)
(3.69)
(3.70)
If it is assumed that there are n number of fibers present in the crosssection of the sliver then the expression (Eq. 3.71) is valid to write
11. We use this equation also more generally; each crimped fiber length divided by
the straightened fiber length is equal to kn in this model.
84
T nta n
t
T
, n kn ,
kn
t
(3.71)
B'
y2 ,
n2
O
y,
n
H'
y,
n
y1 ,
n1
C'
l
D'
d'
(a)
(b)
a
A
B B'
y2 ,
n2
O
y2 ,
n2
O
H'
y,
n
y1,
n1
O'
E
y1 ,
n1
O'
K'
y3 ,
n3
D'
y,
n
y3 ,
n3
d'
(c)
(d)
The first fringe, earlier shown in Fig. 3.23a, is now shown in detail in
Fig. 3.27a. Here, the number of fibers gripped along the clamping line BD
of length y is equal to the total number n of fibers present in the cross-
85
section of the sliver. This fringe of straight fibers has the triangular shape
BCD with the longest fiber of length l. The total length of fibers in such
a fringe is n l/2. By applying Eq. (3.71) we obtain the expression for the
weight (mass) W of the fringe using fiber fineness t as follows
l
T l
l
W n t k n t T kn .
2
t
2
2
(3.72)
(3.73)
Let the number of combed-out fibers be n1 and these fibers are lying at
a distance y1, the remaining n2 fibers are lying at a distance y2. Let us take
that the distance y = y 1 + y 2 is proportional to the number of fibers n
= n1 + n2, y1 is proportional to n1 and y2 is proportional to n2, all with a
common constant of proportionality.
It is shown that the triangle ABD is similar to the triangle OHD.
Therefore,
a
d
,
y1 y2 y1
a y1 y2
y
1 2 ,
d
y1
y1
(3.74)
y2
a d 1.
y1
According to the proportionality the following equations are valid to
write
y2/y1 = n 2/n1,
(3.75)
n1 + n 2 = n.
(3.76)
86
n2 n1
a d 1
n n2
a d 1 ,
n2 n2
a d 1 n
a d 1 ,
n2 1
a d 1 n2 a d n
a d 1 ,
(3.77)
a d 1
d
n2 n
n 1 ,
a
ad
and
n1 n n2 n
d
.
a
(3.78)
T
d
kn t Td . 12
t
kn
(3.79)
Using Eqs. (3.78), (3.73), (3.69), and (3.71), the weight of the combedout fibers (dashed lines in the triangle OH;D;) is expressed as follows
C n1
d;
d d
d 1
d
t n t n
t
a 2 kn
lkn 2kn
2
T
d d
d2
kn
t T
2lkn
t lkn 2kn
(3.80)
d; d; d
y1
y3
(3.81)
y3 y1 1 d d ;
12.
This result is immediately visible with a view to the sliver shown in Fig.
3.25b.
87
(3.82)
Substituting Eqs. (3.82) and (3.73) into Eq. (3.81) and then using
Eq. (3.78), the following expression for n3 is obtained
n3 n1
1 kn n
1 kn .
a
(3.83)
By using Eqs. (3.73), (3.77), (3.83), (3.71), and (3.69), the weight E of
the third fringe of fibers (rectangle BB;H;H and triangle ODD) is
expressed as follows
E n2
d ; d t n3
n
d; d
t t
d ; d n2 3
2
2
d
d 1 d
t ' d ( ' n 1 n
1 kn (
k
a
a
2
n
1
d 1d 1d
kn
tdn 1 '1
k
a 2 a 2 a (
n
1
d 1d
T
1
d
1 d
tdn 1 '1
kn ( td kn 1 '1
kn (
t kn 2lkn 2 lkn
kn 2 a 2 a
d
d
Td
1 k n '1
(.
2
l
2
lk
n
(3.84)
Figure 3.23d displays the arrangement of fibers after cutting the third
fringe of fibers. The weight N of the last fringe of fibers, lying in the
rectangle FBOK, can be expressed by using of Eqs. (3.79), (3.80), and
(3.84) as follows
N GTOTAL C E Td T
d
d2
d
Td
1 kn '1
(
2lk n
2l 2lkn
dk
d
d
d
d
Td 1
1
kn n
2l 2lkn
2l 2l
2lk n
(3.85)
d
Tdkn 1 .
2l
Naturally, an identical result can also be obtained by using the
geometrical dimensions directly from Fig. 3.23d, i.e., after rearrangement
of the following equation: N
n1 n3 d 2
n2 n3 d t .
88
Geometry and weights of fringes when a < d and l > d. The detailed
geometry of these fringes, introduced generally in Fig. 3.24, is characterized
in the following four schemes shown in Fig. 3.28.
The first fringe, introduced in Fig. 3.24a, is now shown in Fig. 3.28a.
This scheme is fully analogical to the scheme shown in Fig. 3.27a so that
Eq. (3.72) is valid for the weight W in this case too. Also, Eq. (3.79) is
valid for the total weight GTOTAL of the sliver, i.e., for all fibers in the
rectangle FGD;B; in Fig. 3.28b gripped fibers (continuous line) and tobe-combed-out fibers (dashed lines).
O'
l
y,
n
y,
n
C'
D
(a)
(b)
O'
O'
y,
n
H J
y4 , y ,
n4 n
E
E
(c)
y4 ,
n4
E
D
(d)
3.28 Schematic representation of fiber geometry in Case (2) when a < d and also l > d.
89
The fibers gripped by clamping line FG are lying in the trapezoid OJAF
and in the triangle GEO. Nevertheless, the triangle GEO can be
replaced onto the position FAO. If the length y = GF = OO then the
weight of gripped fibers can be expressed as follows
l
T l
l
F n t kn t Tk n .
2
t
2
2
(3.86)
lk
l
T d n .
2
2
(3.87)
(3.88)
The triangle KHJ is similar to the triangle OOJ and the line OO has
the same length as that of FG. Then,
y4 n4 l d
ld
, n4 n
y
n
l
l .
(3.89)
l d
ld
ld ld
T
l d
E n4
tn
t kn
t Tkn
.
2
l
2
t
2l
2l
2
(3.90)
The last fringe, after removing the plate Q, is shown in Fig. 3.28d. The
weight N of this fringe is equal to the weight F, derived earlier, minus the
weight E of the triangular fringe KHJ. This is written below
l d Tkn
l
N F C Tkn Tk n
2
2l
2
2
d
Tdkn 1
2l
d2
l
l
2
d
l
(3.91)
90
Geometry and weights of fringes when a < d and also l < d. The detailed
geometry of these fringes, introduced generally in Fig. 3.25, is characterized
in the following four schemes displayed in Fig. 3.29.
O'
B
l
A F
y,
n
y,
n
l
B
C'
E G
(a)
(b)
O'
O'
A F
y,
n
A F
O
l
y,
n
O
l
E
E G
(c)
E G
(d)
3.29 Schematic representation of fiber geometry in Case (3) when a < d and also l < d.
91
The first fringe, introduced in Fig. 3.25a, is now displayed in Fig. 3.29a.
This scheme is fully analogical to the scheme shown in Fig. 3.27a so that
Eq. (3.72) is valid for the weight W in this case too. Also, Eq. (3.79) is
valid for the total weight GTOTAL of the sliver, i.e., for all fibers in the
rectangle FGD;B; in Fig. 3.29b gripped fibers (continuous line) and tobe-combed-out fibers (dashed lines).
In analogy to the previous discussion for obtaining Eq. (3.86), the fibers
gripped by the clamping line FG are lying in the trapezium OJAF and in
the triangle GEO, and the triangle GEO can be replaced onto the
position FAO. So, Eqs. (3.86) and (3.86) are also valid in this case for
the weight F of the gripped fibers and for the weight C of the combed-out
fibers (dashed lines).
Nevertheless, in this case, no one fiber is touching the edge BD after
returning the top plate Q back to its original position. This is shown in
Fig. 3.29c. Therefore, the following equation is valid to write
E = 0.
(3.92)
The last fringe obtained after removing the plate Q is shown in Fig. 3.29d.
The weight N of this fringe is equal to the weight F of the fibers gripped by
the plates along the edge FG. Applying Eq. (3.86) we obtain
l
.
(3.93)
2
Summary of results. The results of weight of the fiber fringes derived
earlier are summarized in Table 3.1.
N F Tkn
Fiber length13:
Case (1)
Case (2)
Case (3)
l " d kn
l
d , d kn
l<d
Equation (3.72)
Weight W
Weight C
Equation (3.80)
Weight E
Equation (3.84)
Equation (3.87)
Weight N
Equation (3.90)
Equation (3.92)
Equation (3.93)
92
W
5W
(3.72)
l dl ,
(3.94)
lmax
C
C
l dl
d kn
lmax
E
5
0
lmax
N
5
0
C(3.87)
l dl
C(3.80)
l dl ,
(3.95)
d kn
d kn
E
l dl 5 E(3.92)
l dl
lmax
E(3.90)
l dl
E(3.84)
l dl , (3.96)
N (3.85)
l dl .
(3.97)
d kn
lmax
N
l dl 5 N (3.93)
l dl
lmax
93
T M L
Original sliver
T1 M 1 L 1T
T2 M 2 L 2T
i-th partial sliver, fiber length li,
Ti M i L iT
m-th partial sliver, fiber length lm,
Tm M m L mT
L
3.30 Scheme of original and partial slivers.
m
i 1
i 1
i 1
i 1
E Ei , and N N i .
Practical example. A polyester drawn sliver of 5.37 ktex linear density
was taken for this study. The weight-based fiber length distribution of this
sliver is shown in Table 3.2. The fineness of the polyester fiber was found
to be 1.67 dtex.
The methodology as reported in mathematical model was followed to
obtain the values of W, C, E, and N in the forward and the backward
directions of the sliver. The average of one hundred such readings carried
out on the aforementioned sliver is reported in the column named
experimental in Table 3.3. The width of the plate d was kept at 12.7 mm.
A computer program was developed to find out the corresponding values
by using the equations derived earlier. They are reported in the column
named theoretical in Table 3.3. It was observed that the summation of the
squares of the deviations between the experimental and theoretical readings
for the four variables corresponding to the forward and backward directions
was found to be minimum at k n = 0.73 and k n = 0.71, respectively.
94
Length (mm)
Relative frequency ()
36
35
30
14
Total
0.65625
0.19792
0.07812
0.06771
1.00000
Wf
Cf
Ef
Nf
Weight (mg)
Experimental
Theoretical at kn = 0.73
0.0590
0.0157
0.0067
0.0461
0.0663
0.0185
0.0098
0.0399
Backward direction
Variable
Wb
Cb
Eb
Nb
Weight (mg)
Experimental
Theoretical at kn = 0.71
0.0568
0.0163
0.0048
0.0466
0.0645
0.0190
0.0104
0.0388
2 , 2 , in the forward and backward
0.4554
0.5096
95
these functions is displayed in Fig. 3.31. As expected, the fibers are more
anisotropically oriented in the forward direction as compared to the
backward direction in the case of this drawn sliver.
0.55
0.5
C=1.6 (Forward)
0.45
C=1.43 (Backward)
0.4
0.35
0.3
0.25
0.2
-2
-1.5
-1
-0.5
0.5
1.5
2
radian
3.8
1.
2.
3.
4.
5.
6.
7.
References
Neck, B. (1998). Morphology and Structural Mechanics of General Fibrous
Assemblies, Technical University of Liberec, Czech Republic, pp. 2139.
Folgar, F., and Tucker III, C. L. (1984). Orientation behaviour of fibres in
concentrated suspensions, Journal of Reinforced Plastics and Composites, 3,
98119.
Jackson, G. W., and James, D. F. (1986). The permeability of fibrous porous
media, The Canadian Journal of Chemical Engineering, 64, 364374.
Mao, N., and Russell, S. J. (2003). Anisotropic liquid absorption in homogeneous
two-dimensional nonwoven structures, Journal of Applied Physics, 94, 4135
4138.
Murugan, R., and Ramakrishna, S. (2007). Design strategies of tissue
engineering scaffolds with controlled fibre orientation, Tissue Engineering, 13,
18451868.
Pourdeyhimi, B., Ramanathan, R., and Dent, R. (1996). Measuring fibre
orientation in nonwovens, Part II: Direct tracking, Textile Research Journal,
66, 747753.
Pourdeyhimi, B., Dent, R., and Davies, H. (1997). Measuring fibre orientation
in nonwovens, Part III: Fourier transformation, Textile Research Journal, 67,
143151.
96
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
Fiber-to-fiber contacts
97
4
Fiber-to-fiber contacts
4.1
The fibers contact with each other in a fibrous assembly. The region of contact,
schematically shown by the dotted line in Fig. 4.1, is defined as contact between
two fibers. The place on fiber surface, which is in contact with other fiber, is
called contact place. Hence, it is clear that one contact is created from two
contact places. The number and position of mutual contacts among the fibers
is known to influence the mechanical behavior of a fibrous assembly, for
example, the mutual forces causing the deformation of a fibrous assembly are
transmitting due to fiber-to-fiber contacts, etc.
Basic idea of van Wyk model. The simple textile fibrous assemblies are
usually consisted of randomly oriented fibers. Therefore, the contacts
between fibers also originate randomly. In contact places, some fibers try
to be shifted to the place of other fibers while the other fibers resist this
shifting. As the fibers cannot penetrate into each other, this process is
ended with creation of fiber contacts.
In an abstract sense, we can consider the fibers as geometrical bodies
(negligible mass or weightless), which can mutually penetrate into each
other. In such cases, the shifting of fibers will not end only with fiber
contact, but they will penetrate into each other. In other words, the contact
between fibers occurs in those places where the imaginatively geometrical
fiber bodies together create a common penetration. With this great idea,
van Wyk [1] converted the problem of contact to a classical concept of
intersections of bodies.
Idealized fibers and their position in space. It is a fact that the geometry
of fibers is very much complicated in a real fibrous assembly. The model
97
98
C
B
R
X
2a
G
S
H
d 2
d
d 2
2c
A
l
P
E
i1
B
Q
F
i2
X
2b
Two fibers in a space. Let us now consider two fibers (fiber No. 1
and fiber No. 2) in a space. Let us also consider that fiber No. 1 has a
given direction and position. Its direction is expressed by the unit
orientation vector i 1, which is defined by the two spherical coordinates
-1, 2 1. The coordinates of end point P shown in Fig. 4.2 describe its
position. Fiber No. 2 has only a given direction, expressed by the unit
orientation vector i 2 , which is defined by the couple of spherical
Fiber-to-fiber contacts
99
(4.1)
Note: So, the angle B is always a function of four variables -1, 21, -2,
22 only.
The area of the trapezoid ABCD in Fig. 4.2 is l*l sinB and the volume of
the oblique prism is
(4.2)
100
(b) The point X lies at the right hand side of the plane BCGF or at the
left hand side of the plane ADHE,1
(c) The point X lies in front of the plane ABFE or behind the plane
DCGH. 2
To the contrary, cylinders 1 and 2 have a common intersection if the
point X lies inside the oblique prism ABCDEFGH (see e.g. the position 2c).
The position of the point X is random and it may lie at any place in the
fibrous assembly of total volume Vc. The probability that point X just lies
inside the oblique prism is P V1,2 Vc (geometrical definition of
probability). By applying Eq. (4.2) we find
(4.3)
If point X lies inside the oblique prism, the intersection of both cylinders
is realized. Under that condition, according to van Wyks model, a contact
between these (material) fibers occurs. Then, Eq. (4.3) expresses the
probability P of occurrence of contact between two fibers.
Mean number of contact places on the fiber No. 1. Komori and
Makishima [2] reported the number of contacts, density of contacts, and
other quantitative characteristics regarding fiber-to-fiber contact in a
fibrous assembly, based on C. M. van Wyks model [1].
We assume that a fibrous assembly is created from a large number of
(idealized) randomly organized fibers whose total number N is very large.
The distribution of fiber orientation is described by the joint probability
density function w(-, 2) and the domain of directions is expressed by
symbol 3 : -
0, 2 4 2
0, 2 according to Eq. (3.1).
Let us now choose (arbitrarily) fiber No. 1, having an orientation vector
i1 defined by spherical coordinates -1, 21. Hence, the number of other fibers
in that assembly is
N2 = N 1.
1.
2.
(4.4)
The concept is considered as a rough idea, because of neglecting fiber ends and
fiber thickness. For example, if point X lies in the plane PQRS between the two
lines PQ and RS and only a small part (less than d/2) in the right hand side
of the plane BCGF, or in the left hand side of the plane ADHE, then intersection
occurs. However, this is insignificant, because d << l .
This consideration is also an approximate one. For example, if point X just lies
in the plane PQRS between the lines PS and QR and only a small part of it
(less than d/2) lies in front of the plane ABFE or behind the plane DCGH, then
intersection occurs. However, this is also insignificant, because d << l .
Fiber-to-fiber contacts
101
(4.5)
From the fibers lying in the mentioned elementary class, the mean
value of number of contact places on fiber No. 1 can be expressed by
dm1 = P dN2. By applying Eqs. (4.3) and (4.5), we get
dm1 P dN 2
2dl 2 sin B
N 2 w
- 2 , 2 2 d- 2 d2 2
Vc
(4.6)
2dl 2 N 2
sin B w
- 2 , 2 2 d- 2 d2 2 .
Vc
The mean number m1 of all contact places on fiber No. 1 (from all N2
fibers) is
m1 = 5 dm1
32
2dl 2 N 2
Vc
55 sin B w
- , 2 d2
d2 2 ,
(4.7)
32
102
(4.8)
The directions of these fibers are expressed by the vector i1, which is
defined by the couples of spherical coordinates -1, 21. The distribution of
the directions of the considered N1 fibers is given by the joint probability
with
defined
domain
density
function
w(- 1 , 2 1 )
31 : -1
0, 2 4 21
0, 2 .
A directional elementary class of mentioned N1 fibers is determined
by the intervals (-1, -1 + d-1) and (21, 21 + d21). The relative frequency
of fibers in such elementary class is w(-1, 21)d-1d21 and the absolute
frequency (number) of fibers in this elementary class is then
dN1 N1 w
-1 , 21 d-1 d21 .
(4.9)
The mean number of contact places m1 on one fiber having this direction
is given by Eq. (4.7). Also the other fibers belonging to the considered
class have the same mean number of contacts. The number of contact places
dM on all fibers in the elementary class is then dM = m1 dN1 and by applying
Eqs. (4.7) and (4.9) we find
2d l 2 N
2
dM m1dN1 '
V
c
'
55 sin B w
- , 2 d- d2
2
32
( N1w
-1 , 21 d-1d21
(
2d l N 2 N1
' 55 sin B w
- 2 , 2 2 d- 2 d2 2 ( w
-1 , 21 d-1d21 .
Vc
' 3 2
(
2
(4.10)
The number M of contact places on all fibers (not only on the considered
fibers in one elementary class, but also on the fibers having other directions)
is
G7 2d l 2 N N
67
2 1
M 5 dM 55 H
' 55 sin B w
- 2 , 2 2 d- 2 d2 2 ( w
-1 , 21 d-1d21 8
Vc
31
31 7
79
' 32
(
I
2
2d l N 2 N1
55 55 sin B w
-1 , 21 w
- 2 , 2 2 d-1d21d- 2d2 2 . (4.11)
Vc
31 , 3 2
Let us introduce a dimensionless quantity I as follows
I 55 55 sin B w
-1 , 21 w
- 2 , 2 2 d-1d21d- 2 d2 2 .
31 , 3 2
(4.12)
103
Fiber-to-fiber contacts
M
2d l 2 N 2 I
.
Vc
(4.13)
We know that one contact creates two contact places. Therefore, the
number n of contacts in the fibrous assembly is
n
4.2
M d l2N 2I
.
Vc
2
(4.14)
Vc
2
V N d 4 l N d 2l
.
4
(4.15)
&
n d l 2N 2I
16 2
16 I
2
2
d
l
N
I
2 3 2 .
2
2 2 4 2
Vc
Vc
N d l
d
(4.16)
16 I .
2d 3
(4.17)
104
(4.18)
:1
:3
Fiber-to-fiber contacts
:
Nl
Nl
Nl
N d 2l d 1
.
M
2d l 2 N 2 I Vc 2d l 2 N 2 I 4
8I
105
(4.19)
k:
(4.20)
(The note to the parameter k& stated after Eq. (4.18) is valid for the
parameter k: similarly). Now the mean distance between the contact places
can be expressed as follows
:
k:
.
(4.21)
Note: The relationship between k& and k: can be found from Eqs. (4.17),
(4.20), and (1.7) as follows
k & k:
16 I d
2
1
1 .
2 3
2
2
d 8 I d
2s
2 d 4
(4.22)
The value of the product k&k: depends only the cross-sectional area of
fibers s. The product k&k: is independent of the arrangement of fibers.
The following examples illustrate the numerical values resulting from
the previous equations.
Example 4.1: We consider a special regular arrangement of cylindrical fibers
of diameter d as shown in Fig. 4.4a. The structure is created from cubes each
of side d, as shown in Fig. 4.4b. The volume of the cube is Vc = d 3. The material
volume creates two halves of fiber of length d, i.e., V d 2 4 d d 3 4 .
The packing density is V Vc d 3 4 d 3 4 ! 0.7854 . In the cube,
there is only one contact (the black arrow in Fig. 4.4b); n = 1. The density of
contacts is & n Vc 1 d 3 . By applying and & into Eq. (4.16), we find
2
1 d 3 16 I 2 d 3 4 I d 3 . Accordingly, we find I = 13 . For example,
we consider a polyester fiber of fineness t = 0.17 tex and density
= 1360 kg m1. According to Eq. (1.6), fiber diameter is d = 0.012616 mm.
We find k& = 807338 mm3 and k: = 0.00495430 mm from Eqs. (4.17) and
(4.20). The numerical results mentioned in Table 4.1 are obtained by applying
3.
We may get the same result if we realize that the contact fibers make always a right
angle B 2 ; therefore, constantly, sin B 1 in Eq. (4.12). Nevertheless, a little
problem can bring here the joint probability density function w, which is equal to
zero for all angles except 0 and 2 and is equal to infinity for these two angles.
Therefore a suitable limitation procedure must be used in this case.
106
k& and k: into Eq. (4.18) and Eq. (4.21), respectively for similar type of
structures (fibers contact at right angle) having different packing densities.
d
(a)
(b)
Density of contacts
& mm3
Mean distance
498007
201835
0.006308
0.009909
0.3
0.1 (orientation value for roving)
0.01 (orientation value for sliver)
72660
8073
80.73
0.016514
0.049543
0.495429
: mm
I 55 55 sin B w
-1 , 21 w
- 2 , 2 2 d-1d21d- 2 d2 2
31 , 3 2
55 55 1 cos 2 B
31 , 3 2
55 55
-1
0, 2
21
0,2
- 2
0, 2
2 2
0,2
sin -1 sin - 2
d-1d21d- 2 d2 2
2
2
1
sin -1 cos 21 sin - 2 cos 2 2 sin -1 sin 21 sin - 2 sin 2 2
12
2
cos -1 cos - 2
sin -1 sin - 2
d-1d21d- 2d2 2 .
4 2
107
Fiber-to-fiber contacts
Density of contacts
Mean distance
& mm
: mm
0.7854
391136
0.008032
158521
57067
0.012616
0.021027
0.3
0.1 (approx. roving packing density) 6341
0.01 (approx. sliver packing density) 63.41
4.3
0.063080
0.630799
Probability (p). In Section 4.1, Eq. (4.3) and Fig. 4.2 define the probability
P 2d l 2 sin B Vc that fiber No. 2 contacts fiber No. 1. We may remind
that the angle B can be expressed by Eq. (4.1) as a function of spherical
coordinates -1, 21 and -2, 22. The probability of the randomly selected
fiber No. 2 belonging to the elementary class, which is defined by the
angle intervals (-2, -2 + d-2) and (22, 22 + d22), is w(-2, 22) d-2d22 (see
also the considerations mentioned before Eq. (4.5)). The probability d p of
the randomly selected fiber No. 2 belonging to the elementary class and at
the same time is in contact with fiber No. 1 is given by the product of
these independent probabilities
dp P w
- 2 , 2 2 d- 2 d2 2
2d l 2 sin B
w
- 2 , 2 2 d- 2 d2 2 .
Vc
(4.23)
p 55 dp
32
4.
2d l 2
sin B w
- 2 , 2 2 d- 2 d2 2 .
Vc 55
32
This value will be later derived also analytically see Eq. 4.44.
(4.24)
108
If we denote
J
-1 , 21 55 sin B w
- 2 , 2 2 d- 2 d2 2
(4.25)
32
p
2d l 2
J
- 1 , 21 .
Vc
(4.26)
(As B is a function of -1, 21, -2, 22, the integral of J is a function of -1, 21).
It is possible to rearrange the last equation by the following manner. The
length and equivalent diameter of an individual fiber are l and d, respectively.
Hence, the volume of individual fiber is d 2 4 l and the volume of N
fibers in the fibrous assembly is V N d 2 4 l . We also consider the total
volume of the fibrous assembly is Vc. From the definition of packing density
expressed in Eq. (1.22), we can write
V
Vc
d 2
l
2
4 , N 2d l 8 l ,
Vc
Vc
d
2d l 2
8 l
.
Vc
N d
(4.27)
8 l
J
-1 , 21 .
N d
(4.28)
(4.29)
m1
N 1 p
N 1 8 l J
N d
-1 , 21 .
(4.30)
Fiber-to-fiber contacts
109
(4.31)
(4.32)
In other words,
P
0 e m1 , P
x
m1x
P
0 .
x!
(4.33)
A
m1
8 l
2d l 2
J
-1 , 21 N
J
-1 , 21
d
Vc
2d l 2
N
Vc
55 sin B w
- , 2 d2
(4.34)
d2 2 .
32
dK
x P
x * w
-1 , 21 d-1 d21
5.
m1x m1
e w
-1 , 21 d-1 d21 .
x!
(4.35)
110
The probability K (x) of the randomly selected fiber having just x contact
places is obtained from the integration of the previous expression. By
applying Eqs. (4.31) and (4.35), we obtain
K
x = 55 dK
x 55
31
31
m1x m1
e w
-1 , 21 d-1d21
x!
x
55
31
1 8 l
8 l
J
-1 , 21 ( exp '
J
-1 , 21 ( w
-1 , 21 d-1d21 ,
'
x ! d
d
(4.36)
where the integral of J (-1, 21) is defined by Eq. (4.25) and it implicates
the function of sin B 1 cos 2 B , which is deduced from Eq. (4.1).
In fact, K (x) is a special type of distribution, which depends on fiber
orientation (via joint probability density function w) and packing density
of the fibrous assembly. In general, a suitable numerical integration method
can be used for evaluation of the expression of K (x) and of course, this is
going to be difficult.
It is relatively easy to find out the mean value m of the number of contact
A
55
31
x P
x w
- , 2 d1
d21
x0
8 l
J
-1 , 21 w
-1 , 21 d-1 d21
d 55
31
8 l
55 sin B w
- 2 , 2 2 d- 2 d2 2 w
-1 , 21 d-1 d21
55
d 31 32
I
8 l I
8 l
w
w
. (4.37)
sin
,
,
d
d
d
d
B
2
2
2
1
1
2
2
2
2
1 21
55
55
d
d 31 32
Fiber-to-fiber contacts
MN
8 l I
2d l 2 I
.
N2
Vc
d
111
(4.38)
(4.39)
(4.40)
(4.41)
112
We can write the integration of Eq. (4.25), from Eqs. (4.40), (4.41),
and (4.42).
J
-1 , 21 J
-1; , 21; 55 sin B w
- ;2 , 2 ;2 d- 2; d2 ;2
3;2
55 sin - ;2
3;2
2
5
0
sin -;2
d- ;2 d2 ;2
2
1
2
2
1
2
sin 2 - ;2
d
2
d
;
;
5 2 2 5 sin - ;2 d- ;2
2
0
(4.43)
2
1
1
' - 2 sin
2- 2; ( .
4
4
2
0
By applying Eqs. (4.25) and (4.43) into Eq. (4.12), we obtain the value
of the integral I for the fibrous assembly with spatial isotropic orientation
as follows
I 55 55 sin B w
-1 , 21 w
- 2 , 2 2 d-1d21d- 2 d2 2
31 , 3 2
J
-1 , 21 4
55 w
-1 , 21 '55 sin B w
- 2 , 2 2 d- 2 d2 2 ( d-1d21
' 32
(
31
1
( 0, 7854).
55 w
-1 , 21 d-1 d21
4 31
4
(4.44)
Komori and Makishima [2] and Stern [5] also found the constant
J
-1 , 2 1 I 4 .
Fiber-to-fiber contacts
113
m1
8 l
8 l 2 l
8 l I 8 l 2 l
J
-1 , 21
, m
,
d
d 4
d
d
d 4
d
(4.45)
2 l
m m1
.
d
K
x = 55
31
m1x m1
e w
-1 , 21 d-1 d21
x!
1
2 l 2d l
e w
-1 , 21 d-1 d21
x! d
x
55
31
1
2 l 2d l
e
x! d
x
(4.46)
1
x
1
2 l 2d l
553 w
-1 , 21 d-1 d21 x ! d e ,
1
J
-1 , 21 J
2
1 5
sinB f
2
2
sin
(4.47)
2
2d
2d
114
Integral of I is defined by Eq. (4.12) and using Eq. (4.25) it takes the
following form
I 55 55 sin B w
-1 , 21 w
- 2 , 2 2 d-1 d21 d- 2 d2 2
31 , 3 2
J
-1 ,21
55 ' 55 sin B w
- 2 , 2 2 d- 2 d2 2 ( w
-1 , 21 d-1 d21
(
31 '
32
(4.48)
2
5 J
f
d
1
2
From Eqs. (4.31) and (4.47), we can re-write the mean number of
contacts on fiber No. 1 as follows
m1
8 l
J
d
(4.49)
K
x =
2
m1x m1
5 x! e f
- 2
1 d
(4.50)
2
1 8 l
J
x ! ' d
- 2
8 l
J
exp '
d
(
( f
1 d
115
Fiber-to-fiber contacts
2
1 J
1; J
0 5
;2 0
sin
2
1
d ;2
2
1
2 sin ;2 d
50
2
2
2
cos ; 0 .
(4.51)
2
From Eqs. (4.48) and (4.51), we obtain the value of the integral I as
follows
2
I
2
1
J
1; f
1; d
2
1;
2
f
5 2
1; d
2
.
;
(4.52)
Komori and Makishima [2] and Pan [6] also found the parameter
J
1 I 2 .
The mean value of contact places on each fiber is equal regardless to its
direction. Accordingly, by applying Eq. (4.51) into Eq. (4.49) and Eq. (4.52)
into Eq. (4.37), we find
8 l
8 l 2 16 l
8 l I 8 l 2 16 l
J
1
2 , m
,
d
d d
d
d 2 d
16 l
m m1 2 .
d
m1
(4.53)
The distribution K (x) of the number of contact places on the fibers can
be found by applying Eq. (4.53) into Eq. (4.50)
K
x =
2
m1x m1
5 x! e f
2
x
1 d
1
16 l
2 e
x! d
2
1
1
16 l 2
2
d
2
16 l
1
16 l 2 d
f
e
5
x! 2d
2
x
1 d
16 l
1
16 l 2 d
.
1 d 1
e
x! 2d
(4.54)
2
It is the Poisons distribution having the mean value m 16 l d .
116
fiber aspect ratio is assumed for cotton fibers as = l/d = 2000 and the
packing density is chosen as = 0.1, 0.3 and 0.5. (The graphs are defined
and therefore calculated only for integers of x values).
0.02
0.1
l d 2000
K
x
0.3
0.5
0.01
b
500
1000
1500
2000
4.5 Probability density K (x) of the number of contact places on fiber in case:
aJ fibrous assembly with spatial isotropic orientation Equation (4.46).
bJfibrous assembly with plane isotropic orientation Equation (4.54).
4.4
Fiber-to-fiber contacts
117
2
a2
a1
1
3b
3a
generally said that every other fiber will realize its contact with fiber No.
1 only when its random position will be outside of the earlier created
blocking sections a1, a 2,. Evidently, the probability of every other
contact with fiber No 1 is not the same but is a decreasing quantity.
Interaction. If a fiber wants to be in contact with another fiber then
they must be mutually very close. The sufficient vicinity of fiber No. 2 to
fiber No. 1 was earlier also the sufficient condition for creation of contact
(Section 3.1, Fig. 4.2). This situation, where a fiber is very close to other
fiber will be called by interaction, now. Fiber No. 1 may (but need not)
contact fiber No. 2 only if fiber No. 2 is in interaction with fiber No. 1.
Earlier, van Wyks probability p according to Eq. (4.26) meant the
probability that a randomly selected fiber No. 2 (regardless of its direction)
is in contact with the given fiber No. 1. The same equation says now that
the so randomly selected fiber No. 2 is only in interaction with fiber No.
1. Similarly, the Poison distribution in Eq. (4.32) or (4.33) expresses only
the probability P (x) of the given fiber No. 1 having just x interactions
with other fibers (never x contacts, as it is valid in the original theory
according to van Wyks concept).
Note: We assume number of fibers NA in this model; then the
mentioned Poison distribution of interactions is valid.
Occurrence of contact place. We imagine that a fibrous assembly is
formed in such a manner that at first we place the given fiber No. 1 in the
available space, and then in a random way, we place the other fibers. In
this way, some of the fibers will interact with fiber No. 1 and some of the
interacting fibers will create contact places with it.
118
l
l l
i 1 z
i 1
ri
1
, i z 1 .
l
l
z
(4.55)
The complementary probability si, that the new interacting fiber do not
create i-th contact place (similarly as fiber 3b on Fig. 4.6) just now on
fiber No. 1 is
i 1
, i z 1.
si 1 ri
(4.56)
z
For i = 1, the probability of creating a contact place is r1 = 1 and hence
s1 = 0; the first interacting fiber creates a contact place on fiber No. 1 for
sure, because no other fiber prevents it to make a contact. In opposite to
this, if already the whole fiber No. 1 was completely occupied then the
probability of creating other contact place is equal to zero.
Like wise for the probability si, it is valid from Eq. (4.56)
1
s2 , si
i 1 s2 , si 1 si s2 .
z
(4.57)
Fiber-to-fiber contacts
119
(1) Let us imagine that e.g. eight fibers (sequence numbers I, II, ,VIII)
are interacting step-by-step with a given fiber No. 1. The interacting
fibers I, IV, VI and VII will contact fiber No. 1, but the fibers II, III, V and
VIII will not contact fiber No. 1. Fiber I surely create contact with
probability r1 = 1. Fiber II have a latent possibility to create the second
contact, but it does not to realize it; the probability of non-creation of
second contact is s2. It therefore stays only one contact on fiber No. 1.
Also, the fiber III can create the second contact, nevertheless it also does
not realize it; the probability of non-creation of second contact is again s2.
The next interacting fiber IV successfully creates the second contact with
the probability r2. Now, there are two contacts on fiber No. 1. The following
fiber V does not realize the possibility to create the third contact and the
probability of this is s3. The sixth and seventh fibers create step-by-step
the third and fourth contacts with corresponding probabilities r3 and r4,
respectively. Finally, the last, i.e. the eighth interacting fiber does not create
the fifth contact with fiber No. 1 and the probability of this is s5. Overall,
the probability of the mentioned configuration of contacts from eight
interacting fibers is the product of all individual probabilities and it can
r4 s5 . Nevertheless, this
be write down by the expression rI1 sII2 sIII2 rIV2 sV3 rVI3 VII
VIII
r1s2 s2 r2 s3r3 r4 s5 in our example and analogical expressions will be used for
each other configuration.
(2) If we create combinations with repetition of k-th class, from j 1
values s2 , s3 , , s j ( j + 2,3,
, z 1, ), then we obtain number of
j 1 k 1
j k 2
combinations equals to
k . Further, the values in
k
each combination will be multiplied and then the resultant products will
be added to obtain the final value which is denoted by the symbol kj .
Note: The following convention will be used in this part of text: a
superscript written by a Greek letter will have the sense of upper index,
not exponent; e.g. k in the symbol kj , but a subscript written by other
quantities means an exponent, e.g. s2k
s2 .
Example 4.3: Let us consider the values of s2, s3, s4 (i.e., the highest
index is j = 4), we create from them 2nd combinations (k = 2) with repetition.
k
120
4 2 2
4
The number of these combinations will be
6 . These
2 2
combinations are (s2, s2), (s2, s3), (s2, s4), (s3, s3), (s3, s4), (s4, s4). Thus, we
obtain the value of 24 s2 s2 s2 s3 s2 s4 s3 s3 s3 s4 s4 s4 .
If j = 2 then it is evidently valid that
k factors
s2 * s2 * * s2 s2k .
(4.58)
k
2
j + 2,3,
, z 1, .
0j 1,
(4.59)
Finally, if j > 2 then, from the definition of kj and from Eqs. (4.57)
and (4.59), we can write the following expression
s2
j
sj
2
k
k
k 1
j j 1
j 1 s2 j 1
j 1 s22 kj 12
s kj
(4.60)
k
k
i
j 1 s2k 0j 1
j 1 s2i kj 1i .
i 0
2
4
32
s2 s2 s2 s3 s3 s3
4 4
kj
s2k
k
j
s2k B kj .
(4.61)
B0j 1, B k2 1 .
6.
(4.62)
Fiber-to-fiber contacts
121
The following equation is obtained by applying Eq. (4.61) into Eq. (4.60)
B kj
i k i
k i
k
k
i s2 j 1
i j 1
1
1
j
j
'
(
'
(
s2k i 0
s2k i 0
s2k i
kj
k
(4.63)
j 1 B kj 1i .
i0
i
(4.64)
1 contact (y = 1). For realizing one contact place, the number x of all
possible interactions with fiber No. 1 must be at least 1, i.e., x = 1, 2, .
1 interaction (x = 1). The interacting fibers create further contact place
(r1 = 1). The probability of possible occurrence of such interaction is found
by applying Eqs. (4.33) and (4.62) as follows
K1 P
1 r1 P
0
m1 r1
r
sm
s2 P
0 1 B 02 2 1 .
s2 1 1!
1! s2
(4.65)
2 interactions (x = 2). The first interacting fiber creates for sure a contact
place (r1 = 1), 2nd interacting fiber will not realize a contact place. The
probability of this possibility is estimated by applying Eqs. (4.33) and
(4.62) as follows
K 2 P
2 r1 s2 P
0
m12 r1 2
r
s 2 m2
s2 P
0 1 B12 2 1 .
2! s2
s2 1 2!
(4.66)
3 interactions (x = 3). The first interacting fiber creates for sure a contact
place (r1 = 1). Neither the 2nd nor the 3 rd fiber creates contact places. The
probability of this possibility is found by applying Eqs. (4.33) and (4.62)
as follows
K 3 P
3 r1s2 s2 P
0
m13 r1 3
r
s 3 m3
s2 P
0 1 B22 2 1 .
3! s2
s2 1 3!
(4.67)
122
r1
s2
r1
0 s2 m1
s 2m2
s3m3
B2
B12 2 1 B 22 2 1
s2
1!
2!
3!
i 1 s2i m1i
' B 2 i ! (.
i 1
A
(4.68)
% 2 P
2 r1r2 P
0
m12 r1r2 2
r1r2 0 s22 m12 .
s
P
0
2 B3
2
2! s22
s2 1 2!
(4.69)
3 interactions (x = 3). The 1st interacting fiber creates for sure a contact
(r1 = 1). Then followed by a) the 2nd interaction fiber creates a contact place
and the 3rd does not or b) the 2nd interacting fiber does not create a contact
place and the 3rd fiber creates a contact place. The probability of this
possibility is found by applying Eq. (4.33), definition of kj and Eq. (4.61)
1
3
a)
b)
% 3 P
3 r1r2 s3 r1s2 r2 P
3 r1r2
s3 s2
B13
m 3 r r
1
rr
s 3 m3
P
0 1 1 22 s22 3 s2 P
0 1 22 B13 2 1 .
s2
3! s2
3!
s2
(4.70)
Fiber-to-fiber contacts
123
2
3
a)
b)
c)
2
% 4 P
4 r1r2 s3 s3 r1s2 r2 s3 r1s2 s2 r2 P
4 r1r2 s3 s2 s3 s22
m14 r1r2 2
32 2
r1r2 2 s24 m14
s
s
P
0
.
P
0
B3
2
2
4! s22 s22
4!
s22
B32
(4.71)
r1r2
s22
r1r2
0 s22 m12
s 3m3
s 4m4
B
B13 2 1 B 32 2 1
2 3
s2
2!
3!
4!
i 2 s2i m1i
'B3 i ! ( .
i2
A
(4.72)
&3 P
3 r1r2 r3 P
0
r1r2 r3 0 s23m13
m13 r1r2 r3 3
s
P
0
3 B4
2
3! s23
s2 1 3! .
(4.73)
124
5 interactions (x = 5). The 1st interacting fiber creates for sure a contact
place (r1 = 1). Then further a) the 2nd and 3rd interacting fibers create contact
places, the 4th and 5th do not create or b) the 2nd interacting fiber creates a
contact place, the 3rd do not create a contact place, the 4th creates a contact
place and the 5th does not create a contact place or c) the 2nd interacting fiber
creates a contact place, the 3rd and 4th do not create a contact place and the
5th creates a contact place or d) the 2nd interacting fiber does not create a
contact place, the 3rd and 4th create a contact place and the 5th does not create
any contact or e) the 2nd interacting fiber does not create a contact place, the
3rd creates a contact place, the 4th does not create a contact place, the 5th
creates a contact place or f) the 2 nd and 3rd interacting fibers do not create
any contact, the 4 th and 5 th realize contact. The probability of these
possibilities is estimated by applying Eq. (4.33), definition of kj and
Eq. (4.61) as follows
a)
b)
c)
d)
e)
f)
&5 P
5 r1r2 r3 s4 s4 r1r2 s3r3 s4 r1r2 s3 s3r3 r1s2 r2 r3 s4 r1 s2 r2 s3r3 r1s2 s2 r2 r3
24
m15
P
0
r1r2 r3 s 42 s3 s4 s 32 s2 s4 s2 s3 s 22
5!
B 24
5
m r r r
2
s 5 m5
rr r
(4.75)
P
0 1 1 23 3 s23 24 s22 P
0 1 23 3 B 24 2 1 .
5! s2
5!
s2
s2
Q
3 &3 &4 &5 P
0
r1r2 r3
0 s23 m13
s 4m4
s 5 m5
B
B14 2 1 B 24 2 1
3 4
3!
4!
5!
s2
(4.76)
4.5
125
Fiber-to-fiber contacts
y
Q
0 P
0 , Q
y P
0
Lr
i y
'B
t 1
y
2
i y
y 1
s2i m1i
,
i ! (
y 1, 2 . (4.77)
A
A
y
s i mi
s i mi
e s2 m1 1 ' 2 1 y ! Biyy1 ( y ! ' 2 1 B iyy1 (,
i y i !
i y i !
y 1, 2 . (4.78)
Q
0 P
0 , Q
y P
0
Lr
t 1
y
2
y! s
s2 m1
1 ,
y 1, 2 .
(4.79)
Further, rearranging and applying Eqs. (4.56) and (4.57) we find the
following equation
y
L rt
t 1
y ! s2y
L
1 st
t 1
y ! s2y
z y 1
L
jz
y!
y
y
1 st
t 1
z
z
L
L
z t 1
s2 L
z t 1
t 1 s2
t 1
y!
y!
y!
y
j
j z y 1
y!
j
Lj
j 1
z y
y! L j
z
z!
y !
z y ! y
(4.80)
j 1
126
P
0 e
s2 m1
1 e
m1
m1
z
m
m
1
1
z
e z
m1 1
m1 1 e
e
m1 e
m1
m1
y
z
e z
e z
e
1 e
m1
z
m1
m1
y
z
1 e
m1
z
e
y
m1 z y
z
(4.81)
By applying Eqs. (4.33), (4.80) and (4.81) into Eq. (4.79), we find
y
m
m
1
1
z
Q
0 e m1 , Q
y 1 e z e z
y
z y
y 1, 2 .
(4.82)
Although the second equation of the last expression is defined only for
y = 1, 2, substituting y = 0 and calculating, we find that Q
0 e m1 ,
which is in agreement with the first equation. Therefore, it is possible to
describe the distribution of number of contact places on fiber No. 1 by
one collective expression
y
m
m
1
1
z
Q
y 1 e z e z
y
z y
y 0,1, ,
(4.83)
q 1 e
m1
z
1 e
8 l
J
-1 ,21
z d
(4.84)
y 0,1, .
(4.85)
127
Fiber-to-fiber contacts
m1 y Q
y zq z 1 e
y 0
m1
z
z
1 e
8 l
J
-1 ,21
z d
(4.86)
1
1
m m1
0.5
z m1
4.7 Relative reduction in mean value of number of contact places on fiber No. 1,
according to the Eq. (4.86); m1 m1
z m1 1 e1
z m1 .
Note: Fibers will not prevent creating contact places if each contact
place on fiber No. 1 is infinitely small. In this case each interaction is
realized in successive contact places. The maximum number of contact
places zA occurs when the mean number of contact places on fiber No. 1
is m1 lim z 1 e m1
z A
m
and q lim 1 e m1
z A
y
1
m1
This is identical with the results obtained from Eq. (4.32) for calculating
the number of contact places on fiber No. 1.
Distribution of contact places C (y) on fibers.7 The probability of the
randomly selected fiber No. 1 belonging to the elementary directional class
limited by the interval (-1, -1 + d-1) and (21, 21 + d21), is w(-1, 21)d-1d21.
The probability dC (y) of this selected fiber fiber No. 1 belonging to the
considered elementary class and at the same time a number of y contacting
places lie on it, is the product of Q
y , w
-1 , 21 d-1 d21 . By applying
Eq. (4.84) we get
z
z y
dC
y Q
y * w
-1 , 21 d-1 d21 q y
1 q w
-1 , 21 d-1 d21 .
y
(4.87)
7.
128
The probability C(y) of the randomly selected fiber having just y contact
places is obtained by the integral of the previous expression. By applying
Eqs. (4.84) and (4.85), we find the distribution of contact places C(y) on
the fiber in the following form
z
z y
C
y = 55 dC
y 55 q y
1 q w
-1 , 21 d-1d21
y
31
31
y
m
m
1
1
z
55 1 e z e z
y
31
z y
(4.88)
w
-1 , 21 d-1d21
8 l
8 l
J
-1 ,21
J
-1 ,21
z
z d
e
55 1 e zd
y
31
z y
w
-1 , 21 d-1d21 .
y
Q
(
m1 y
m1 z y
z
(
z
z
y 1 e e w
-1 , 21 d-1d21 (
m1
55
31
8 l
J
-1 ,21
z d
,
2
d
d
2
1
y
Q
y
w
z
e
1
1
1
1
55
w
-1 , 21 d-1d21 .
y 0
31
(4.89)
z
8 l
8 l
z
C
y 55 1 e zd 4 e zd 4
y
31
y
2 l
2 l
z
1 e zd e zd
y
z
1 e
y
2 l
zd
e
2 l
zd
z y
z y
w
-1 , 21 d-1 d21
1
w
2
,
d
d
55 1 1 1 21
31
z y
(4.90)
Fiber-to-fiber contacts
129
By applying Eq. (4.43) into Eq. (4.89), we find the mean value of contact
places on a fiber as follows
1
8 l
2 l
m z 55 1 e zd 4 w
-1 , 21 d-1d21 z 1 e zd 55 w
-1 , 21 d-1d21
31
31
2 l
z 1 e zd .
(4.91)
8 l 2
8 l 2
z
C
y 55 1 e zd e zd
y
31
y
16 l
16 l
2
2
z
1 e zd e zd
y
y
2
2
z
1 e zd e zd
y
16 l
16 l
z y
w
-1 , 21 d-1d21
1
55 w
-1 , 21 d-1d21
31
z y
(4.92)
By applying Eq. (4.51) into Eq. (4.89), we obtain the mean number of
contact places on a fiber as follows
8 l 2
m z 55 1 e zd w
-1 , 21 d-1d21
31
1
16 l
2
zd
z 1 e
55 w
-1 , 21 d-1d21
31
16 l
2
z 1 e zd .
(4.93)
130
l
l
, d .
zd
z
(4.94)
0.03
K
x,
C
y
C
y , 1.5
0.5
C
y, 1
0.02
C
y , 0.5
K
x , l d 2000
0.01
0
1000
1500
2000 x, y
4.6
(4.95)
131
Fiber-to-fiber contacts
It is known that two contact places create one contact so that the number
n of contacts in the fibrous assembly is
n
M N m
.
2
2
(4.96)
&
n N m 4
2m .
Vc
2 N d 2l d 2l
(4.97)
The mean distance between neighboring contact places : (see Fig. 4.3)
is
:
Nl
Nl
l
.
M
N m m
(4.98)
Nl
1 e
d 55
31
8
J
-1 ,21
(4.99)
w
-1 , 21 d-1d21 .
n
M
Nl
2 2 d
8
J
-1 ,21
1
e
55
w
-1 , 21 d-1 d21 .
(4.100)
31
&
8 l
J
-1 ,21
2 m
2
z 55 1 e zd
2
2
w
-1 , 21 d-1d21
d l d l 31
2
d 3
8 l
J
-1 ,21
z d
e
1
55
w
-1 , 21 d-1d21 .
31
(4.101)
132
:
l
l
m
z
55 1 e
8 l
J
-1 ,21
z d
31
1
w
-1 , 21 d-1d21
(4.102)
d
.
8 l
J
-1 ,21
z d
55 1 e
w
-1 , 21 d-1d21
31
(4.103)
n
M
Nl
1 e2 .
2 2 d
(4.104)
&
2 l
2m 2 zd
2
1 e zd 3 1 e 2 ,
2
3
d l
d l
d
(4.105)
:
l
l
d
.
2 l
m
1 e 2
zd
z 1 e
(4.106)
2
Note: It is true that when 0, lim & 2 d lim 1 e 2
0
0
according to van Wyk in Eq. (4.16). Here, for spatial isotropic orientation,
it is the integral of I according to Eq. (4.44). From these two equations,
133
Fiber-to-fiber contacts
(4.107)
n
(4.108)
16 l
16
2
2
2m 2 dz
2
zd
1
e
1
e
d 3
,
d 2 l
d 3l
(4.109)
:
l
l
d
16 .
16 l
2
2
m
zd
e
z 1 e
1
(4.110)
Table 4.3 shows some typical values for density of contacts & and the
mean distance : between contact places. Comparing these values with
the values obtained from Table 4.2, we can conclude that the reduction in
& is a result of mutual obstruction of fibers. Obviously, the reduction of
& is accompanied by increasing in and decreasing in values.
Table 4.3 Values of & and : for isotropic orientation.
Material. Polyester fibers, t 0.17 tex , 1360 kg m-3 ( d 0.012616 mm )
Packing
0.5
0.5
1
1
density
0.7
0.5
0.3
0.1
0.01
& mm-3
: mm
& mm-3
: mm
223444
124746
49303
6034
63.09
0.012530
0.016032
0.024338
0.066287
0.633959
167202
100204
42914
5747
62.78
192191
105642
41062
4937
51.19
0.014568
0.018931
0.029222
0.081018
0.781376
150581
88041
36630
4745
50.98
0.016745
0.019958
0.027962
0.069598
0.637129
0.018594
0.022715
0.032758
0.084300
0.784543
134
4.7
References
1.
2.
3.
4.
5.
6.
7.
135
5
Compression behavior of fibrous assemblies
5.1
1.
2.
135
136
fiber material up to the height c0. Let us consider that the fiber packing
density in this structure takes a very small value 0.
Note: Here, the quantities related to the non-compressed free fiber
material are expressed with the subscript 0. But, the quantities related to
the compressed fiber material are expressed without this superscript.
Let us apply a pressure p (black arrow) onto this fiber material so that
it gets compressed (shown in Fig. 5.1b). Consequently, the height of the
column of fiber material decreases to a smaller value c, but the nondeformable rigid box maintains the constant dimensions a and b. The length
% = c0 c represents the whole trajectory of movement. Let us assume
that the resulting fiber packing density in the deformed structure be .
%
c0
p
c
%
c
b
a
(a)
(b)
(5.1)
The volume of air gaps among the fibers is reduced by the process of
compression, but we assume that the fiber volume V stays roughly the
same as in the non-deformable rigid box. (We think about a reasonably
moderate compression, not about hard briquetting). Then, the following
equations are valid to write
V abc0 0 abc abc , c c0 0 , c c0 0 .
(5.2)
137
The following derivatives are resulting from Eqs. (5.1) and (5.2)
d % d % dc
c
2
1 c0 0 0 20 .
d
dc d
(5.3)
p p , 0 p
0 , p p
.
(5.4)
dA p ab d% p
abc0 0
2
d V
p
2
(5.5)
A
0
dA V 5
d .
p
(5.6)
p
dA
d p
p
V
V 2 .
d
V
'5
(
2
2
d
d ' 0
(
(5.7)
y
Let us remember that the general mathematical formula d ' 5k f
x dx ( dy f
y
is valid, when the quantity y stays as the upper limit of the integral.
138
A
B
a)
c0
a
b
p
A
b)
L
V
4V
.
2
d 4 d 2
(5.8)
139
The length of the fibers (fiber segments) lying only in the direction 1
(i.e., in the elemental angular class ranging from 1 to 1 + d 1) is then
dL L f
1 d
4V
f
d 2
1 d
(5.9)
h0 : 0 h : , h
h0 : 0 :, : h :
1 h0 : 0
(5.10)
B
:0
F0
F0
a)
F0
h0
y0
h0
X
b)
A :
F
B
F
F
h
F
h
140
Let us note that h0/:0 is a parameter that characterizes the initial state of
the fibrous assembly so that the horizontal lengths h and : are mutually
proportional and independent to the level of compression which is, for
example, characterized by the value of packing density .
The forces F are acting on the fiber at the points of fiber-to-fiber contacts
as shown in Fig. 5.3b.
Note: At the initial state as shown in Fig. 5.3a, very small forces F0 can
also act, because the initial fiber is crimped a little. However, it is the
natural crimp of the fibers or the forces F0 which is compensated by the
internal friction of a real fibrous assembly. Therefore, the real outer pressure
is equal to zero at the initial state.
Strategy of model creation. The current model is based on the assumption
that the work done is proportional to the deformation energy required to
compress the fibers in accordance with Eq. (5.6). The deformation energy
can be determined by using the previous relations and following the strategy
mentioned hereunder.
(1) The number m1 (or m1 ) of fiber-to-fiber contacts on one fiber of
length l was derived earlier (Equation (4.49) or (4.86)). These equations
can be expressed by a general formula m1 l :
, 1 , so that the length
: of one fiber segment is a function of and 1 as shown below
: l m1 :
,
(5.11)
Then, the following equation is also valid for the initial length :0
:0 = :( 0,
).
(5.12)
Using Eq. (5.9) the number of fiber segments lying in the direction of
angle 1 is obtained as follows
dN
dL 4V f
: d 2 :
(5.13)
(2) We assume that the force F and the length : of fiber segment shown
in Fig. 5.3b are functions of the deflection y, the horizontal distance h,
and the state of compression, given by the value B, as shown below
F = F (y, h, B),
(5.14)
: : ;
y , h, B ,
(5.15)
(5.16)
B0 = B( 0).
(5.17)
141
The deflection y can be explicitly expressed from Eq. (5.15) in the form
of a general function as shown below
y = y (:, h, B).
(5.18)
1 y
F d0 .
2 50
This expression can be written as follows by applying the function
expressed in Eq. (5.14) using 0 in place of y, further the function expressed
in Eq. (5.18) and then Eq. (5.10)
that the whole deformation energy in one fiber segment is E:
6
7
7
8
E: E:;;
:, h, B E:; y
:, h, B , h, B , 7
E: E:
:, B E:;;:,
h0 : 0 :, B . 79
E: E:;
y, h, B
1
F
0, h, B d0,
2 50
(5.19)
dE E: dN E:
:, B
4V f
d 2 :
4V E:
:, B
f
d 2
:
1 d
(5.20)
1
1
dE .
2
(5.21)
142
E0 E
: 0 , B0 .
(5.22)
(Because :0 and B0 are not the functions but the initial parameters, the
energy E0 is a parameter, too).
So, the increment of the deformation energy, in consequence of
compression, as shown in Fig. 5.1, is expressed as follows
/E = E E 0 = E (:, B) E0.
(5.23)
(4) The variables : and B are the functions of according to Eqs. (5.11)
and (5.16) so that the derivative of /E with respect to is
d/E dE
:, B
.
(5.24)
d
d
(5) The increment of deformation energy is equal to the work done in
the case of so-called conservative system. Such theoretical system is not
really applicable to fibrous assemblies, because the dissipation of energy
by mechanical straining of such materials is often significant. Let us
therefore introduce a more general assumption that the increment of
deformation energy /E is proportional to the work done A in our case.
Using C as the coefficient of proportionality the relation /E = CA is valid.
After differentiating this expression with respect to packing density and
applying Eqs. (5.7) and (5.24) we obtain
d/E
dA dE
:, B
p
,
C
CV 2 ,
d
d
d
p
2 dE
: , B
.
CV
d
(5.25)
The last equation states that the pressure p is a function of packing density .
5.2
Length and number of fiber segments. The (mean) number of contact places
m 1 on a fiber of length l is des cribed by the expressio n
m1 l 8
d J
1 according to Eq. (4.49)5 in Section 4.3. Beside the
fiber length l, this variable depends on the packing density , the equivalent
fiber diameter d, and the angle 1 which is expressed in the integral J ( 1)
given by Eq. (4.47). In accordance with Eq. (5.11), thelength of fiber
4. Here, we use our own way for the derivation of the traditional results of van Wyk.
5. This equation corresponds to the idea of van Wyk; see Section 4.1.
143
:
l
d
m1 8 J
(5.26)
Regular beam. The following relations specify the force F and the length
of fiber segment : in terms of deflection of a regular beam
F F
y, h
y
kF ,
h3
(5.27)
y 2
: : ;
y, h h '1 q ( ,
' h (
(5.28)
yh
1
:
1 .
qh
(5.29)
1
1 0
kF
kF y2
0
0
0
0
0
F
,
h
d
k
d
d
,
F
2 50
2 50 h3
2h3 50
4h3
1
:
' h 1 (
' q h (
h
: 1
k F 2 1
: k F : h k F :
,
3 h 1
4h
4q
q h 4q h 2
h2
k
E: E:;;
:, h E:; y
:, h , h F3
4h
144
h : :
: 1 0 0
:
k F 1 h0 : 0
k
.
E: E:
: E:;;:, h
h0 : 0 : F
2
4q
h0 : 0 :
4q
h0 : 0 2 :
(5.30)
Increment of the deformation energy. Using Eqs. (5.20) and (5.30) the
following expression is valid to write for the deformation energy in all
fiber segments having the direction 1
k F 1 h0 : 0
2
4V 4q
h0 : 0 :
f
d
1
1
d 2
:
4V E:
:
f
d 2 :
k V 1 h0 : 0
f
F2
d q
h0 : 0 2 : 2
dE
1 d
1 d
(5.31)
E E
:
1
dE
1 2
2
k FV 1 h0 : 0
d 2 q
h0 : 0 2
1
f
2
:
2
1 d
(5.32)
k FV 1 h0 : 0
d 2 q
h0 : 0 2
2
1
f
2
:
2 0
1 d
(5.33)
k FV 1 h0 : 0
d 2 q
h0 : 0 2
2
1 1
: 2 : 2 f
0
2
1 d
(5.34)
Using Eqs. (5.24), (5.32) and (5.26), the following derivative is valid
to write
d/E dE
: k FV 1 h0 : 0
2
d
d
d q
h0 : 0 2
k V 1 h0 : 0
F2
d q
h0 : 0 2
k V 1 h0 : 0
F2
d q
h0 : 0 2
2
d
1
f
d : 2
2
d 64 J
2d 2
d
2
128 J 2
2d 2
2
1 d
2
2
2
128k V 1 h0 : 0
3 4F
5 J2
2
d q
h0 : 0 2
1 d
f
f
1 d
1 d
145
(5.35)
Pressure. Finally, applying Eq. (5.35) into Eq. (5.25) we find the
expression for the pressure p as shown below
2 dE
: 2 128k FV 1 h0 : 0
J2
C V d
C V 3 d 4 q
h0 : 0 2 5 2
2
p
2
128k F 1 h0 : 0
J2
'
C 3 d 4 q
h0 : 0 2 ' 5 2
f
1 d
f
1 d
3
( .
(
(5.36)
128k F 1 h0 : 0
C 3 d 4 q
h0 : 0 2
2
2
J2
f
1 d
(5.37)
so that
p = k p 3 .
(5.38)
2
4 1
d
2
2
512 k F 1 h0 : 0
5 d 4 q
h0 : 0 2 .
(5.39)
146
when the packing density was low, especially around 0.2 or 0.3. But, for
moderate values of packing density (e.g. when the value of packing density
was around 0.5 which is typical for common yarns), this equation was not
found too precise and at high compression the equation was found to be
quite false. Also, there are difficulties for extremely small compression
where -values are very near to the initial value 0 of free fiber
materials 6 . Summarily, there are two problems encountered with Eq. (5.38):
(1) problem at very high pressure and (2) problem at very low pressure.
Problem at very high pressure. The behavior of the resulting function
expressed in Eq. (5.38) is shown in Fig. 5.5a. It can be seen that the packing
density is increasing with the increase in pressure, and when p = kp the
corresponding packing density is equal to one. Then, at p > k p the
corresponding packing density > 1, which is logically nonsense,
evidently. The obvious question is why did we obtain such result? It appears
that we did not think about the effect of mutual obstruction experienced
by the fibers inside the fibrous assembly. Our model fibers were
mechanically understood as infinitely thin beams having point contacts
with supports (fiber-to-fiber contacts). Such a volume-free structure is
possible to be compressed to the thickness c = 0, as shown in Fig. 5.1b, so
that the total volume is limiting to zero. But, we still geometrically assumed
the constant volume V of fibers during the whole process of compression.
Therefore, the packing density must be limited to infinity (see later on).
0.06
?
0.02
0
0.5
0.02
0
0
0.5
p kp
1.5
(a)
30 0.0001 0.0002
p kp
(b)
Typically, 0
0.01, 0.03 .
147
p k p 3 30 .
(5.40)
5.3
Non-compressible region. The traditional concept interprets fiber-tofiber contacts as contact points (black points shown in Fig. 5.6); then the
whole fiber length is free for bending deformation. Nevertheless, in reality,
each fiber-to fiber contact occupies an area of contact. The fibrous material
is compressed at the surrounding of the contact area and such surrounding
creates a no-more-compressible region, something as a non-compressible
solid inside the fibrous assembly see gray color region in Fig. 5.6. The
more deformable fiber material is created only by free fiber segments lying
outside the mentioned non-compressible regions.
Empirical idea of modification. Wyks Eq. (5.38) can be rearranged by
using the definition of packing density according to Eq. (1.22) as follows
p k p 3 k p
V3
,
Vc3
(5.41)
148
where the common parameter kp and the total volume of fibers V are
considered as suitable constants. It is evident that the increasing pressure
p is the function of the decreasing total volume Vc only. (If the pressure is
infinitely high then the total volume must limit to an impossible value
equal to zero). In spite of them, let us empirically assume that van Wyks
equation could be right, but only for a deformable region; it is the total
volume V c of fibrous assembly minus the total volume of all noncompressible volumes Wc. Thus we obtain
p kp
V3
Vc Wc 3 .
(5.42)
F
a)
F
b)
Fiber volume in non-compressible region. Let us take that one noncompressible region occupies a total volume W1. The fiber volume inside
the region is W f,1 so that the packing density inside the region is
s Wf,1 W1 . (We can expect that s is very near to the value 1. The air
gaps between fibers are probably very small inside the non-compressible
region). We can also write
W1 Wf,1 s .
(5.43)
Wf,1 f
K a ,
149
(5.44)
where K and a are two suitable parameters. Such function can describe a
convex bow, a concave bow, and a straight line as shown
Wf,1 a " 1 Wf,1 a < 1 Wf,1 a 1
.
k& 2 aVc
.
s
(5.45)
(3) The non-compressible regions fully occupy the total volume of the
fibrous assembly; Wc = Vc,min.
Note: Let us imagine that the non-compressible regions do not
fully occupy the total volume of fibrous assembly. Then, Wc < Vc,min,
so that the deformable volume (V c,min W c) > 0, and the fibrous
assembly can be further compressed. However, it is not the limit state
as it was assumed to be.
(4) The packing density of the fibrous assembly is the same as the packing
density of the non-compressible regions; m = s. (Fibrous assembly
is created only from non-compressible regions. This is analogous to
the wall created from bricks).
Applying the quantities of the limit state to general Eq. (5.45) the
parameter K can be evaluated as follows
Vc,min K
, K
1
.
k & 1m a
(5.46)
150
Wc
1 k& 2 aVc
2 a
V
c
.
s
1m a s
k& 1m a
(5.47)
Pressure. In the limit case, the packing density inside the noncompressible region follows the point (4). In a general case, the packing
density inside the non-compressible region need not to be precisely equal
to m, but, for simplification, let us assume that the relation s = m is
valid independent to the level of compression. Using this equality in
Eq. (5.47) we obtain
Wc Vc
2 a
2 a
Vc
m .
1 a
m m
(5.48)
Finally, let us use the last equation in our modified idea expressed by
Eq. (5.42) and then we obtain
p kp
V3
Vc Vc
m 2 a
kp
3
1
m 2 a
(5.49)
m 1
Eq.(5.38)
Eq.(5.49)
0.5
0.5
p kp
1.5
151
where x
3m k p
p
(5.50)
b 3 1
1 a m
G
c 1 H 1 m
I
2 a
2 a 3
1
m
b 3
6
8,
9
2 a
,
(5.51)
(5.52)
and then the approximated equation around * takes the following form
p = k pc b.
(5.53)
152
2
0.6
0.4 4
m 1,
0.2
a 1
0.2
0.4 p k p
) and approximated (
) results.
Problem at very low pressure. The problem at very low pressure is also
actual in this model, because at = 0, the pressure according to Eq. (5.49)
is very negligible, but it is higher than zero. So as to obtain precisely the
zero value for = 0 we need (empirically) to correct Eq. (5.49) in the
following manner
p k p 2
2
0 ,
(5.54)
where
2
3
1
m 2 a
, 2
0
3
30
1
0 m 2 a
(5.55)
5.4
8 l
J
-1 ,21
z d
according to
153
Eq. (4.86), where l/z = d according to Eq. (4.94). Here, z is the maximum
possible contact places on a fiber of length l, d expresses the length of
the fiber segment that is covered by one contact (the length occupied by
one contact place is equal to times fiber diameter d), and the integral
J (-1, 21) is reduced to the integral J
1 , 1
2 , 2 according to
Eq. (4.47) for planar orientation. Using all these expressions the following
expression is valid to write for m1
m1
8
8 l
J
1
J
1
l
l
l
1
e
d
.
d
d
d 1 e
(5.56)
:
l
m1
d
1 e
8
J
d
:0
1 e
8
0 J
, 1
, 1
8
J
d
e
:
8
0 J
d
e
:0
(5.57)
(5.58)
(:0 is the length of fiber segment at the initial position, that is, before
compression, when = 0).
Note: Let us denote x 8 J
Eq. (5.57) : l
m1
lim
x0
x
d
e 1 8 J
x
, so that 8 J
. Then from
1
x
d
8 J
1
d
x
1 e
8 J
d
x and
lim :
0
e 1
8 J
154
C
A
u
d
in packing density of the fibrous assembly7 . (At the initial stage, that is,
at = 0, we imagine the gripped length is reduced to a point). Thus
u Bd , B B
1 ,
(5.59)
and especially
B0 = B( 0).
(5.60)
u
0
hu
h
hu
h
:u
5.11 Regular fiber as an infinity long beam with regularly distributed grips.
Instead of exact contacts between the beam and the supports and the
exact action of the force, the new fiber beam has a firm and straight
support and force-grips8 along the width u. The fiber segments inside the
grips of length u are considered to be straight. The segments outside the
grips of length u are bent and we assume that Eqs. (5.27) and (5.28) are
valid analogically. Nevertheless, in place of earlier h we must use h u
and in place of : we must use : u. Then, we obtain
F F
y , h, B
7.
8.
h u
kF
h Bd 3
kF ,
(5.61)
155
y 2
: u
h u '1
q( ,
' h u (
y 2
: : ;
y, h, B
h Bd '1
q ( Bd ,
' h Bd (
(5.62)
y y
: , h, B
(5.63)
E: E:;
y, h, B
kF
2
h Bd
1
1
0
F
0, h, B d0 5
k F d0
5
20
2 0
h Bd 3
y
0 d0
3 5
0
kF y 2
4
h Bd
E: E:;;
:, h, B E:; y
:, h, B , h, B
1
: Bd
1
h
B
d
3
4
h Bd q h Bd
kF
kF
4
h Bd
h Bd 2
1
: Bd k F
: Bd
h Bd
1
q h Bd 4q
h Bd 2
kF : h
,
4q
h Bd 2
E: E:
: , B E:;;:, h
h0 : 0 :, B
:
1 h0 : 0
kF
.
4q
h0 : 0 : Bd 2
:
h0 : 0 :
kF
4q
h0 : 0 : Bd 2
(5.64)
156
:
1 h0 : 0
kF
2
4V E:
:, B
4V 4q
h0 : 0 : Bd
dE 2
f
1 d 1
f
d
:
d 2
:
k V
1 h0 : 0
f
1 d 1.
F 2
qd
h0 : 0 : Bd 2
1 d
(5.65)
E E
:, B =
1
dE
1
k FV
qd 2
2
5
h
2
1 h0 : 0
: 0 : Bd
1 d
. (5.66)
E0
1
k FV
qd 2
dE
2
k FV
qd 2
2
5
h
2
1 h0 : 0
5 h
2
2
B 0 d
1 h0 : 0
: 0 : 0 B 0 d
1 d
1 d
(5.67)
k FV
qd 2
1 h0 : 0
1 h0 : 0
5
h : : Bd 2 h B d 2 f
2
0
0
0 0
2
1 d
(5.68)
Derivatives. The derivative of the deformation energy E with respect to
is necessary to know for the determination of pressure according to
Eq. (5.25). The following expression is valid according to Eq. (A4.10) in
Appendix 4.
157
d
q d 2 7
7
I7
1
1
1 h0 : 0 d
h0 : 0 2 :
d
1
:
8J
1 dB 6
:
h0 d 7
1
77
d :0
8f
3
7
d : 0
7
1 B : h
0
97
1 d
(5.69)
Pressure. Using the last expression in Eq. (5.25) we obtain the following
expression for the pressure p as a function of packing density
p
2 dE
:, B
CV
d
G
7
7
2 2 7
16k F
HJ
Cq 2 d 4 5 2 7
7
I7
1
1
1 h0 : 0 d
h0 : 0 2 :
d
1
: 8J
1 dB 6
: h0 d 7
1
77
d :0
8f
3
7
d : 0
7
1 B : h
79
0
1 d
(5.70)
Limit case. Here, the limit case denotes the theoretical maximum
compression. In this situation, the following statements are valid to write:
(1) the packing density is maximum, m, (2) the length of the fiber
segment is minimum, : :m, (3) the distance h is minimum, h hm, the
quantity B is maximum, B = Bm. According to Eqs. (5.57), (5.10) and (5.59)
it is valid to write
d
:m
1 e
8 m
J
(5.71)
h0 : 0 hm : m , hm
h0 : 0 : m ,
(5.72)
Bm B
m .
(5.73)
158
d : 0
d : 0
1 Bm
0,
lim 1 B
m
: h0
: m h0
6
7
7
7
7
8
7
7
7
7
9
d
h0
Bm
,
:0
:m
d : 0
d : m
B :m
1 B
1
,
: h0
: d B m
Bm :
1 B
1
1
1 :m
.
: h0
:
d B m :
Bm
d :0 d
:m
(5.74)
Applying these relations to Eq. (5.70), the expression for the pressure
is obtained as follows
G
7
16k F
7J
p
H
Cq 2 d 4 5 2 7
7
I
2
h0
d
: m dB d 6
1 : d 1 : 1 8 J
: B 7
1
m
7f
0
1
8
2
3
:
h0
B :m
7
1
7
B m :
: 0
9
2
1 d
(5.75)
Value (). It is valid to write the following expression from Eqs. (5.71)
and (5.74)
d
h0
h
:m 0
:0Bm
: 0Bm
8 m
J
d
1 e
8
m J
, e
8 m
J
1
,
h0
J
8
m
1 : B
0 m
ln
1
ln
h0
,
: 0B m
1
.
h0
1 : B
0 m
(5.76)
d
:
note mentioned after Eq. (5.58) it is valid lim
0
8 J
d
: m lim lim : lim '
' 0 ( m A 8 J
m A
and especially
( 0 . Further, th e express io n
1
159
d
0d
Bm
:m
0 is the indefinite expression now; its value is equal to the
Bm
Cq 2 d 4 5 2 7
7
7
I
1
1
1 h0 : 0
h0 : 0
2
128k F 1 h0 : 0
J2
'
Cq 3d 4
h0 : 0 2 ' 5 2
d
d
8 J
0d
d
8 J
0
B
1
Bm
d
8 J
f
1 d
d
1
1
h0
d
8 J
1 : 8 J
1
0
1
dB 6
d 7
7
7
8f
7
7
7
9
3
3
( k p .
(
1 d
(5.77)
This result is identical with that obtained from Eqs. (5.36) to (5.38),
describing the traditional van Wyks model.
Case of planar isotropic orientation. As a special case of planar isotropic
orientation of fibers it is valid that f ( 1) = 1/ according to Eq. (3.26),
and J ( 1) = 2/ according to Eq. (4.51). According to Eq. (5.57) it is valid
to write
d
:
1 e
8 2
d
1 e
16
2
d
1 e
:
16
2
(5.78)
:m
d
1 e
8 m 2
d
1 e
16 m
2
d
,
1 e
:m
16 m
2
(5.79)
The following equation for pressure p is valid to write from Eq. (5.75)
G
h0
d
2 : m dB d 6
7 1 1 1
7
: 16 : B m 7 1
16k F
7 2 :0 d
p
H
8 d
3
Cq 2 d 4 5 2 7
h 2 :
B
:
7
0
m
1
B :
7 : 0
7
m
I
9
2
2
160
h0
d
2 : m dB d
1 1 1
: 16 : B m
32k F : 0 d
p
3
Cq 3d 4
h 2 :
B :m
0
1 B :
:
2
512k 1 h : 2 d
F
0
0
'
2 (
5 4
' Cq d
h0 : 0 ( 16 :
d
16 :
2
p kp
d
2 : m dB d
1
: 16 : B m
B :m
1 B :
m
d
2 : m dB d
1
: 16 : B m
B :m
1 B :
m
2.
2 ,
(5.80)
128k F 1 h0 : 0
p
Cq 3 d 4
h0 : 0 2
2
2 2 1
d
' 5
' 2
3 512k 1 h : 3
F
0
0
k p 3.
( '
2 (
5 4
' Cq d
h0 : 0 (
(
(5.81)
Note: In the last two equations, the symbol k p for planar isotropic
orientation was used in accordance with the expression mentioned in
Eq. (5.39).
Force in fiber-to-fiber contact. The force F in one fiber-to-fiber contact,
shown in Fig. 5.3 b or Fig. 5.11, was introduced by Eq. (5.61). Applying
Eq. (5.63) in Eq. (5.61) we obtain the square of the force as follows
1
: Bd
k F2
2
'
1
h Bd (
6
h Bd q h Bd
h Bd
: Bd h Bd
h
1 :
:h
h Bd
:
2
2
k
k
k F2
F
F
4
5
5
q
h Bd
q
h Bd
d
h
q B :5
:
:
h
1
2
:
(5.82)
.
kF
5
d 4
h
q B :
:
:
F 2 y2
k F2
161
Substituting the ratio h/: by the ratio h0/:0 according to Eq. (5.10), we
can write
1
F 2 k F2
h0
:0
5
h
d
q 0 B :4
:
:0
4
k F2
h0
:0
k
d
qd 4 :
h 5
0
:
0
2
F
h0
:0
h
d : 0 4
:
q 0 1 B
: h0
:0
1
d : 0
1 B : h
0
(5.83)
Finally, we apply the second expression from Eq. (5.74) in the last
equation and we obtain
h0
4 1
2
:
k
d
0
F 2 F4
qd :
h 5
0
:
0
B :m
1 B :
m
h0
2
:0
d
kF
52
q d 2
h0 :
:
1
F
1
B :m
1 B :
m
52
(5.84)
162
of our fibers see Fig. 5.10) as a function of applied force F, stated below
u CF 3 F ,
(5.85)
h
1 0
'
(
2
:0
d
1
' k
(
Bd C F ' F 2
52
52 (
:
q d
h0
B :m (
'
1
B m : (
:0
'
k 1 3
1 h0 : 0
d
CF 1 6 F 2 3
q d
h0 : 0 5 6 :
16
1 h0 : 0
B
k1 3
CF 1 6 F5 3
Bm
q d B m
h0 : 0 5 6
16
B
B :m
1
B m B m :
56
23
1
B :m
1 B :
m
1
d
:
23
56
1
B :m
1 B :
m
k1 3
1 h0 : 0
CF 1 6 F5 3
q d B m
h0 : 0 5 6
16
1
d
:
56
23
(5.86)
For each given value of packing density (this is present in the expression
for the quantity :) the corresponding value of B/Bm can be found out by a
suitable numerical way.
Numerical determination of B/Bm. For a given value of packing density
let us write
1 h0 : 0
k F1 3
16 53
q d B m
h0 : 0 5 6
16
CF
1
d
:
23
Q
parameter,
(5.87)
and also
:m
c
parameter,
:
(5.88)
B
x
obtained value.
Bm
(5.89)
163
Then we obtain
x
1 xc
56
Q
(5.90)
Let us express the function of the left-hand side of the last expression
by y as follows
x
1 xc
56
y
(5.91)
dy
0 1 xc
dx
x
56
x
5
c
6 1 x c
16
, 1 x c
5
x c,
6
6
0.5454
!
,
11c
c
y x 1 x c
56
(5.92)
6
6
1
11c
11
56
0.2828
.
c
(5.93)
y
y
Q
Root
(real)
0
Root
(unreal)
x
1c
164
than the value x*) is not real10 . Therefore, if Q y* then we must find the
true root in the interval
0 , x (e.g. using the algorithm of interval
halving). The finally obtained value x B B m (shown by symbol / in
Fig. 5.12) is the solution of Eq. (5.86).
Enlargement and empirical generalization of B
B m . The quantity Q
according to Eq. (5.87) has the character of a parameter for only one given
value of the packing density , which is present in the expression for : see
Eq. (5.57) and/or Eq. (5.78). Therefore, let us introduce the parameter K
at the first instance
k F1 3
1 h0 : 0 ,
16 53
q d B m
h0 : 0 5 6
16
K CF
(5.94)
so that
QK
1
d
:
23
(5.95)
56
K
1
d
:
23
(5.96)
165
1
B
Bm
C
B
0.5
A
0.5
dB d
a1b1 b1 1 (1 a1 )b2 b2 1 ,
Bm
(5.97)
166
6
7
for the curve A (K 0.1), 7
77
B B m 0.57 0.7 0.43 5
8
for the curve B (K 0.2), 7
7
B B m 0.89 0.7 0.11 3
7
for the curve C (K 0.3). 79
B B m 0.27 0.7 0.73 8
(5.98)
1
W
A
B
C
0.8
0.6
0.4
MW
0.2
0
0.2
0.4
0.6
0.8
p kp
6
7
for the curve D (K 0.3), 7
77
B B m 0.58 0.65 0.42 5
8
for the curve E (K 0.45), 7
7
B B m 0.82 0.7 0.18 9
7
for the curve F (K 0.57). 79
167
(5.99)
1
B
Bm
0.5
0.5
1
0.8
D
E
F
0.6
MW
0.4
0.2
0
0.2
0.4
0.6
0.8
p kp
168
Also in the previous example, the thin curves represent the classical
van Wyks function shown in Eq. (5.38) curve W, and the modified van
Wyks function shown in Eq. (5.49) when m = 1 and a = 1 curve MW.
Also, it is possible here to show practically the same behavior of the
thick curve F (when K = 0.57) and the earlier derived thin curve MW.
Problem at very low pressure. There exists a problem at very low
pressure in this model, too. At = 0 it is valid that : = :0 according to
Eq. (5.58) and B = B0 according to Eq. (5.60). The pressure obtained
according to Eq. (5.75) is very negligible, but it is higher than zero. In
order to precisely obtain the zero value for = 0 we need (empirically)
to correct Eq. (5.75) as follows
p
16k F 2
Cq 2 d 4
2
1 h :
5 J
h :
0
0
2
2
0 f
1 d
(5.100)
where
d
:
0 :
d
: m dB d
1
: 8 J
1 : B m
B :m
1 B :
m
d
: m
dB d 0
1
:0 8J
1 :0
Bm
B0 : m
1 B :
m
0
(5.101)
p kp
2
16
0 2 ,
where
d
:
d
2 : m dB d
1
: 16 : B m
B :m
1 B :
m
(5.102)
0 :
d
2 : m
dB d 0
1
: 0 16 : 0
Bm
B :m
1 B :
m
0
169
(5.103)
5.5
There exists hardly any literature reporting the relation between pressure
and packing density from extremely small to extremely high values. One
170
of such rare work is due to Baljasov [2]. He compressed the fiber materials
in a rigid metal box whose length was 60 mm and width was 50 mm.
Every time just 20g of fiber material was put inside the box and this material
was compressed. He used a tensile testing equipment from the textile
laboratory for measurement at smaller values of pressure and employed a
pressing machine from the engineering laboratory for measurement at high
values of pressure.
The primary measured quantity was the height of the fiber material in
the rigid box at a given value of pressure. The area of the box multiplied
by the measured height was equal to the total volume Vc of the tested fiber
material. Baljasov published the values of these volumes together with
the pressure values in his book [2].
We calculated the fiber volume V from the known relation V = m/,
where the mass of the fiber material was m = 20g and the value for the
fiber density was taken from Table 1.1. In this way, we determined the
actual values of for the packing density = V/Vc for all the pressure values.
Baljasov [2] investigated four different types of fiber materials. They
were (a) cotton, fibers 108F11 of staple length 31/32 mm, (b) viscose fibers
of 0.17 tex fineness and 38/40 mm length, (c) merino wool 64s, middle
length 67.7 mm, and (d) polyester fibers 12 of 0.21 tex fineness and
38/40 mm length. Baljasov studied two types of fiber directional
organization for each material. He put in the rigid box either chaotically
organized fibers (roughly isotropic orientation) or roughly parallel
organized fibers (significant preferential direction).
The experimental results of Baljasov [2] represents the points ( and )
in the semi-logarithmic graphs, from (a) to (d), in Fig. 5.17. It is shown
that the different types of fiber materials, the orientation does not play too
significant (systematic) role.
Also, three types of theoretical curves are displayed in Fig. 5.17. The
thin curves marked by number 1 correspond to the original van Wyks
model Section 5.2, the thin curves marked by number 2 correspond to
the empirically modified van Wyks model Section 5.3, and the thick
curves marked by number 3 correspond to the alternative model Section
5.4 (Curves 2 and 3 are practically identical which was already mentioned
in Fig. 5.16, curve F).
It appears that the experimentally found trends are in agreement with
the behavior of the curves 2 and 3, while the behavior of the curve 1 is
markedly different at higher values of packing density. It means that the
empirically modified van Wyks model as well as the alternative model
11. Former Soviet system of description of cotton variety.
12. Former Soviet polyester fibers Lavsan.
171
can describe this compression or the used fibrous material was quite good.
However, another experimental research in the field of high values of
pressure and packing density is required to be studied in future.
1
0.8
1
0.8
cotton
1
0.6
viscose
1
0.6
0.4
0.4
2
0.2
p MPa
0
0.0003 0.0196
0.9807 19.61 98.07
0.0029 0.1667 4.903 49.03 321.6
0.2
p MPa
0
0.0003 0.0196
0.9807 19.61 98.07
0.0029 0.1667 4.903 49.03 321.6
(a)
1
0.8
(b)
wool
1
1
0.8
0.6
0.6
0.4
0.4
2
0.2
p MPa
0
0.0003 0.0196
0.9807 19.61 98.07
0.0029 0.1667 4.903 49.03 321.6
(c)
polyester
0.2
2
p MPa
0
0.0003 0.0196
0.9807 19.61 98.07
0.0029 0.1667 4.903 49.03 321.6
(d)
Note: Many attempts were made to solve the relation between pressure
and packing density as a masters research work at the Technical university
172
5.6
x1
11
22
x2
22
1 N1
x3
1 N 2
1
Basic relations. Let us imagine the initial fibrous assembly in the form
of a unit cube as shown in Fig. 5.18. The normal forces 11
, 22 act on
the unit area of the initial fibrous cube.
Note: Traditionally, the vectors of positive values 11
, 22 direct outside
from the body (tensile loads). Nevertheless, the vectors of forces direct
173
inside of the body (pressure) as shown in Fig. 5.18. Therefore, the negative
symbols 11, 22 are used there.
Generally, the tensile forces 11
, 22 act on the unit area of the starting
fibrous cube so that they are also called as engineering or nominal (fictive)
stresses at the same time.
In consequence of acting forces, the initial unit dimensions of the starting
cube are changed (deformed) to new dimensions 1 + N1 and 1 + N2 as shown
in Fig. 5.18 (The values of N1 and N2 are negative for compression). Only
the dimension in the direction of x3- axis stays equal to one13 .
After deformation the forces 11
, 22 act on the real areas 1(1+N1) and
1(1+N2). The following expressions are then valid for Cauchys (real) stresses
11
11
22
11 , 22
22 .
1
1 N 2 1 N 2
1
1 N1 1 N1
(5.104)
All the time we assume that the volume V of fibers stays the same after
compression of the initial cube (Only the volume of air is decreasing due
to compression). Then the starting value of the packing density is
0 V
1*1*1 V ,
(5.105)
1
1 N1
1 N 2
0
.
1 N1
1 N 2
(5.106)
0
, 2 0 , i.e. 1 N , 1 N . (5.107)
1
0
1
2
0
2
1 N1
1 N2
Note: While and 0 have the logical sense of packing densities, the
quantities 1 and 2, determined by Eq. (5.107), do not have this sense.
They can only be thought as suitable quantities expressing the deformation
in another way.
The following expression is valid from Eqs. (5.106) and (5.107)
1 2
13.
0 0
0 .
1 N1 1 N 2
(5.108)
It is possible to obtain, for example, by placing the initial cube between two
rigid plates. However, this dimension will not probably much change from the
starting value 1 in case of a roughly parallel fiber bundle.
174
d 1
d
0
0
12
,
dN1 dN1 1 N1
1 N1 2
0
1 N 2
d 2
d
0
dN 2 dN 2 1 N 2
22
0
O
0
0
1 6
d
,7
'
(
2
ON1 dN1
1 N1
1 N 2
1 N1
1 N 2
0 7
8
O
0
0
2 7
d
.
'
(
ON 2 dN 2
1 N1
1 N 2
1 N1
1 N 2 2
0 79
(5.109)
(5.110)
(5.111)
OE
OE
dN1
dN 2 .
ON1
ON 2
(5.112)
The elemental increment of the work done is easy to derive with a view
to Fig. 5.19. The two forces (engineering stresses) 11
, 22 act on the fibrous
body in a general moment. In consequence of the force 11
the body extends
its strain N1 to the elemental increment dN1. Similarly, in consequence of
14.
A part of work done is usually dissipated in the form of thermal energy due to
fiber-to-fiber friction, etc., so that C < 1 . Only for the conservative system it is
valid that C 1 . Our assumption, given by Equation (5.111), bears then a little
looser sense than that meant by the strict conservative system.
175
11
dN1
22
1 N1
1
1 N 2
dN 2
the force 22 , the body extends its strain N2 to the elemental increment
dN2. Then the increment of the work done is
dA 11
dN1 22 dN 2 .
(5.113)
(5.114)
The last equation is valid for all couples of values N1, N2, so that the
following two equations must be right
OE
1 OE 6
dN1 C 11
dN1 , 11
,
ON1
C ON1 77
8
OE
1 OE 7
dN 2 C 22 dN 2 , 22
.
ON 2
C ON 2 97
(5.115)
(5.116)
(5.117)
176
(5.118)
Similar behaviors are also shown by other materials. Table 5.1 illustrates
the estimated values of 3.
Table 5.1 Estimated values of 3 according to experimental results of Gurov [12].
Estimated
value 3
for about
> 0.05
Polyester
Polypropylene
Polypropylene
Polypropylene
Kevlar
Fineness
[dtex]
Length
[mm]
6.7
6.7
17
70
2.4
80
60
85
90
40
0.57
0.40
0.32
0.65
0.30
1.2
0.9
0.6
0.3
0.05
0.10
0.15
0.20
177
(5.119)
(5.120)
(5.121)
0
02
0
1
f
f
0 , f
0 0 . (5.123)
)
E f
0 0 0 (
0
'
C
0
0
C
The expressions given in Eqs. (5.117) and (5.123) are valid as a special
relation mentioned in Eq. (5.22) in the case of uniaxial deformation (in
the direction x1), then
6
7
7
8
20
0 1
1
7
22 ') f
0
E f
( E f
.7
C
0
0 C
9
11
2
2 1
2
1
)
E
,
f
f
f
'
(
C
0
0 C
0
(5.124)
178
11
(5.125)
6
7
7
8
0
2 1
2 7
1
22 ') f
0
E f
.
( E f
C
0
0 C
0 79
11
2 1
2
1
E f
,
') f
( f
C
0
0 C
0
(5.126)
Note: Comparing Eq. (5.124) with Eq. (5.126) we can see that 11
11
in the case of uniaxial deformation. Nevertheless, it was already found in
Eq. (5.117).
The ratio 3 defined according to Eq. (5.118) takes the following
expression
2
1
E f
C
0
3 22
E
constant.
2
1
11
f
C
0
(5.127)
The last equation expresses the logical meaning of the parameter E and
currently fulfills the earlier assumed condition according to Eq. (5.118).
The parameter ) also has a logical meaning. For uniaxial deformation,
the following relation is valid from Eq. (5.127)
11 22
1 E ) .
11
179
(5.128)
2 P max 11 ) .
(5.129)
The logical meaning of the parameter ) is now evident from the last
equation.
Homogenous stress. In the case of homogenous stress the following
relations are valid
11 22 , so that 1 2 0 .
(5.130)
1
') f
C '
') f
C
0
2
E f
(
0
0 (
E f
( .
0
0
(5.131)
11
1
2
f
p
, f
C 02 p
.
C
0
(5.132)
180
(b) Equation (5.54) with Eq. (5.55) in the case of modified van Wyks
model mentioned in Section 5.3 or
(c) Equation (5.100) with Eq. (5.101) in the case of alternative model
mentioned in Section 5.4. In the special case of planar isotropic
orientation of fibers the relevant expressions are mentioned in
Eqs. (5.102) and (5.103).
Application of modified van Wyks model. Let us illustrate the described
procedure in case (b), i.e. application of modified van Wyks model. Using
Eq. (5.55) in Eq. (5.132) we obtain
f
C
0
k p 2
2
0 ,
2
(5.133)
f
1 C
0
k p 2
1 2
0 ,
12
f
2 C 02 k p 2
2 2
0 .
2
(5.134)
1G
0
12
1 6
)
2
2
E C 02 k p 2
2
0
C
k
H
8
p
1
0
2
CI
1
0
0 9
) k p 2
1 2
0 E
1
k p 2
2
0 ,
6
11
k p ') 2
1 E 2
1 ) E 1 2
0 ( , 7
7
8
7
2
2
22 k p ') 2
2 E2
) E 2
0 ( . 7
9
(5.135)
181
6
7
7
8
1
0
0
2
2
0 .7
22 E C 2 k p 2
2
0
k pE
79
C
11
1 0
2
C 2 k p 2
2
0
k p 2
2
0 ,
0
C
(5.136)
Cauchys stresses: According to Eqs. (5.104) and (5.107) it is valid
11 11
2 0 and 22 22 1 0 . Applying Eqs. (5.135) and Eq. (5.108)
in these two relations we obtain the following expressions
11 11
2
k p ') 2
1 E 2
1 ) E 1 2
0 ( 2
0
0
k p ') 2
1 2 E 2
1 2 ) 2 E 1 2 2
0 ( ,
0
0 0
0
6
11 k p ') 2
1 E 2
) E 2
0 ( , 7
1
1
7
8
7
22 k p ') 2
2 E 2
) E 2
0 ( .7
2
2
9
(5.137)
6
11 k p ') 2
E 2
) E 2
0 ( k p 2
2
0 ,
7
7
8
7
22 k p ') 2
0 E 2
) E 2
0 ( k pE 2
2
0 .7
0
0
9
(5.138)
Homogenous stress: The general expression for homogenous stress
describes the expression mentioned in Eq. (5.131) by applying the
conditions in accordance with Eq. (5.130). Now, Eq. (5.133) is valid, so
that Eq. (5.131) can be rearranged as follows
182
'
)C
C '
'
k p ' ) 2
k p ') 2
'
k p 2
0 2
0
E C 02 k p 2
2
0 (
0
(
(
0 2
0
E 2
2
0 (
0
E 2
)
E 2
0 ( .
0
0
(
(5.139)
1 N1
1 N 2 0
, so that ln
1 N1 ln
0 ln
1 N1 ln
0 1 .
Let us denote
g1
ln
1 N1 ln
0 1
ln
1 N 2 ln
0 2
, g2
, g1 g 2 1 .
ln
0 ln
0
ln
0 ln
0
(5.140)
g2
(5.141)
15.
183
0.04
0.03
0.02
22
kp
11
kp
0.01
0.2
0.8
0
1
0.4
0.6
0.6
0.4
0.8
0.2
1
0
g1
g2
1 N1
1 N 2
(a)
0.5
0.4
0.3
0.2
22
kp
11
kp
0.1
0
0
1
1
0.2
0.8
0.4
0.6
0.6
0.4
0.8
0.2
1
0
g1
g2
1 N1
1 N 2
(b)
5.21 Behavior of engineering stresses (pressures), according to Eqs. (5.135) and (5.55).
Parameters used: m = 1, a = 1, E = 0.4, 0 = 0.02;
(a) 0.3 ( 0 0.0667) , (b) 0.6 ( 0 0.0333) .
184
0.04
0.03
22
kp
11
kp
0.02
0.01
0.2
0.8
0
1
0.4
0.6
0.6
0.4
0.8
0.2
1
0
g1
g2
1 N1
1 N 2
(a)
0.5
0.4
0.3
22
kp
11
kp
0.2
0.1
0
0
1
1
0.2
0.8
0.4
0.6
0.6
0.4
0.8
0.2
1 g1
0 g2
1N1
1N2
(b)
5.22 Behavior of Cauchys stresses (pressures) according to Eqs. (5.137) and (5.55).
Parameters used: m = 1, a = 1, E = 0.4, 0 = 0.02;
(a) 0.3 ( 0 0.0667) , (b) 0.6 ( 0 0.0333) .
185
1
E2
k p
0.95
E 0.6
E 0.5
E 0.4
E 0.3
0.90
0.85
0
0
0.2
0.4
0.6
0.8 1
5.23 Size of addend E2( ) in relation to the whole value of homogenous stress
(pressure), according the Eq. (5.139) with (5.55).
Parameters: m = 1, a = 1, 0 = 0.02.
k pE 2
.
(5.142)
Note: Comparing Eq. (5.49) with Eq. (5.142) it is shown that the
expression for uniaxial compression can be roughly applied also for the
homogenous compression, but with another (smaller) parameter, i.e. kpE
in place of kp only.
186
5.7
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
187
6
Mechanics of parallel fiber bundles
6.1
187
188
Fb
F
A
A
h
1 N
A
h
1 N
B
B
B
F
(a)
Fb
(b)
6.1 Parallel fiber bundle between jaws. (a) One fiber. (b) Set of fibers.
(6.1)
189
ab a,
Fb n F n F
a .
(6.2)
The bundle force per one fiber Ff and the bundle breaking force per
one fiber Ff are
Ff Fb
N n F
N ,
Ff Fb n F F
a .
(6.3)
b
N
b
Fb
N n F
N F
N
N ,
T
nt
t
b
(6.4)
F
nF
F
.
T
nt
t
(6.5)
Using the foregone equations the following expressions are valid for
the bundle stress b;
N and the bundle strength ;b
Fb
N n F
N F
N
;
N ,
S
ns
s
F n F F
;b b
; .
S
ns
s
;b
N
6.2
(6.6)
Let us now change our earlier assumption number 3 and take the following
alternative assumptions:
3*. In an ideal fiber bundle,
(I) there are two components (for example, cotton and polyester are the
two components in a cotton-polyester fiber bundle),
(II) each fiber of the same component exhibits
(a) same fiber density ,
(b) same fineness t,
190
First component
Second component
Fiber density
Fiber fineness
Forcestrain function of fiber
Specific stress of fiber
Stress of fiber
Breaking strain of fiber
Breaking force of fiber
Specific strength of fiber
Strength of fiber
Number of fibers in bundle
Mass fraction in bundle
Mass(+) of component in bundle
1
2
t1
F1(N)
t2
F2(N)
1(N)
1;
N
a1 ..................
2(N)
;2
N
F1
a2
F2
1
;1
2
;2
n1
g1
m1
n2
g2
m2
(+) This corresponds to the (gauge) length h between jaws A and B shown in Fig. 6.1b.
s1 t1 1 , s2 t2 2 .
(6.7)
(6.8)
(6.9)
The following expressions are valid to write for fiber breaking forces,
fiber specific strengths (tenacities) and fiber strengths.
F1 F1
a1 , F2 F2
a2 .
191
(6.10)2
1 1
a1 F1 t1 , 2 2
a2 F2 t2 .
(6.11)
1; 1;
a1 F1 s1 ,
(6.12)
;2 2;
a2 F2 s2 .
(6.13)
(6.14)
(6.15)
m1 g1m, m2 g 2 m .
(6.16)
The total fineness (count) T of the bundle with mass m and (gauge)
length h is
T m h.
(6.17)
In the bundle, the total length of fibers of the first component is n1h and
the total length of fibers of the second component is n 2h. Using the
definition of fineness and Eqs. (6.16) and (6.17) it is valid to write that
6
7
7
8
m2
m2 g 2 m
T 7
, n2
g2 .
t2
n2 h
t2 h t2 h
t2 79
t1
m1
m gm
T
, n1 1 1 g1 ,
n1h
t1h t1h
t1
(6.18)
Forces in fibers. The forcestrain functions of one fiber form the first
and the second components are schematically illustrated in Fig. 6.2. The
end point of function F1(N) characterizes the breaking strain a1 and the
breaking force F1 of the fiber from the first component. Similarly, the end
point of function F2(N) characterizes the breaking strain a2 (a2 a1) and
the breaking force F2 of the fiber from the second component. Further,,
the force F2(a1) corresponds to the fiber of the second component at the
strain value N = a1.
2. We think that a fiber is not broken at the value of breaking strain (a1 and/or a2 ), but
it shows the maximum force, i.e. the breaking force at this value of strain. However,
a fiber is shown to be broken at each and every higher than this value of strain.
192
F2
Part A
Part B
Part
C
F1
F2
a1
F1
N
F2
N
a1
a2
Force and specific stress in bundle. Let us divide the strain area into
three parts as shown in Fig. 6.2. The strain values N 0, a1 correspond to
part A. No one fiber is broken in this interval given for the strain. The
strain N
a1 , a2 falls in part B. All the fibers of the first component are
broken in this interval, but each fiber of the second component is able to
carry the force. In part C, the strain N > a2 and evidently, all fibers are
broken.
In part A, each of n1 fibers of the first component carries the tensile
force F1(N) and each of n2 fibers of the second component carries the tensile
force F2(N). Using Eqs. (6.8) and (6.18), the bundle force Fb(N) and the
bundle specific stress b(N) can be expressed as follows
6
F
N
F
N
Fb
N n1 F1
N n2 F2
N T ' g1 1
g2 2 ( , 7
t1
t2
7
8 N 0, a1 (6.19)
Fb
N
F1
N
F2
N
b
N
g1
g2
g1 1
N g 2 2
N .77
T
t1
t2
9
In part B each of n1 fibers of the first component is broken and each of
n2 fibers of the second component carries the force F2(N). Now, using
Eqs. (6.8) and (6.18), the bundle force Fb(N) and the bundle specific stress
b(N) can be expressed as follows
F2
N
,
t2
6
7
7
8 N
a1 , a2
Fb
N
F2
N
b
N
g2
g 2 2
N . 7
79
T
t2
Fb
N n1 0 n2 F2
N Tg 2
(6.20)
193
In part C all the fibers form both components are broken and therefore
the bundle force as well as the bundle specific stress is equal to zero. This
is shown below
Fb
N b
N 0, N " a2 .
(6.21)
F
a
F
F
a
F
a
Fb
a1 T ' g1 1 1 g 2 2 1 ( T ' g1 1 g 2 2 1 (
t1
t2
t2 ,
t1
b
a1 g1 1
a1 g 2 2
a1 g1 1 g 2 2
a1 .
6
7
8 (6.22)
7
9
This expression is obtained from Eq. (6.19) by using Eqs. (6.10) and
(6.11).
Note: The specific stress 2(a1) is given by the ratio F2
a1 t2 ; the force
F2(a1) is shown in Fig. 6.2.
In part B, the maximum force in the bundle is originated at N = a2. This
force and the corresponding maximum specific stress can be expressed
from Eq. (6.20) using Eqs. (6.10) and (6.11) as follows
F2
a2
F 6
Tg 2 2 , 7
t2
t2 8
b
a2 g 2 2
a2 g 2 2 . 79
Fb
a2 Tg 2
(6.23)
The bundle breaking force Fb and the corresponding bundle specific
strength b are equal to the total maximum of the bundle force and the
bundle specific stress. Using Eqs. (6.22) and (6.23) the following
expression must be valid to write
Fb max + Fb
a1 , Fb
a2 ,
F
a
F 76
7G F
max HT ' g1 1 g 2 2 1 ( , Tg 2 2 8 .
t2
t2 97
I7 t1
(6.24)
194
b
N
Part A
b
a1
b
b
a2
a)
g 2 2
N
g1 1
N
g 2 2
N
a1
0
b
N
Part A
b
a2
g1 1
N
g 2 2
N
b
N
Part
C
g 2 2
N
a1
a2
N
Part B
b
b
a1
b)
Part
C
Part B
Part A
b
a1 b
a2
a2
N
Part B
b
Part
C
c)
g1 1
N
g 2 2
N
g 22
N
a1
a2
6.3 Scheme of specific stress of bundle according to Eqs. from (6.19) to (6.21). Total
maximum: (a) at a 1, (b) at a 2, (c) both same.
b max + b
a1 , b
a2 , max g1 1 g 2 2
a1 , g 2 2 .
(6.25)
ab a1 if b b
a1 " b
a2 , 6
7
ab a2 if b b
a2 " b
a1 , 8
7
ab is undefined if b b
a1 b
a2 . 9
(6.26)
Note: If both of the local maxima possess the same value in accordance
with Fig. 6.3c then purely mathematically there should exist two values
a1, a 2 of the bundle braking strain a b. However, this result cannot be
195
obtained in reality. Practically, such a bundle shall have the breaking strain
lying somewhere in the interval from a1 to a2 in relation to very small
difference from that of the real bundle.
Influence of blending. The function of the bundle specific strength and
the corresponding values of the bundle braking strain are shown in Fig. 6.4
in relation to the mass fraction of fibers in the bundle, according to Eqs.
(6.25), (6.26) and (6.15). The mass fraction of the second fiber component
g2 increases from left to right and the mass fraction of the first component
g1 increases from right to left on the abscissa (X-axis) of the each graph.
Two dashed lines show the two parts of the summation g1 1 g 2 2
a1
separately and one full line represents the summation of the two parts.
The second full line represents g 2 2 . The maximum of the mentioned two
full lines expresses the bundle specific strength b as the function of mass
fraction g2 (or g1) according to Eq. (6.25). It is graphically illustrated by
the triple-lines as shown in the four graphs in Fig. 6.4.
The point C refers to the point of intersection of the two full lines
mentioned earlier. In this point g 1 = g 1,C, g 2 = g 2,C, and evidently the
following expressions are valid
(6.27)
1 g
2,C
g 2,C
1
g 2,C 2
a1 g 2,C 2 , g 2,C 1 g 2,C 2
a1 g 2,C 2 1 ,
2 2
a1
1
,
g
1
g
.
1,C
2,C
1 2 2
a1
1 2 2
a1
(6.28)
Note: The point g1,C or g2,C is marked by the gray triangle on the abscissa
(X-axis).
The value of bundle specific strength b,C shown by point C results
from Eqs. (6.25), (6.27) and (6.28) as follows
b,C g1,C 1 g 2,C 2
a1 g 2,C 2
1 2
2 2
a1 .
1
(6.29)
196
ab undefined
ab a1
ab a2
b
2
1
C
g2 2
g22
a1
g1,C, g2,C
g2
g1
a)
g11
g22
a1
g2 2
2
a1
2
a1
g22
a1
g1,C, g2,C
g2
1
0
1
0
g1
b)
2
g11
ab undefined
ab a1
ab a2
b
1
g11
g22
a1
g11
ab undefined
ab a2
ab a1
b
ab undefined
ab a1
ab a2
b
2
2
g11
C
g22
a1
1
2 2
g
0
1
2
a1
g11
g22
a1 C
1
g22
a1
2 2
g
1 1
g
g22
a1
g1,C, g2,C
g2
g1
c)
2
a1
g11
g1,C, g2,C
1
0
g2
g1
d)
1
0
6.4 Behavior of specific strength and breaking strain of bundle in relation to mass
fraction of fibers in bundle; (a) 1 < 2 and 1 " 2
a1 , (b) 1 " 2 and 1 " 2
a1 ,
(c) 1 < 2
a1 , (d) 1 2
a1 .
197
Case 3: If 1 2
a1 then the graph shown in Fig. 6.4d is valid and
according to Eq. (6.29) b,C g1,C 1 g 2,C 1 g1,C g 2,C 1 1 . Then
6
7
8
7
9
(6.30)
For the previous equation we calculate the required values g1,C, g2,C and
b,C from Eqs. (6.28) and (6.29).
Example 6.1: Let us think about a cottonpolyester blended fiber bundle.
The cotton fibers possess specific strength (tenacity) 1 0.36 N tex -1 and
breaking strain (extension) a1 = 0.11 (11%). The polyester fibers possess
specific strength 2 0.41 N tex -1 , breaking strain a2 = 0.46 (46%), and
specific stress 2(a1) = 0.12Ntex 1 2 ( a1 ) 0.12 N tex -1 corresponding to
strain N = a1 = 0.11. Because a1 < a2, the choice of subscripts is valid and
the scheme shown in Fig. 6.4a is also valid in this case. The mass fraction
of the second material g 2,C = 0.554 according to Eq. (6.28) and the
corresponding mass fraction of the first component (cotton) is
g 1,C = 1 g 2,C = 0.446. The value of specific stress of the bundle is
b,C 0.227 N tex -1 and this corresponds to the point C according to Eq.
(6.29). Because 1 " 2
a1 , the minimum bundle specific strength (bundle
tenacity) is b min b,C 0.227 N tex -1 according to Eq. (6.30). It is shown
that the specific strength of the blended fiber bundle (44.6% cotton and
55.4% polyester) is even smaller than that of the less tenacious singlecomponent, i.e., cotton fiber bundle.
Let us be conscious of a paradox which is observed while blending
of two components. If we blend fibers having higher value of tenacity
with fibers having smaller value of tenacity then by a choice of bad
blending ratio we can obtain a bundle having smaller tenacity than that
198
S n1s1 n2 s2
g1 T t1
t1 1
g 2 T t2
t2 2
T g1 1 g 2 2 .
(6.31)
F
N
F
N
F
N
F
N
T ' g1 1
g 2 2 ( g1 1
g2 2
t
t
Fb
N
s
s22
1
2
1 1
;b
N
S
T g1 1 g 2 2
g1 1 g 2 2
g1
g
1;
N 2 ;2
N
2
, N 0, a1 .
1
g1 g 2
1 2
(6.32)
g1 1 v1 ,
g 2 2 v2 ,
v1 v2 1 .
(6.33)
v1
v
1;
N 2 2;
N
;b
N
v11;
N v2 ;2
N , N 0, a1 .
v1 v2
(6.34)
199
F2
N
F
N
g2
v2
;
N
g2 2
;2
N
Fb
N
2 2
t2
s22
,
b;
N
g1 g 2
v1 v2
S
T g1 1 g 2 2 g1 1 g 2 2
1 2
Tg 2
;b
N v2 ;2
N , N
a1 , a2 .
(6.35)
(6.36)
Let us now compare Eqs. (6.19), (6.20), (6.21), which are valid for the
specific stress b(N), with Eqs. (6.34), (6.35), (6.36), which are valid for
the stress ;b
N . The mathematical structure of the corresponding
expressions is the same, but in the place of specific stresses b(N), 1(N),
2(N) the stresses ;b
N , 1;
N , ;2
N are used and in the place of mass
fractions g1, g2 the volume fractions v1, v2 are used.
It is possible to derive all other expressions in the same way as before.
Then, in place of Eq. (6.22) we obtain
;b
a1 v1 1;
a1 v2 ;2
a1 v1 1; v2 ;2
a1 .
(6.37)
(6.38)
;b max + ;b
a1 , b;
a2 , max v1 1; v2 2;
a1 , v2 ;2
(6.39)
ab a1 if b; ;b
a1 " ;b
a2
ab a2 if ;b ;b
a2 " ;b
a1
ab is undefined if b; ;b
a1 ;b
a2 .
6
7
8
7
9
(6.40)
(6.41)
200
v2,C
;2 ;2
a1
;1
,
v
1
v
1,C
2,C
1; ;2 ;2
a1
1; ;2 ;2
a1 .
(6.42)
1; ;2
1; ;2 2;
a1
(6.43)
6
7
8
7
9
(6.44)
After changing of the symbols, the graphical interpretation looks
analogically to Figs. 6.3 and 6.4.
Table 6.2 Replacement of fiber characteristics by yarn characteristics.
Symbols
Original sense
1 , 2
;1 , ;2
Strengths of fibers
a1 a2
;2
a1
;2
a1
Stress in the fiber of the second Stress in the single-component yarn from
component at the breaking strain second material at the breaking strain of
of the first component
the single-component yarn from the first
material
g 1, g 2
v1, v2
201
twist, and the fibers do not possess same mechanical properties. Therefore,
the immediate application of the previous results cannot be recommended.
The idea, based on the replacement of earlier fiber characteristics to
analogical yarn characteristics, offers better results. This substitution is
shown in Table 6.2.
6.3
Let us think of an ideal fiber bundle in accordance with Section 6.2. Let
us then modify the assumption 3 * mentioned in Section 6.2 as follows.
3** In an ideal fiber bundle,
(I) there are K components, K 2.
(II) Each fiber of the same component exhibits
(a) same fiber density ,
(b) same fineness t,
(c) same continuously increasing forcestrain relation F = F(N),
(d) same breaking force F* and
(e) same breaking strain a.
(III) The properties of fibers between components are generally different.
Table 6.3 introduces the symbols used for such a fiber bundle.
We use the convention, by which the quantities related to a component
of smaller breaking strain are displayed with the smaller subscript i. Strictly
speaking,
ai 1 ai ,
i 1, 2,
, K ,
(6.45)
(6.46)
progression.
Initial quantities. The following equations are valid for each general
i-th component i = 1, 2,, K, in an analogy4 to earlier Eqs. from (6.7) to
(6.16).
The cross-sections of the fibers are expressed as follows
si ti i .
(6.47)
202
i-th component
i = 1, 2, , K
Fiber density
i
Fiber fineness
ti
Fi(N)
i(N)
Stress of fiber
;i
N
ai
Fi* Fi ai
i a
Strength of fiber(+)
Number of fibers in bundle
i; i;
ai
ni
gi
(+)
(++)
Mass
of component in bundle
i i
mi
The fiber specific stresses and the stresses are stated below
i
N Fi
N ti , i;
N Fi
N si .
(6.48)
The fiber breaking forces, the specific strengths and the strengths are
expressed as follows
Fi Fi
ai ,
(6.49)
i i
ai Fi ti ,
(6.50)
;i i;
ai Fi si .
(6.51)
n ni .
(6.52)
i 1
m mi
i 1
(6.53)
203
mi
m
gm
T
, ni i i g i .
ni h
ti h ti h
ti
Fj
Fi
N
Fi
(6.54)
Fj
N
Fj
ai
a1
ai
aj
aK
aK 1 , aK ,
aK , A . If N
a0 , a1
0 , a1 then no one fiber will break
because no one component will have its braking strain smaller than N. If
N
a1 , a2 then the fibers from the first component will be broken (a1 < N),
but all the other components will not break. That is, their breaking strains
are not smaller than N. If N
a2 , a3 then the fibers from the first and second
components will be broken (a1 < N and a2 < N) and the fibers from other
204
Fj
N
j i
j i
tj
If N
ai 1 , ai , i 1, 2,
K , then Fb
N n j Fj
N T g j
If N " aK then Fb
N 0.
6
. 7
8
7
9
(6.55)
Applying the first expression of Eq. (6.48) in the previous expressions,
the bundle specific stress b(N) is given by the following expressions
6
Fb
N K
g j j
N , if N
ai 1 , ai , i 1, 2,
K , 7
T
7
j i
8
Fb
N
7 (6.56)
b
N
0, if N " aK .
79
T
b
N
Note: If K = 2 then Eqs. (6.55) and (6.56) are identical to Eqs. (6.19),
(6.20) and (6.21).
Breaking force and specific strength in bundle. The local maxima of
breaking force and/or specific stress of the bundle lie in the upper limits
of the strain intervals
a0 , a1 ,
a1 , a2 ,
a2 , a3 , ,
aK 1 , aK . The values
of these local maxima of breaking force and/or specific stress of the bundle
accord to Eqs. (6.55) and (6.56)
K
Fb
ai T g j
j i
Fj
ai
tj
b
ai g j j
ai
i 1, 2,
K , (6.57)
j i
Fj
ai 76
7G K
F max + Fb
ai , i 1 max HT g j
8
t j 97 ,
I7 j i
i 1
K
b
(6.58)
205
7G K
76
max + b
ai , i 1 max H g j j
ai 8 .
7I j i
97i 1
K
b
(6.59)
Note: If K = 2 then Eqs. (6.58) and (6.59) change to Eqs. (6.24) and
(6.25). (The symbols according to Eqs. (6.49) and (6.50) are still valid).
The bundle breaking strain ab is given by the following expression
ab ai k if b max + b
ai , i 1 b
ak is valid
6
7
for only one subscript k +1, 2, K , ,
7
8
K
ab is undefined if b max + b
ai , i 1 b
ak is valid 7
for minimum two subscripts k +1, 2, K , . 79
K
(6.60)
Note: 1. Usually, there is only one value b(ak) from the set + b
ai , i 1
which is the highest. Then the bundle breaking strain exists and it
corresponds to the value ak. However, if there are two or more values from
K
the set + b
ai , i 1 that have same value and they are all the maximum
value, then the bundle has two or more breaking strains. Of course, it is
physically not real (see the similar note preceding Eq. (6.26)). If K = 2
then Eq. (6.60) is changed to Eq. (6.26).
Mathematical rearrangement of b(a i). At some set of mass fractions
+ gi ,iK1 the bundle specific strength (tenacity) is the smallest in the case of
bi-component fiber bundle (see Fig. 6.4 and subsequent discussions).
Nevertheless, finding such a worst blending ratio is more difficult in
the general case of K components. To solve such a problem, the following
mathematical rearrangement is useful.
Let us define the quantities
K
Gi g j , i 1, 2, , K .
(6.61)
j i
j 1
jK
G1 g j 1, GK g j g K , Gi 1
j i 1
g j g j gi Gi gi.
j i
(6.62)
206
gi
gi
, i 1, 2, , K .
Gi
(6.63)
g1 g1 G1 g1 1 g1 , g K g K GK g K g K 1 . (6.64)
We also define the following quantities
ri 1 g i (g i ri 1), i 1, 2, , K .
(6.65)
The value ri represents a relative mass fraction of the other remaining
fibers in the mentioned partial group of components. Using Eqs. (6.62) and
(6.63) in the last expression, the following equation is obtained
ri 1
gi Gi gi Gi 1
, i<K.
Gi
Gi
Gi
(6.66)
r1 1 g1 1 g1 ,
rK 1 g K 1 1 0 .
(6.67)
6
7
Gj
Gj
7
8
g j j
ai rj Z j 1
ai , j 1, 2, K 1, i j. 7
7
Z K
ai K
ai (i.e. by j K , i j ).
9
Z j
ai
gj
j
ai
G j 1
Z j 1
ai
(6.68)
The sense of the function Zj (ai) is shown by the result of the following
rearrangement, based on a repeated use of Eq. (6.68) and then a final use
of Eq. (6.62).
Z j
ai
gj
Gj
gj
Gj
j
ai
j
ai
G j 1
Gj
Z j 1
ai
G j 1 g j 1
G j2
j 1
ai
Z j 2
ai (
'
G j G j 1
G j 1
gj
Gj
j
ai
g j 1
Gj
j 1
ai
G j2
Gj
207
Z j 2
ai
G j2 g j2
G j3
Z j 3
ai (
j 2
ai
'
Gj
Gj
G j G j2
G j2
gj
g j 1
g j2
G j 3
Z j 3
ai
j 2
ai
j
ai
j 1
ai
Gj
Gj
Gj
Gj
gj
j
ai
g j 1
j 1
ai
gj
Gj
j
ai
g j 1
Gj
j 1
ai
g j 2
Gj
j 2
ai
g K 1
G
K 1
ai K Z K
ai
Gj
Gj
1
g j j
ai g j 1 j 1
ai g j 2 j 2
ai g K 1 K 1
ai g K K
ai .
Gj
(6.69)
It is valid j (ai1) .j (ai) for each i j. (That is to say, each of j (N) is
an increasing function according to the assumption 3 **II)c) mentioned in
the introductory part of this chapter). So, the following inequality arising
from Eq. (6.69) is valid
Z j
ai 1 Z j
ai
(6.70)
Now, we can express the values b(ai) of the local maxima of bundle
specific stress according to Eqs. (6.57) and (6.69) by the following form
b
ai Gi Z i
ai
i 1, 2,
K .
(6.71)
b
ai 1 Gi 1 gi1 i 1
ai 1 ri1Z i
ai 1 , 67
8
b
ai Gi 1ri1Z i
ai .
79
(6.72)
208
Local
BASIC FORM
FORM A
maximum according to Eq. Basic form after
(6.71)
using Eq. (6.66)
b
a1
G1Z1
a1
b
a2
G2 Z 2
a2
FORM B
Basic form after using Eq. (6.68)
G1 g11
a1 r1 Z 2
a1
G1r1 Z 2
a2
G2 g 22
a2 r2 Z 3
a2
b
ai 1
Gi 1Z i 1
ai 1
Gi 2 ri 2 Z i 1
ai 1
Gi 1 gi1 i 1
ai 1 ri1Z i
ai 1
b
ai
Gi Zi
ai
Gi 1ri1Zi
ai
Gi g ii
ai ri Zi 1
ai
b
aK 3 GK 3 Z K 3
aK 3 G
K 4 K 4
Z K 3
aK 3 GK 3 g
K 3
K 3
a K 3 rK 3 Z K 2
aK 3
b
aK 2 GK 2 Z K 2
aK 2 GK 3rK 3 Z K 2
aK 2 GK 2 g K 2 K 2
aK 2 rK 2 Z K 1
aK 2
b
aK 1 GK 1Z K 1
aK 1 GK 2 rK 2 Z K 1
aK 1
b
aK
GK Z K
a K
GK 1 g K 1 K 1
aK 1 rK 1Z K
aK 1
GK 1rK 1Z K
aK
Let us assume that we know three values in the last couple of equations:
i 1
ai 1 , Z i
ai 1 and Z i (a i ). However, the values of next triplet of
parameters Gi 1 , g i1 , ri1 are not known at this moment.
We introduce the symbol Xi1 for the higher value from b(ai1) and b(ai).
Using Eq. (6.72) we can write
X i 1 max + b
ai 1 , b
ai , ,
(6.73)
X i 1 Gi 1Yi 1 , Yi 1 max g i1 i 1
ai 1 ri1Z i
ai 1 , ri1Z i
ai . (6.74)
209
a1
( a1 a2 )
a2
(+)
2
a1
2
( 2
a1 )
( g1 g 2 1 )
ai 1
( ai 1 ai )
aY , i 1
i 1
ai 1 i 1
(++)
Zi
ai 1
2
g1
g2
ai
ab
1
a1 1
Zi
ai ( Zi
ai 1 Zi
ai ) (+++)
gi1
The poi nt C of i ntersecti on has the val ues ri1 ri1,C (and /o r
gi1 gi1,C 1 ri1,C )
and
Y i 1 = Y i1 ,C ;
th e
equality
gi1,C i 1
ai 1 ri1,C Zi
ai 1 ri1, C Zi
ai is valid at this point (similar to
Eq. (6.27) in the previous chapter). Rearranging this equation together
with gi1 1 ri1,C we obtain the following expressions as follows:
ri1,C
i 1
ai 1
,
i 1
ai 1 Z i
ai Z i
ai 1
(6.75)
Zi
ai Zi
ai 1
i 1
ai 1 Z i
ai Z i
ai 1
i 1
ai 1 Z i
ai
i 1
ai 1 Z i
ai Z i
ai 1
(6.76)
210
Yi 1
Z i
ai
i1
ai 1
ri1Z i
ai
i1
ai 1
g i 1
ai 1
i 1
Yi 1
ri1
i 1
g
a)
aY , i 1undefined
aY , i 1 ai 1
aY , i 1 ai
ri1Z i
ai 1
Yi 1
g
b)
aY , i 1undefined
aY , i 1 ai 1
aY , i 1 ai
Z i
ai1
i 1
ai 1
gi1i1
ai1
ri1Z i
ai1 C
ri1
i 1
g
c)
1
0
Z i
ai
ri1 Z i
ai
Z i
ai1
ri1 Z i
ai 1
gi1i 1
ai 1
g i1i 1
ai 1
Z i
ai1
gi1,C ,
ri1,C
i 1
Z i
ai
ri1Z i
ai 1
i 1
ai 1
Z i
ai
ri1
1
0
gi1i 1
ai 1
ri1Z i
ai 1 C
ri1 Z i
ai
ri1Z i
ai
gi1i 1
ai 1
Z i
ai1
ri1 Z i
ai 1
gi1i1
ai1
ri1Z i
ai1
gi1i 1
ai 1
ri1Z i
ai 1
aY , i 1undefined
aY , i 1 ai
aY , i 1 ai 1
Yi 1
ri1
gi1
d)
1
0
*
6.6 Behavior of Y i1 and corresponding a Y, i1 in relation to ri 1 and/or gi1 ;
(a) i 1
ai 1 < Z i
ai and i 1
ai 1 " Z i
ai 1 ,
(b) i 1
ai 1 " Z i
ai and i 1
ai 1 " Z i
ai 1 ,
(c) i 1
ai 1 < Z i
ai 1 ,
(d) i 1
ai 1 Z i
ai 1 .
Point C of intersection: ri 1 ri 1,C and/or g i 1 g i 1,C .
2. If ri1 " ri1,C (right part of each graph) then Yi 1 ri1Z i
ai see the
second thick line in the graphs and the corresponding value of
strain is aY, i1 = ai.
3. If ri 1 ri 1,C (the intersection of both thick lines) then Eq. (6.76) is
valid, i.e., Yi 1 Yi 1,C i 1
ai 1 Z i
ai i 1
ai 1 Z i
ai Z i
ai 1 .
The corresponding value of strain aY,i1 is undefined. (Formally, aY,i1 is
equal to ai1 and ai at the same time, but two values are not physically
real).
Minimum value of Y i1 and Xi1. Let us continue our consideration to the
211
g i1 gi1,min 1 ri1,min ); see the gray triangles in the graphs in Fig. 6.6.
We denote a value aY,i1 corresponding to Yi1 = Yi1,min as aY,i1,min.
In consideration of the graphs shown in Fig. 6.6 and the considerations
similar to Section 6.2, we obtain the value Yi1,min, and the values gi1,min ,
(6.77)
Let us remember that it is necessary to use Eqs. (6.75) and (6.76) for
practical evaluation of the quantities g i1,C , ri1,C and Y i1,C . The last
expression defines the practical way for calculation of Y i1,min and the
corresponding values gi1,min , ri1,min .
The minimum of Xi1, called Xi1,min, can be formulated according to
Eq. (6.74) as follows
X i 1, min Gi 1,minYi 1, min
(6.78)
Note: We will discuss about the suitable value of Gi1 = Gi1,min later on.
Minimum specific strength evaluation. Equation (6.77) determines the
quantity ri1,min (and also the quantity gi1,min 1 ri1, min ) by only one value
in the cases according to the graphs (a), (b) and (c) shown in Fig. 6.6.
However, this quantity can be an arbitrary value taken from the whole interval
0, ri1,C in graph (d). It infers that there are a lot of blending ratios that can
212
the one hand, this case of absolutely precise equality i1(ai1) = Zi(ai1) is
practically very unique, and on the other hand, the mathematical solving of
this special case brings lot of difficulty5 . Therefore, let us assume that the
mentioned equality is valid for no one subscript i. (Only the graphs (a), (b)
and (c) shown in Fig. 6.6 can be actual).
The bundle specific strength b , determined by Eq. (6.59), obtains its
Note: Let us remember that j a j j is the specific strength of the
fiber from j-th component, and the values j(ai) for i < j express the specific
stresses in the mentioned fiber at N = a1,a2,aj1, , i.e., by breaking strains
of the fibers from all previous components.
The scheme of the respective calculation is shown in Table 6.5. All the
values ZK(ai) in column 1 are known from the last expression mentioned
in Eq. (6.68) they are directly the input values K(ai), because j = K in
this equation. Using the first (bold printed) couple of values, i.e.
Z K
aK K
aK , and Z K
aK 1 K
aK 1 , and Eq. (6.75) at i = K, we
can evaluate the quantities rK 1,min , g K 1,min from Eq. (6.77)7 at i = K.
Further, using the values rK 1 rK 1,min , g K 1 g K 1,min , known now, and
the values ZK(ai), all from the first column, we can calculate all values
ZK1(ai) in the second column according to Eq. (6.68) by j = K 1, as shown.
Using the first (bold printed) couple of values, i.e., ZK1(aK1) and ZK1(aK2),
and i = K 1 in Eqs. (6.75) and (6.77), we can analogically as formerly
evaluate the quantities rK 2,min , g K 2,min . But the known values rK 2, min ,
g K 2, min and the values ZK1(ai) from the second column allow us to calculate
all values ZK2(ai) in the third column according to Eq. (6.68) by j = K 2,
as shown, etc., to the last column K 1. In this manner, we get beside
others all the values gi,min and ri,min for i = 1,2,,K1 and the last values,
5. Generally, such a bundle shows the same minimum specific strength for the whole
area of blending ratios.
6. See the previous assumption.
7. Only the first two alternatives can be actual, based on our previous assumption.
213
Calculated values of Z j
ai according to Eq. (6.68)
Input values,
Z K
ai K
ai
Using
, rK 1, min and
K 1, min
Using
g 2, min , r2, min
Using
, rK 2, min and
K 2, min
and
Z 3
ai from
Z K
ai from the
previous
column
Z K 1
ai from the
previous
column
Z K
aK
Z 2
a2
Z K
a K -1
Z K -1
a K -1
Z K
aK 3
Z K 1
aK 3
Z K
a2
Z K 1
a2
Z K
aK 2
Z K -1
a K -2
Z K
a1
Z K -2
a K -2
Z K -2
a K -3
Z K 2
a2
Z K 1
a1
Z K 2
a1
the previous
column
K-1
Z 2
a1
Using Z K 1
aK 1
Using Z K 2
aK 2
and Z K
aK 1
and Z K 1
aK 2
and Z K 2
aK 3
K 1, min
K 2, min
K 1, min
Using Z 2
a2
and Z 2
a1
K 3, min
r1,min
K 3,min
g1,min
K 2, min
+g ,
K
i ,min i 1
and/or ri,min
K
i 1
The second part of evaluation utilizes Eqs. (6.62) and (6.63) for
K
i 1
strength of the bundle. The mentioned equations have the following form
now.
8.
(6.79)
(6.80)
g K , min and rK , min take some values of g K and rK , so that these equations must
be valid too.
214
The equality
g1,min
g1, min
(6.81)
is also valid, according to Eq. (6.64). We also keep the set of values
+g ,
K
i ,min i 1
g 2, min g 2, min G2, min according to Eq. (6.80). Then, it is possible to
G3,min , e t c ., a n d
c a l c u l a t e G 3 , m in = G 2 , m i n g 2 , m in a n d g3, min g3,min
and
finally
+g ,
i , min i 1
b, min
b, min
G7 K
67
max H g j ,min j
ai 8 .
I7 j i
97i 1
(6.82)
b
a1 g11
a1 g 2 2
a1 g3 3
a1 , 6
7
b
a2
g 2 2
a2 g3 3
a2 , 8
7
g3 3
a3 . 9
b
a3
b max + b
a1 , b
a2 , b
a2 , .
215
(6.83)
(6.84)
Z3
a3 3
a3 , Z 3
a2 3
a2 , Z3
a1 3
a1 .
(6.85)
2
a2
1 r2,C
, g 2,C
.
2
a2 Z 3
a3 Z 3
a2
(6.86)
Now, we shall use Eq. (6.77) for i = 3 to determine the values g 2, min
and r2,min
as follows
If 2
a2 " Z 3
a2 then g 2,min
g 2,C
, r2,min
r2,C
, 67
8.
if 2
a2 < Z3
a2 then g 2, min 1, r2,min
0.
79
(6.87) 9
(6.88)
1
a1
, g1,C
1 r1,C
.
1
a1 Z 2
a2 Z 2
a1
(6.89)
9. Do not forget that the case of equality was neglected according to our assumption.
216
Now, we shall use Eq. (6.77) for i = 2 to determine the values g1, min and
r1,min
as follows
If 1
a1 " Z 2
a1 then g1, min g1, C , r1,min r1,C
, 67
8 . (6.90)
if 1
a1 < Z 2
a1 then g1, min 1, r1,min
0.
79
We can also write the following expressions from Eqs. (6.64) and (6.65)
for K = 3.
g3, min 1, r3,min
0.
(6.91) 10
In this way, we find the complete triplet of values g1, min , g 2, min , g3, min .
Equation (6.81) determines the quantity
g1,min g1,min
.
We get the values G1, min , G2,min from Eq. (6.79) for i = 1
G1, min 1, G2, min G1, min g1, min .
(6.92)
(6.93)
(6.94)
Finally, we use Eq. (6.80) once more for i = 3 and also g3, min 1
according to Eq. (6.91)
g3,min g3,min
G3, min G3,min .
(6.95)
So, we find the triplet of mass fractions g1,min, g2,min, g3,min pertaining to
the minimum bundle specific strength b,min . Applying Eqs. (6.83) and
(6.84) we determine this value as follows
b,min
a1 g1, min 1
a1 g 2,min 2
a1 g3, min 3
a1 , 6
77
g 2, min 2
a2 g3,min 3
a2 , 8
b,min
a2
7
g 3, min 3
a3 . 79
b,min
a3
(6.96)
10. g 3, min and r3, min take some values of g 3 and r3 , so that these equations must be
valid too.
217
(6.97)
Example 6.2: Hypothetical input data. Let us think about the triplet of
(hypothetical) fibers according to Fig. 6.7. The fiber number 1 possesses
the breaking strain a1, the fiber number 2 shows its breaking strain a2 > a1,
and the fiber number 3 displays the highest breaking strain a3 > a2. The
necessary values of tenacities and specific stresses are determined
immediately from Fig. 6.7.
We find the values as follows:
1
a1 10,
6
7
2
a1 8, 2
a2 15,
8
7
3
a1 7, 3
a2 12, 3
a3 20.9
20
i
N
15
10
5
0
3
a3
2
a2
1
a1
2
a1
(6.98)
3
a2
3
a1
a1
a2
a3
218
g2
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
10
0.9
9.7
9.8
0.8
9.4
9.5
9.6
0.7
9.1
9.2
9.3
9.4
0.6
8.8
8.9
9.0
9.1
9.2
g1 0.5
10.0
8.6
8.7
8.8
8.9
0.4
12.0
10.0
8.4
8.5
8.6
8.7
9.0
0.3
14.0
12.0
10.0
9.3
9.6
9.9
10.2
10.5
0.2
16.0
14.0
12.0
10.5
10.8
11.1
11.4
11.7
12.0
0.1
18.0
16.0
14.0
12.0
12.0
12.3
12.6
12.9
13.2
13.5
20.0
18.0
16.0
14.0
13.2
13.5
13.8
14.1
14.4
14.7
9.0
15.0
(2) The smallest calculated value in our table is b 8.4 , which was
obtained by g1 = 0.4, g2 = 0.2 and g3 = 0.4, so that we can wait, the
total minimum of specific strength of the bundle is near to this case.
Further, we calculated the minimum specific strength of the bundle using
the input data according to Eq. (6.98) and step by step Equations from
(6.85) to (6.97), and we obtained b, min 8.310 for mass fractions
g1,min = 0.3629, g2,min = 0.2216 and g3,min = 0.4155. (Our expectation from
the table was right).
Example 6.3: Cotton/viscose/polyester. Let us think about three type of
real fibers: (1) cotton fibers11 (breaking strain a1 = 0.08), (2) viscose fibers
(breaking strain a 2 = 0.17), and (3) polyester fibers (breaking strain
a3 = 0.27). Beside the breaking strain values we found the necessary values
for the tenacities and the specific stresses from (mean) stressstrain curves,
available for these fibers
Cotton: 1
a1 40 cN tex 1 ,
6
7
Viscose: 2
a1 16 cN tex , 2
a2 26 cN tex ,
8
1
1
1 7
Polyester: 3
a1 30 cN tex , 3
a2 50 cN tex , 3
a3 53cN tex .9
1
(6.99)
11. We assume that each cotton fiber is gripped by both jaws of the tensile tester in
this case.
219
g2
0
g1
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
40.0
0.9
39.0
37.6
0.8
38.0
36.6
35.2
0.7
37.0
35.6
34.2
32.8
0.6
36.0
34.6
33.2
31.8
30.4
0.5
35.0
33.6
32.2
30.8
29.4
28.0
0.4
34.0
32.6
31.2
29.8
28.4
27.0
25.6
0.3
37.1
32.6
30.2
28.8
27.4
26.0
24.6
23.2
0.2
42.4
37.6
35.2
32.8
30.4
28.0
25.6
23.2
20.8
0.1
47.7
42.6
40.2
37.8
35.4
33.0
30.6
28.2
25.8
23.4
53.0
47.7
45.2
42.8
40.4
38.0
35.6
33.2
30.8
28.4
26.0
The broken trends are shown only for the constant value of g3 on
diagonal directions in this case; e.g., for g3 = 0 (diagonal from 40 to 26)
the minimum value is 20.8 cN tex1 and this is also the smallest value in
the whole table. So, we can wait, the total minimum is near to g1 = 0.2,
g2 = 0.8 and g3 = 0.
The minimum specific strength of the bundle was calculated using the
input data according to Eq. (6.99) and step by step Equations from
(6.85) to (6.97). We obtained b, min 20.8000cN tex 1 for mass fractions
g1,min = 0.2000 (cotton), g2,min = 0.8000 (viscose) and g3,min = 0 (polyester)
the blend without polyester fibers. (Coincidentally, the result is quite
the same as our expectation from the table, thanks to values from Eq. (6.99).
Usually, it is not so).
6.4
A real fiber bundle consists of many single fibers that have different
physical and mechanical characteristics. When such a fiber bundle is tested
for its mechanical behavior, we observe its (specific) stressstrain curve,
220
Area I
d h
1
l0
AREA II
2
d l0
F 0
F 0
Area III
d l
d " l0
d < la
d la
F<P
FP
d " la
F 0
Let us now increase the distance d. The slack fiber is straightened and
obtained a quite straight position, where d = l0 (Figure 6.8, Case 2). In
Area I, the general fiber length is l = l0 and the acting force is F = 0.
Further increase of the distance d causes elongation of the fiber so that
l = d >l0 (Case 3). In Area II, the distance increases from d = l0 (Case 2)
to fiber breaking length d = l = la (Case 4). And the force increases from
F = 0 (Case 2) to fiber breaking force (strength) F = P (Case 4).
If the distance d is higher than the fiber breaking length la (Area III)
then the fiber breaks (Case 5), the fiber length l cannot be defined, and the
force becomes F = 0 again.
Table 6.8 summarizes the obtained relations. It can be noted that a
positive value of force F relates only to the fiber in Area II.
221
Fiber length l
Distance d
Acting force F
Area I
l l0
d h , l0
F 0
Area II
l d l0 , l a
d l0 , la
F 0, P
Area III
Broken fiber
d " la
F 0
(6.100)
N f
l l0 l0 l l0 1, N f 0 ,
(6.101)
a f
la l0 l0 la l0 1, a f 0 ,
(6.102)
%
l0 h h l0 h 1, % 0 .
(6.103)
d h
1 N ,
(6.104)
l0 h
1 % .
(6.105)
(6.106)
d
l
l l
1 a 1 a 0 1 1 a f
1 % 1 a f % a f %. (6.107)
h
h
l0 h
From the last equation we also obtain the following expressions for the
upper limit of the Area II
222
N %
1 a f a f , %
N a f
1 a f ,
(6.108)
N a f
1 % %, a f
N %
1 % .
(6.109)
Note: The upper limit of Area II is given by relation among the triplet
of variables N, af, %. Therefore, all three Eqs. (6.107), (6.108), and (6.109)
say the same information.
The limit lines are shown in Fig. 6.9.
N a f % a f %,
or %
N a f
1 a f
N a f % a f %,
or a f
N %
1 %
N Area
III AREA II
af
Area
III
N%
Area I
AREA II
N%
%
Area I
(a)
(b)
af
6.9 Graphical representation of the borders among Areas I, II, and III.
Let us now consider that we know the values of N and af and we want
to define the fiber in Area II by averages of the quantity %. This can be
done by following expressions, obtained from Fig. 6.9a by using Eq. (6.108)
67
8
1 a f , N , (See Z). 79
(6.110)
In another case, let us consider that we know the values of N and % and
we want to define the fiber in Area II by the quantity af. Then, it is valid
to write the following expressions, obtained from Fig. 6.9b, by using
Eq. (6.109).
If N % then a f
N %
1 % , A , (See Z). 67
8
If N < % then a f do not exist for the Area II.
79
(6.111)
Note: The fiber in Area I is given by the relation N < %.and the fiber in
Area III co rrespon ds to th e expressio n N > a f % + a f + %, and/or
% <
N a f
1 a f , and/or a f <
N %
1 % .
223
For the whole Area II, it is valid to write that l = d see Fig. 6.8 and/
or Table 6.8. Then the strain N, described generally according to Eq. (6.100),
can have the form N l h 1 for Area II, and we can rearrange it using
Eqs. (6.101) and (6.103) as follows
N
d
l
l l
1 1 0 1 1 N f
1 % 1 N f % N f % ,
h
h
l0 h
Nf
N%
.
1 %
(6.112)
N %
F S
N f S
, N f 0, a f , ( 0 S
0 , P S
a f ). (6.113)
1 %
Each fiber in a bundle has its own forcestrain curve S(Nf), strength P,
breaking strain a f , and crimp %. The quantities af , P, and % are random
variables having average values a f , P , and % .
We often assume that there exists an average forcestrain function
S N f that is common for all fibers (Figure 6.10). For this, it is valid to
so
that P S a f .
(c) The so-called theorem of similarity says that the forcestrain
function of each fiber (before breaking point) is proportional to the
expressions P K S a f , K P S a f
F
P
P
N %
S Nf
S
1 % .
S af
S af
(6.114)
224
S
N f , P
S
N f
af
N, a f
(Equation (6.112) was used for the second expression). The breaking
force P and the breaking strain af are only two scalar characteristics of
individual fiber; S
N f is a common function for all fibers. Assuming, for
S Nf
P
P N%
Nf
af
af 1 % ,
P N%
N % P
N
F S
N f S
1 % a f f a f 1 % .
(6.115)
(6.116)
N a f % a f % .
(6.117)
225
F
555
F w a f , P, % da f dP d%
a f
0, A
P
0, A
%
0, A
(6.118)
w a f , P, % v
% u a f , P ,
(6.119)
g a f 5 u a f , P dP ,
0
A
(6.120)
P a f 5 P u a f , P dP g a f ,
0
(6.121)
F
N
G A A
6
7 ' F w a , P, % dP ( da 7 d%
F
w
a
,
P
,
%
d
a
d
P
d
%
f
f
f
f
555
5
5
5
H
8
a f
0,A
0 N% 0
.
7I 1%
P
0,A
97
%
0, A
(6.122)
13. It tells about the distribution of a f only.
14. This is the average value of fiber strength P that can be obtained from a subset of
fibers having same value of breaking strain a f . See Fig. 6.13 preliminarily.
226
Let us rearrange Eq. (6.122) by using Eqs. (6.114), (6.119), and (6.121)
N
G A A
6
F 5 7H 5 ' 5 F w a f , P, % dP ( da f 78 d%
0 N% 0
7I 1%
79
N
G A A P
6
N %
S
v % u a f , P dP ( da f 78 d%
5 7H 5 ' 5
S
a f 1 %
(
0 N% '
7I 1% 0
97
N
G A A
6
N %
1
v
% d% ,
5 7H 5 ' 5 P u a f , P dP (
da f 78 S
1 %
S
a f
0 N% 0
7I 1%
79
N G A
6
N %
P
af
F 5 7H 5
g
a f da f 78 S
v
% d%.
1 %
S
a
N%
f
07
I 1%
97
(6.123)
(6.124)
0a a N a a f % a f % .
(6.125)
to F 50 F v
% d% , and using Eq. (6.110) we derive the following
expression
N
6
7
77
8
N %
S
v
% d%. 7
1 %
7
79
N %
If N a f then F 5 S
v
% d%.
1 %
0
N
If N " a f then F
N a f
1 a f
(6.126)
227
1
e
2 %
ln % 2
22
, ln
%
CV%2 1
, ln CV%2 1 . (6.127)
The quantity % indicates the average value of % and CV% denotes the
coefficient of variation15 of %.
We calculated F from Eq. (6.126) using Eqs. (6.116) and (6.127)
applying a numerical method of integration. Some results are shown in
Fig. 6.11. Each thick curve represents the relation between the jaw
displacement N and the ratio F P . (If average force per fiber in the bundle
reaches to the strength of single fiber then we obtain F P 1 ).
In Fig. 6.11, the breaking points of the bundles ({ ) are shown to lie on
the thin line. Their coordinates express the strength utilization coefficients
Fmax P 0 P Equation (6.124) and the breaking strain of bundle is
N = a. Also, using Eq. (6.125), we can evaluate the breaking strain
utilization coefficient as 0a a a f % a f % .
Note: Do not forget that still each P P and each a f a f in this case.
The strength utilization coefficient is equal to the maximum of the ratio
F P so that dF dN 0 at N = a. This idea allows us to determine the breaking
strain of bundle a and the bundle breaking force per one fiber Fmax .
Let us rewrite the second expression of Eq. (6.126) as follows
2 2
N
F
N %
5
f
%, N d%, where f
%, N S 1 % v
% ,
21 N
21
N
N af
1 a f
, 2 2
N N.
(6.128)
Then, it is valid to write that
O f
%, N
ON
N %
OS
1 %
d21
N
1
d2 2
N
v
% ,
,
1 . (6.129)
ON
dN
1 af
dN
15. We understand CV as a simple ratio of the standard deviation to the average value,
not as percentage.
228
CV% 0.01
1
F
P 0.8
0.1
%0
0.6
0.3
0.5
0.4
0.7
0.9
0.2
0
0.05
0.10
(a)
0.15 N 0.20
CV% 0.01
1
F
P 0.8
0.1
0.6 % 0
0.3
0.5
0.4
0.2
0.7
0.05
0.10
(b)
0.9
0.15 N 0.20
2
N
2 2
N O f
% , N
dF d 2
d2
N
d2
N
f 2 2
N , N 1
f 21
N , N
' 5 f
%, N d% ( 5
d% 2
dN dN ' 21
N
dN
dN
ON
( 21
N
N af
N %
N
N OS
af
N af
1
1 %
N
N
1
v
5
*S
v
% d% 1 * S
v
N
N
a f 1 a f
1 N
1 a f
ON
N a f
1
1 a f
1 a
N
N a f
N %
OS
N af
1 %
1
* S
a f v
v
% d%
1 a f
ON
1 a f
.
1 a f
(6.130)
Note: It was used S(0) = 0 according to Eq. (6.113), and
N af
N 1 a
f
N a f N Na f N a f a f
1 N
1 1 a 1 a N a 1 N a f .
f
f
f
229
0
aa f
a %
OS
a af
1 %
1
* S
a f v
v
% d%
,
Oa
1 a f
1 af
1 a f
aa f
1
a %
OS
1 %
v
% d%
Oa
1 a f
a af
1
* S
a f v
1 a f
1 a f
(6.131)
The root a of the last equation determines the breaking strain of the
bundle.
Applying especially Eq. (6.116) we can derive the following expressions
a %
OS
1 %
O P a% P 1
,
Oa
Oa ' a f 1 % ( a f 1 %
S
a f
P
af P .
af
(6.132)
(6.133)
and then by % a a f
a af
v
1 a f
1 a
f
1
e
a af
2
1 af
we obtain
aa f
ln
1 a f
22
(6.134)
Now, using Eq. (6.127) and Eqs. (6.132), (6.133), (6.134) in Eq. (6.131),
we can write
230
a a f
1
1 a f
a af
1
* S
a f v
1 a f
1 af
aa f
1
a %
OS
1 %
v
% d%
Oa
1
e
1 % %
ln % 2
1 a f
af
e
a af
where
aa f
P 1
af 1 %
1
e
2 %
ln % 2
22
d%
1 a f
1
*P
1 af
1
e
a af
2
1 a f
aa f
ln
1 a f
22
,
(6.135)
d%
2 2
aa f
ln
1 a f
2 2
ln %
CV%2 1
and ln CV%2 1
according to
Eq. (6.26).
The root a of the last equation, evaluated by a numerical method,
represents the breaking strain of the bundle in this special case. We
calculated also Fmax from Eq. (6.126) using Eqs. (6.116), (6.127) and N = a,
applying again a numerical method of integration.
The graphical illustration of both of the utilization coefficients 0P and
0a is shown in Figs. 6.12a and 6.12b. These quantities are decreasing with
the increase in the average value of fiber crimp % as well as with the
increase in the coefficient of variation CV%, but they are increasing with
the increase in the fiber breaking strain af.
Note: Remember the definition of breaking strain utilization coefficient
0a of a bundle consisting of average fibers whose crimp is % , so that
% max
5 v
% d% 5
v
% d% 1 for each values N %max. Now, let us think that
231
1
0P , 0a
0.8
1
0.6
0.4
0.2
0
0.2
0.4
0.6
0.8
a)
1
0P , 0a
0.8
0.6
0.4
CV%
0.2
0
0.2
0.4
0.6
b)
0.8
CV%
%max is limited to zero. Then (1) all fibers came to be straight (2) also all
N
values of % are limited to zero, and (3) the equation 5 v
% d% 1 is valid
0
for all values N 0. Applying this type of limitation in Eq. (6.123) we
obtain the expression as follows
N GA
67
G7 A P
a f
67 N
7 P
af
F 5 H5
g
a f da f 8 S
N v
% d % S
N H 5
g
a f da f 8 5 v
% d%,
7 N S
af
7I N S
a f
0I
97
97 0
A
F S
N 5
N
P
a f
S
a f
g
a f da f .
(6.136)
232
is generally a
to assume that P a f
P a f S
a f , a f 0, A .
(6.137)
Figure 6.13 illustrates this assumption. Let us think that some fibers have
their breaking points (gray) lying inside of an infinitesimally small strip of
width da f at an arbitrary point af. The strengths of such fibers have a
S
N f ,
P
S
N f
af
daf
af
N f , af
F S
N 1 G
N , G
N 5 g a f da f .
0
(6.138)
S
N
dS
N P
P
N,
af
af .
dN
233
(6.139)
F P a f N 5 g a f da f .
N
Let us consider that the fiber breaking strain af follows e.g. lognormal
distribution as shown below
1
e
2 ;a f
g af
; ln
af
CVa2f 1
ln a f ;
2 ;
,
(6.140)
; ln CVa2f 1
A
P af
dF d '
S
N 5
g a f da f (
(
dN dN '
N S af
234
F 1
P 0.8
0.01 CVa f
0.1
0.3
0.5
0.6
0.4
0.7
0.9
0.2
0
0.05
0.1
a)
0.15
N 0.2
0.01 CVa f
0.1
0.3
F 1
P 0.8
0.6
0.5
0.4
0.7
0.9
0.2
0
0.05
0.1
b)
0.15
N 0.2
values of CV a : (a) a f 0.07 (7%), (b) a f 0.15 (15%), {breaking points, --- bundle
f
consisting of identical fibers ( CV a = 0).
f
dS
N P
a
dN 5 S
a
dF dS
N P
a
dN
dN 5 S
a
g a da
g a da
A
dS
N A P a f
d ' P af
(
d
d
g
a
a
S
N
g
a
a
f
f
f
f
(
dN 5N S a f
dN ' 5N S a f
A
f
f
f
f
P
N
S
N '
g
N ( ,
' S
N
(
P
N g
N .
(6.141)
235
We know dF d = 0 at = a. Then,
0=
1=
( )
( )
( )
dS ( a) P a f
g a f da f P ( a) g ( a) ,
da a S a f
dS ( a ) P ( a f )
g ( a f ) da f
da a S (a f )
P (a) g (a)
(6.142)
Further, using the special value a on the place of af in Eq. (6.140) and
introducing the quantities
u=
ln a f
, ua =
ln a
,
(6.143)
1 1
a g ( a) =
e
2
(ln a )2
2 2
(ua )
1 1 u2a (ua )
e =
=
=
2
ln CVa2f + 1 ,
2
(6.144)
where (ua) is the PDF of standard normal distribution. According to
the second expression stated in Eq. (6.138) and Eq. (6.140), it is also valid
to write that
a
( )
G ( a) = g a f da f =
0
1
2 a f
(ln a f )
2 2
da f =
ln a
= ua
1 u2
e du = (ua ).
2
(6.145)
( a ) , was
f
236
1=
dS ( a ) P ( a f )
g ( a f ) da f
da a S (a f )
P (a ) g ( a )
P
af
P
af
g ( a ) da
f
S ( a) g ( a )
g ( a ) da
f
P
a g (a)
af
1 ( ua )
( ua )
,
(6.146)
ln CVa2f + 1
1 = ln CVa2f + 1
1 (ua )
.
(ua )
ua =
af
ln a
, u a = ln a , u a ln CVa2f + 1 = ln a ln
,
CVa2f + 1
u a ln CVa2f + 1 = ln a ln a f + ln CVa2f + 1 = ln
a
+ ln CVa2f + 1
af
= ln a + ln CVa2f + 1,
(6.147)
Applying the relation a = a a f , Eq. (6.139) for = a (that is,
S ( a) = P a f a ), and Eq. (6.145), we can rearrange Eq. (6.138), for = a
(where F = Fmax ), as follows
Fmax = S ( a ) [1 G ( a )] =
P = a 1 (ua ) .
P
F
a
1 (ua ) ,
a 1 (ua ) , max =
af
P
af
(6.148)
16. The probability density function (ua ) and the distribution function (ua ) of
standardized normal (Gaussian) distribution are generally known and still
remaining as the same functions.
237
1
0a
0.8
0.6
0.4
0P
0.2 Common
range
0.2
0.4
0.6
0.8 CV 1
af
Note: Let us remark that the utilization coefficients 0P and 0a are only
functions of CVaf in the studied case (linear forcestrain curves, lognormal
distribution, etc.), but also in a lot of other cases; e.g., for force strain
curves type S
N P a fc N c , where c is a parameter, for normal distribution
of af . May be, CV af is a crucial quantity that determinates both of the
utilization coefficients for a major part of real fiber bundles made up of
straight fibers. (The step-by-step derivation of Eqs. from (6.142) to
(6.148) are mentioned in Ref. [8]).
Fibers having variable crimp, strength, and breaking strain. Because
the universal expression for the average force per fiber in bundle stated in
Eq. (6.123) is concretized, which is also valid in this case, let us accept
the following assumptions that are already mentioned and explained in
previous sections:
theorem of similarity, stated in Eq. (6.114),
mutual independency of crimp, strength, and breaking strain
distributions, stated in Eq. (6.119),
linear forcestrain function of each fiber, stated in Eq. (6.116),
lognormal distribution of fiber crimp, stated in Eq. (6.127),
symmetrical strengths of fibers, stated in Eq. (6.137),
lognormal distribution of fiber breaking strain, stated in Eq. (6.140).
238
Let us rearrange Eq. (6.123) using Eqs. (6.137), (6.115) and the second
expression stated in Eq. (6.138)17 as follows
N A
P af
N %
F 5' 5
g a f da f ( S
v
% d%
1 %
'
(
0 N% S a f
' 1%
(
N
A
P
5 ' 5 g a f da f (
0 ' N%
( af
1%
N
N %
P
5 '1 G
1 % ( a f
F
1
P af
N %
v
% d%
1 %
N %
v
% d% ,
1 %
N%
1%
N % N %
N %
%
%
G
v
G
g a f da f .
1
d
,
where
50 ' 1 % ( 1 %
1 % 5
0
(6.149)
239
F 1
P 0.8
3
0.6
0.4
0.2
0
0.1
a)
0.05
F 1
P 0.8
0.15
N 0.2
3
0.6
0.4
0.2
0
0.1
b)
0.05
0.15
N 0.2
Eq. (6.137) so that Eq. (6.149) is valid too. Applying the third assumption
and using k P a f we obtain F
N
N%
k 5 1 5 g a f da f
N % v
% d% .
0
0
k N 5 v
% d%
0
N%
N 5 Q
% d% 5 % Q
% d% , where the function Q
% v
% 50
g a f da f
Q
% d% , then
k N 5 v
% d% 5 %v
% d% N 5 Q
% d% % 5 Q
% d% .
This is identical to what was derived by Cui, Suh, and Sasser [10].
240
6.5
ne
, % 0,1 .
n
(6.150)
241
II
(a)
Bundle
B
h
E
D
II
x
a
F
(b)
6.17 Bundle (sliver) of staple fibers gripped in jaws.
Simplest ideal fiber bundle with constant fiber length. The model of
simplest ideal fiber bundle is introduced in Section 6.1. This model is
considered here, but with a little modification as stated below:
A. The fiber bundle consists of a large number n of fibers.
B. The fibers are straight and parallel to the axis of the jaws, but they
need not be gripped by both the jaws simultaneously.
C. All of the fibers have the same length l (constant).
D. The positions of the individual fibers are random in the bundle.
According to these assumptions, it is possible to imagine a model of
fiber bundle where the fibers are distributed regularly across the bundle
and thus create a parallelogram as shown in Fig. 6.17b. Consider that the
length of one side of the parallelogram ADFE is l. While clamping such a
bundle in-between two jaws I and II separated by gauge length h l, it
may so happen that some of the fibers are gripped by both the jaws
simultaneously (thick lines in Fig. 6.17b) but the others are not. Further
consider that all n fibers are lying in the width of a of the fiber bundle,
however, a part of them ne are lying in the width x. Because the cross-
242
0
ne x
.
n a
(6.151)
The similarity of the triangles ABC and ADE as shown in Fig. 6.17b
and Eq. (6.151) allow us to write the following expressions
lh x
,
l
a
h
0 1 , h l,
l
0 0, h " l.
6
7
8.
79
(6.152) 18
One may remember that the model of simplest ideal fiber bundle
introduced in Section 6.1 accounts all n fibers for the tensile force applied
to the bundle. But, here, only a partial strip of ne fibers, width BC = x, is
functional while applying tensile force to the bundle and the rest of n ne
fibers are totally non-functional. It means that all forces and stresses are
required to be reduced by the ratio ne n 0 . This modification results in
mechanical quantities of the bundle as displayed in Table 6.9.
Table 6.9 Original and modified equations of mechanical quantities (+) of bundle.
Section 6.1
Quantity
Bundle force
Bundle breaking force
Bundle force per one fiber
Bundle breaking force per one fiber
Bundle specific stress
Equation
No.
Fb
N n F
N
(6.2)
F n F n F
a
b
Ff F
N
(6.3)
F F F
a
f
Bundle stress
;b
N ;
N
b
b
(6.4)
; ;
Ff F
N 0
Ff F F
a 0
b
N
N
Fb
N n F
N 0
Fb n F n F
a 0
Bundle strength
Modified equation
b
N
N 0
b 0
(6.6)
;b
N ;
N 0
;b ; 0
18. If h > l (i.e. if the upper jaw is lying over the point A shown in Fig. 6.17b) then
no one fiber can be gripped by both jaws simultaneously.
243
Note: It is shown in Table 6.9 that the modified equations always contain
two factors namely, length utilization coefficient 0 and force or stress of
fiber (forcestrain relation of fiber F(N), breaking force of fiber F* F(a),
specific stressstrain relation of fiber (N), specific strength of fiber *,
stressstrain relation of fiber ;
N , and strength of fiber ; ). The value
of 0 is 0 < 1 because all n fibers are not contributing to the bundle extension
in reality. In another way, it can be imagined that all n fibers are fully
functional, but all the mentioned functions and values of fiber mechanics
are smaller, reduced by the length utilization coefficient 0. Such
imagination will be found useful later on.
l1
T1 , n1, m1
j 1
l2
T2 , n2 , m2
j2
Tk , nk , mk
T,
n,
m
lk
jk
y
6.18 Doubling of partial slivers.
Simplest ideal fiber bundle with finite number of fiber lengths. Let us
consider the simplest ideal fiber bundle (sliver) which is created by
blending of fibers of different lengths l1, l2,,lj,,lk. Except the length,
all other fiber properties are considered to remain the same. It can be
understood that such a sliver is the result of doubling of k virtual partial
slivers, j = 1, 2,, k, where each of them is created by fibers of same
length. This is illustrated in Fig. 6.18. The fibers in the partial slivers are
arranged to resemble a parallelograms as shown in Fig. 6.18. Consider
that all partial slivers share a common length y.
Each j-th partial sliver has fineness Tj , mass mj = Tj y 19 , and the number
of fibers in the cross-section of the sliver, in accordance with Eq. (6.1) is
n j T j t , where t is fiber fineness. Further, in accordance with Eq. (6.152),
19. This accords to the general standard definition of the quantity fineness stated in
Eq. (1.3).
244
h
6
, h lj, 7
lj
8,
7
0 j 0, h " l j ,
9
0j 1
(6.153)
where h denotes the (common) gauge length. The resulting sliver has
fineness T, mass m = Ty, and contains n T t number of fibers in its crosssection. Then the following expressions are valid to write
k
m mj , T
j 1
k
k
k
mj
m
Tj , n n j .
y j 1 y
j 1
j 1
(6.154)
Let us express the mass fraction j of the j-th partial sliver, having fiber
length lj as follows
j
mj
m
mj
j 1
j 1
j
1.
(6.155)
Tj
nj
mj
y
Tj
t
m j
y
Tj
t
T j,
(6.156)
n j .
(6.157)
The number of fibers in j-th partial sliver that are gripped by both the
jaws simultaneously is expressed in accordance with Eqs. (6.150) and
(6.157) as follows
ne, j n j 0 j n j 0 j .
(6.158)
j 1
j 1
ne ne, j n j 0 j ,
(6.159)
0
k
ne
j 0j ,
n
j 1
245
(6.160)
lmax
l dl 1 .
0
5 0
l dl ,
0
lmax
0
h
5 1 l
l dl ,
0
0 0, h " lmax .
6
h lmax , 7
8.
7
9
(6.161)
246
Modified
First
component
Second
component
First
component
Second
component
F1
F2
F1
%1
F2
N % 2
1
N
1;
N
2
N
1
N %1
2
N % 2
2
;2
N
1;
N %1
;2
N % 2
F %1
F2 % 2
1
2
1 %1
2 % 2
Strength of fiber
;1
2;
1; %1
;2 % 2
1
1
247
Note: A reader can apply an analogous way of thinking also for parallel
fiber bundles with variable properties of staple fibers, i.e. to extend the
problem described in Section 6.4. Nevertheless, the process of solving
such a complex problem is generally complicated. In a general sense, we
need to work with conjugated probability density function of fiber breaking
strain, fiber specific stress, fiber crimp and fiber length. It is usually difficult
to find the required input values and functions in order to solve such a
problem. Even a suitable simplification for a specific fiber material is hardly
possible. (Let us point out an example of the problem of fiber crimp, where
it is necessary to use the crimped fiber lengths for the derivation of length
utilization coefficients).
Note: The cohesive force among fibers was neglected in this chapter
for simplification and because they are usually too small. However, they
can sometimes play a significant role. Remember, the cohesive force exists
among the fibers which are not gripped by both jaws simultaneously and
also among the fibers which are broken during the application of tensile
force to the fiber bundle. Based on our models, the determination of number
of such fibers is relatively easy. Nevertheless, a suitable expression for
fiber-to-fiber cohesion is required to be utilized in the aforesaid model.
6.6
1.
References
248
11.
12.
13.
14.
15.
16.
17.
18.
249
7
Mechanics of non-parallel fiber bundles
7.1
Multi-axial textiles
249
250
4
3
1
a)
j2
j1
j3
j4
1
3
b)
7.1 Scheme of multi-axial textile.
(a) Principle of textile; (b) Directions of systems.
We can imagine it such that both of the jaws are fastened to solid metal rods in
a tensile tester.
251
A
h
-
h;
B
-;
l;
Fp
B;
x
Fl
N
h; h h;
1, h; h
1 N ,
h
h
(7.1)
l; l l;
1, l ; l
1 N l .
l
l
(7.2)
h
h ; h
1 N
1 N
cos , cos - ;
l
l ; l
1 N l
1 Nl .
(7.3)
x 2 l 2 h 2 , x 2 l ; 2 h; 2 .
(7.4)
The strain of the yarn can be derived from the previous equations as
follows
l 2 h 2 l ; 2 h; 2 , 1
h 2 l ; 2 h; 2 l ; 2 h 2
2
2 2 2 2
1 N ,
l2
l
l
l
l
1 cos 2 -
1 Nl
1 N cos 2 -,
2
1 Nl
2.
1 cos -
1 N cos - ,
2
(7.5)
252
2
N l 1
1 N 1 cos 2 - 1 1 N 2 2N cos 2 - 1 .
(7.6)
It can be read from the last equation that a more oblique yarn (higher
value of -) displays smaller strain Nl at the same jaw displacement N.
Especially, if the yarn strain Nl is equal to the yarn-breaking strain a
then the yarn breaks itself. The corresponding jaw displacement is equal
to the value N = N*3 . Then the following relation is valid from Eq. (7.5)
1 a 2 1 cos 2 -
1 N
N
cos 2 -,
1 a 2 1 cos2 - 1
cos 2 -
1
a 2 2a
1.
cos 2 -
(7.7)
N 2 2 N ! 2N,
1 2N cos 2 - ! 1 N cos 2 - .
(7.8)
(7.9)
a N cos 2 -, N
a
.
cos 2 -
(7.10)
Forces on one oblique yarn in-between jaws. Let us express the force
strain relation of the yarn as follows
Fl Fl
N l , N l
0, A , Fl
0 0, Fl
a P .
(7.11)
Notes: We define this function for all values of Nl. The force Fl is equal
to zero for each non-strained yarn. The value P represents the breaking
3.
4.
253
force of the yarn, i.e., the maximum force at Nl = a. If the value of yarn
strain Nl is higher than a then the yarn is thought as broken, so that it
assigns a force generally smaller than P; nevertheless, we often introduce
Fl (Nl > a) is equal to zero.
The force Fl is also shown in Fig. 7.2. The vector of this force assigns
an angle -; taken from the direction of jaw axis. The vertical component
(experimentally measurable component which is parallel to the jaw axis)
of this force is equal to F. Using Eqs. (7.3) and (7.11) the following
expression can be written from Fig. 7.2
F Fl cos - ; Fl
N l cos -
1 N
1 Nl ,
(7.12)
cos 2 - a 2 2a 1 1
1 a
P
1 a 2 sin 2 1 a
P 1
sin 2 -
1 a 2
(7.13)
Note: Let us note that for the determination of vertical rupture force F*,
we need to know yarn-breaking force P and yarn-breaking strain a only,
we do not require to know the forcestrain curve of the yarn.
Simplest case of forces on oblique yarn in-between jaws. The simplest
case of an oblique yarn fulfills the following two assumptions in addition
to the four assumptions stated earlier in case of simple multi-axial textile.
5. The jaw displacement, yarn strain, and yarn-breaking strain are so
small that Eqs. (7.9) and (7.10) are valid enough well.
6. The forcestrain relation of the yarn is linear and it corresponds to
the following expression
254
P
6
N l
N l a, 7
a
8
Fl
N l 0
N l " a. 79
Fl
N l
(7.14)
X Y , 76
8 where X , Y are real quantities .
:
X , Y 0 if X " Y .79
(7.15)
Note: Then the following expression is valid to write for the introduced
function
:
CX , CY :
X , Y , where C is a positive value .
(7.16)
P
Nl .
a
(7.17)
Substituting the last function in Eq. (7.12) and then using Eqs. (7.17),
(7.9), (7.16), and (7.10), the vertical component of force F is expressed as
follows
F :
Nl , a
P
1 N
P
1 N
N l cos : N cos 2 -, a N cos3 a
1 Nl
a
1 N cos 2 -
a P
1 N
P
1 N
: N,
N cos3 : N, N N cos3 .
2
cos 2 - a
a
1 N cos 1 N cos 2 -
(7.18)
Nevertheless, the following approximated expressions can be applied
when the jaw displacement N is very small
1
! 1 N cos 2 -, N N 2 sin 2 - N 3 cos 2 - ! N .
1 N cos 2 -
(7.19)
Pa cos -
1 N cos -
1 N N
Pa cos -
1 N N cos
F : N, N
: N, N
- N 2 cos2 - N
255
Pa cos -
1 N N N sin
Pa cos -
N N
: N, N
: N, N
- N 2 cos 2 - N
sin 2 - N 3 cos 2 - : N, N
Pa N cos
-.
(7.20)
The vertical rupture force F* takes the value of force F at N = N*, where
N* is determined from Eq. (7.10). Then, we can write : N, N : N , N 1
F
P
a cos - P cos - .
a
(7.21)
Note: The last equation can also be derived directly from the scheme
shown in Fig. 7.2. Actually, the angle -; is practically equal to the angle
- in case of very small deformation.
Geometry of one-yarn system in-between jaws. Let us now imagine one
system of yarns, fulfilling the four assumptions of a simple multi-axial
textile, gripped in-between the couple of jaws A and B as shown in Fig. 7.3.
The number of yarns & per length unit (measured across the direction of
yarns) characterizes the density of yarns in the system. In coherence with
the textile terminology5, let us call it sett of yarns.
We arrange to grip a strip of textile of width c, including the yarn system,
in-between the jaws of a tensile tester. This however does not mean that all
the yarns in the yarn system are gripped simultaneously by both of the jaws.
The yarns near to the edge of the strip can be held by only one jaw, so that
they cannot contribute to the force resulting from straining of such yarn
system. As shown in Fig. 7.3, the effective width of the strip is c; and the
non-effective width of the strip is :. Then, the total width of the strip is
c = c; + :.
(7.22)
All the yarns in the yarn system are inclined with the (non-oriented)
angle - taken from the vertical direction of the jaw axis. The following
expression is then valid from the geometry of the yarn system as shown in
Fig. 7.3
tan -
5.
:
.
h
(7.23)
In weaving technology, the traditional terms warp sett and weft sett are
commonly used.
256
c;
-b
A
h
B
c;
Detail:
Detail
c
- P. 1 &
U y Q
c ; c : c h tan - .
(7.24)
The dotted diagonal line shown in Fig. 7.3 includes the boundary angle
-b taken from the vertical direction of the jaw axis. It is shown that
Eqs. (7.22), (7.23), and (7.24) are valid only when the angle - < -b. If
- -b then c; = 0 and : = c, i.e., no one fiber is simultaneously gripped
by both of the jaws.
The following relations are evident from Fig. 7.3
tan - b
c
c
, h
h
tan - b .
(7.25)
6
tan -
c
tan - c 1
- < -b , 7
tan - b
tan - b
8
7
c ; 0
- - b .
9
c; c
(7.26)
c ; c M
-, - b ,
(7.27)
257
where
tan 6
- < -b , 7
tan - b
8 especially, M
0, - b M - , 1.
2
7
M
- , - b 0
- - b ,
9
M
-, - b 1
(7.28)
Evidently, the function M (-, -b) expresses the influence of the borders
of the yarn system.
The distance y between the two gripping points of neighboring yarns is
shown in Fig. 7.3. The perpendicular distance between the two neighboring
yarns must be equal to the reciprocal value of the yarn sett, i.e., 1 & . As the
angle - is one of the angles of the triangle UPQ6, it is valid to write that
y
1&
1
.
cos - & cos -
(7.29)
Using Eqs. (7.27) and (7.29) the number N of gripping points in the
effective width c; is obtained as follows
N
c;
&c cos - M
-, - b .
y
(7.30)
S
SSYSTEM F N
.
c
c
(7.31)
The gripping line UQ of lower jaw is perpendicular to the jaw axis (dot-and-dash
line); the dashed line PQ is perpendicular to the yarn line UP. Therefore, the
angle UQP must be also equal to -.
258
where Eq. (7.6) describes N l and Eq. (7.28) determines the function
M (-, -b). If we know the input quantities, i.e., c, h, -, sett &, jaw
displacement N, and forcestrain function of yarn F l (N l), then we can
calculate the nominal force S from Eq. (7.32) by using Eqs. (7.6), (7.25)
and (7.28).
We obtain the maximum of nominal force at angle - = 0. Then the
following triplet of equations is valid from Eqs. (7.6), (7.7), and (7.28)
N l N, N a, M
-, - b M
0, - b 1 .
(7.33)
(7.34)
S
FN
.
c
(7.35)
Using Eqs. (7.13) and (7.30) the nominal breaking force is derived from
Eq. (7.35) as follows
sin 2 'P 1
1 a 2
'
S
( &c cos - M
-, - b
(
sin 2 &P cos - 1
M
-, - b .
c
1 a 2
(7.36)
(7.37)
259
P
3
': N, N a N cos - ( &c cos - M
-, - b
&P
S
: N, N
N cos 4 - M
-, - b .
c
a
(7.38)
&P
N.
a
(7.39)
Equation (7.21) is also valid in place of Eq. (7.13) in the earlier simplest
case (small deformation and linear forcestrain relation of yarns). Using
the mentioned equation and Eq. (7.30) in Eq. (7.35) we obtain the following
expression for the nominal breaking force
S
&P cos2 - M
-, - b .
(7.40)
We find the maximum nominal breaking force S -
0 from the last equation
at - = 0, i.e., also M (-, -b) = 1 according to Eq. (7.33). So we obtain
S -
0 &P which is identical to the expression mentioned in Eq. (7.37).
Specific stress and specific strength of one-yarn system. Let us imagine
a part of a strip of oblique yarn system, having a width equal to one, shown
in Fig. 7.4a. The total length of yarns in the rectangle TUUT is L. Let us
extrapolate the yarns at the right-hand side of the scheme as shown by
the dashed lines in Fig. 7.4a. It is evident that the length of yarns in the
triangle TUV is same as the dashed length in the triangle TUV.
Therefore, the total yarn length L in the rectangle TUUT is equal to the
total yarn length in the parallelogram TVVT.
1R
y
V
V;
U;
h
T
T;
1
(a)
(b)
260
1 y , where the distance y is given by Eq. (7.29). The length of one yarn,
inclined at angle - taken from the jaw axis, is h cos - . Thus the total
length of yarns in the rectangle TUUT or the total length of yarns in the
parallelogram TVVT is
1
h
h
L
& cos &h .
cos y cos -
(7.41)
(7.42)
The (vertical) length of unitary strip of yarn system is equal to the gauge
length h shown in Fig. 7.4. So, the fineness T1 of such unitary strip of yarn
system is
T1
m
T& G ,
h
(7.43)
S
S
S
.
T1 T & G
(7.44)
(The definition according to Eq. (1.21) was used). Using Eq. (7.32) in
the last equation we can write
7.
261
F
N
1 N
1
1 N
& cos 2 - l l
M
- , - b l
N l cos 2 M
-, - b ,
1 Nl
1 Nl
T&
(7.45)
where Eq. (7.6) describes the quantity N l as a function of N and -,
Eq. (7.28) determines the value M (-, -b), and
l
Nl
Fl
N l
T
(7.46)
(7.47)
S S S
.
T1 T & G
(7.48)
1
sin 2 sin 2 &P cos - 1
M
- , - b l cos - 1
M
-, - b ,
2
T&
1 a
1 a 2
(7.49)
P
T
(7.50)
(7.51)
262
Note: This case corresponds to Fig. 7.4b where the previous equation is
immediately obtained.
Specific stress and specific strength of yarn system for simplest case.
The easier equations are valid in the earlier mentioned simplest case. Using
Eqs. (7.38) and (7.50) in Eq. (7.44) we can write
1
&P
N
: N, N
N cos 4 - M
-, - b : N, N l cos 4 - M
- , - b .
T&
a
a
(7.52)
N
.
a
(7.53)
1
&P cos 2 - M
-, - b l cos 2 - M
- , - b .
T&
(7.54)
8.
The subscript i will be automatically added to all quantities which relate only
to the i-th system in previous equations.
263
Independent Variable: N.
:
X , Y 1 X Y , 76
8
:
X , Y 0 X " Y .79
(7.25) tan -b c h .
(7.28)
M
-, -b 1 tan - tan -b
- < -b , 67
8 especially M
0, -b M -, 1 .
M
-, -b 0
- -b ,
2
79
(7.50) l P T .
General Case
Input function: Fl
Nl .
(7.6)
Nl 1 N 2 2N cos 2 - 1 .
(7.7)
N 1
a 2 2a
1 .
cos 2 -
(7.9)
Nl N cos 2 - .
(7.10) N
a
.
cos 2 -
(7.17) N : N , a P N .(+)
l
l
l
l
Ta
(7.46)
(7.46) l
Nl Fl
Nl T .
2
(7.32) S & cos -
Simplest Case
(small deformations, linear forcestrain
function of yarns)
P (+)
(7.17) Fl
Nl :
Nl , a Nl .
a
Fl
Nl
1 N
&P
M
-, -b . (7.38) S :
N, N
N cos 4 - M
-, -b .
1 Nl
a
(7.39) S-0 :
N, a
(7.36) S &P cos - 1
sin 2 -
1 a
M
-, -b . (7.40) S &P cos 2 - M
-, -b .
(7.37) S-
0 &P .
(7.37) S-
0 &P .
(7.50) l P T
(7.50) l P T .
2
(7.45) l
Nl cos -
1 N
M
-, -b .
1 Nl
(7.47) -0 l
N .
(7.49) l cos - 1
&P
N.
a
(7.52) :
N, N l
N
cos 4 - M
-, -b .
a
(7.53) -0 :
N, a l
sin 2 -
1 a
N
.
a
M
-, -b . (7.54) l cos 2 - M
-, -b .
(7.51) -0 l .
264
A
B
the subscript i. They are jaw displacement N, the parameters of the adjusted
tensile tester, i.e., c, h and -b, and the symbols expressed in function :
Equation (7.15), and function M Equation (7.28).
The nominal (vertical) force SS expresses the vertical force per unitary
strip of the multi-axial textile 9. Applying Eq. (7.32) we obtain
K
K G
67
Fl ,i N l ,i
1 N
7
SS Si H &i cos 2 - i
M
-i , - b 8 ,
1 N l ,i
i 1
i 1 7
I
97
(7.55)
N l ,i 1 N 2 2N cos 2 - i 1 ,
(7.56)
tan - i
6
-i < - b , 7
tan - b
8 especially M
0, - b M - i , 1.
2
M
- i , - b 0
- i - b , 79
M
-i , - b 1
(7.57)
The maximum of nominal (vertical) force SS,- = 0 expresses the vertical
force per unit width of the multi-axial textile at -i = 0, i = 1,2,, K. Then
the following relations are valid similarly to the triplet of Eq. (7.33)
N l ,i N, Ni ai , M
- i , - b M
0, - b 1, i 1, 2, , K .
(7.58)
i 1
i 1
(7.59)
The quantities related to the whole multi-axial textile have the subscript S.
265
S , because the jaw displacement Ni at the time of rupture is different for
different yarn systems; actually it is valid according to Eq. (7.7)
S
N i 1
ai2 2ai
1.
cos 2 - i
(7.60)
(7.61)
(7.62)
266
(7.63)
where Si is the nominal (vertical) force, acts on the unitary width of the
i-th system and the equality
T1,i = T i&i = G i
(7.64)
to Eq. (7.64)
K
i 1
i 1
i 1
TS T1,i
Ti &i Gi GS ,
(7.65)
where GS means the areal mass density of the whole multi-axial textile
(thicker yarns as shown in Fig. 7.6).
The total nominal (vertical) force SS acting on the unitary width (of
unitary strip) of the whole multi-axial textile is equal to the summation of
GS
SS
7.6 Example of unitary strip and areal unit of multi-axial textile.
267
the nominal (vertical) forces acting on all yarn systems. Using Eq. (7.63)
it is possible to write the following expression
K
i 1
i 1
i 1
SS Si
iTi &i
i Gi .
(7.66)
We can write the expression for the specific stress S of the whole multiaxial textile using the definition of specific stress according to Eq. (1.21).
Applying Eqs. (7.65) and (7.66) it is valid
K
S
SS
TS
gi
Gi
,
GS
G
i
i 1
GS
i gi ,
(7.67)
i 1
where
K
g
1
i 1
(7.68)
(7.69)
where Eq. (7.56) is used for Nl,i and Eq. (7.57) is used for M (-i, -b).
The maximum of specific stress of the multi-axial textile S,-=0 is
obtained from Eq. (7.69) at -i = 0, i = 1, 2,, K, where the expressions
mentioned in Eq. (7.58) are also valid. We can therefore write
K
S ,- 0 gi l ,i
N .
(7.70)
i 1
268
l ,i
S N ' g i : N, N i
cos 4 - i M
- i , - b ( .
ai
i 1
K
(7.71)
(7.72)
(7.73)
where S,-=0 > 0. This utilization coefficient generally varies with the
jaw displacement N when Eqs. (7.69) and (7.70) are used in the definition
expressed in Eq. (7.73). On the other hand, we may use Eqs. (7.71) and
(7.72) if the simplest case can be assumed. Further, if the jaw
displacement is smaller than each values a i, i.e., all values : (N, a i) = 1
according to Eq. (7.15), then the angular utilization coefficient
K
K
0a ' gi l ,i cos 4 - i M
- i , - b ( ' gi l ,i ( is independent of the jaw
ai
i 1
i 1 ai
displacement N.
Active and passive groups of yarn systems. The areal mass density of
each yarn system Gi and the total areal mass density GS of the whole multiaxial textile can be evaluated by using Eqs. (7.64) and (7.65). The border
angle -b can be determined by using Eq. (7.25).
Let us now divide all yarn systems creating a multi-axial textile into
two groups.
(a) The first group contains the yarn systems whose angles are given by
-i < -b, i.e., M (-i, -b) > 0 according to Eq. (7.28). We call this group
as the active group of yarn systems in the multi-axial textile in
short active group because such yarn systems are actively
contributing to the forces and stresses of the multi-axial textile. The
total areal mass density of this active group of yarn systems is
GA,S
Gi
i 1,2, K ,
-i <- b
269
(7.74)
i 1,2, K ,
- i - b
Gi
(7.75)
GS Gi
i 1
i 1,2, K ,
- i <- b
Gi
i 1,2, K ,
-i - b
Gi GA,S GP,S
(7.76)
It is also valid
K = K A + K P.
(7.77)
It is evident that only the active group of yarn systems determines the
mechanical behavior (including strength) of the multi-axial textile.
Specific strength (tenacity) of multi-axial textile. The problem of specific
strength (and/or breaking force) of the multi-axial textile can be solved by
means of active group of yarn systems only. Quite generally, the solution
follows the commentary mentioned after Eq. (7.60). We must find out the
total maximum of the specific stress corresponding to all values of jaw
displacement. Such a maximum represents the specific strength (tenacity)
and the corresponding value of jaw displacement refers to the jaw
displacement at the time of rupture of the multi-axial textile.
Nevertheless, the forcestrain relation Fl,i(Nl,i) of yarn from each i-th yarn
system is often equal to zero when its strain Nl,i is higher than the breaking
strain of the yarn ai. Such a function fulfills besides Eq. (7.11) the
mentioned assumptions; see the scheme shown in Fig. 7.7. It is therefore
valid to write that
270
Fl ,i
N l ,i
Pi
a
b
ai
Nl ,i
(Nl,i) of yarn.
Fl ,i N l ,i ai
increasing function, Fl ,i
0 0, Fl ,i
ai Pi , 67
8 i 1, 2, K .
Fl ,i N l ,i " ai 0,
79
(7.78)
Note: The forcestrain function of the multi-axial textile for the simplest
case, determined by Eq. (7.14), is a type of the function Fl,i(Nl,i) according
to Eq. (7.78).
Note: If Nl,i > ai, i.e., also N i " N i according to Eq. (7.7) or Eq. (7.10),
then all the forces and the strengths of the i-th system are equal to zero,
because all yarns are broken.
Let us first study the active group of yarn systems only. (We can imagine
it as a part of multi-axial textile where all the yarn systems are
mechanically active). We can calculate the jaw displacement Ni at the
time of rupture (breaking strain of yarn system) according to Eq. (7.7) or
Eq. (7.10) (also in Table 7.1) for each yarn system of active group. Then
let us rename the values of the subscripts i so that it is valid N i 1 N i ,
i = 1,2,, KA, , where N 0 0 . (Compare it with the introductory part of
Section 6.3). After renaming, Eq. (7.74) takes the following form
KA
GA,S Gi .
(7.79)
i 1
Further, let us define the mass fractions of the individual yarn systems
in the active group (a part of multi-axial textile) analogically to Eq. (7.68)
g A,i
Gi
,
GA,S
KA
g
i 1
A,i
1.
(7.80)
271
S A,S
S A, S
GA, S
GS
GA, S
GS
(7.81)
(7.82)
GA, S
GS
(7.83)
272
Symbol
Initial quantity
Symbol
Subscript of component
KA
Number of components
gA,i
gi
N
Strain of bundle
i
i
i
Ni
ai
i (N)
A,S
A, S
b
A, S ,min
b,min
gA,i,min
gi,min
b(N)
Note: Equation (7.81) produced the same results as Eq. (7.69) and/or
Eq. (7.71).
Similarly, the mass fractions gA,i and gA,i,min can be recalculated to mass
fractions related to the whole multi-axial textile as follows
gi g A,i
GA, S
GS
KA
i 1
GA, S
GS
(7.84)
where the ratio GP, S GS is the areal mass density of overall yarn systems
creating the passive group of yarn system. (The mass relations of the passive
group of yarn systems do not play any role).
7.2
273
General
symbol
Common symbol
by regular textiles
Ti
Sett of yarns
&i
&
Fl,i(Nl,i)
Pi
ai
Fl(Nl,i) (+)
P
a
Further, according to Eq. (7.64), the areal mass density of each yarn
system is G = T&; according to Eq. (7.65), the areal mass density of the
multi-axial textile is GS i 1 G KG ; and according to Eq. (7.68), the
K
G
G
1
GS KG K .
(7.85)
l Nl , i Fl N l , i T .
(7.86)
274
a)
jaw
axis
i 3
i2
i4
2
3
i 1
K
K K
K
K
i5
6
K
i6K
line of jaw
b)
l P T .
(7.87)
by
2 , 2 . In the previous chapter, the non-oriented angle -i is given
275
- i 0, 2 .
(7.88)
It is evident from Fig. 7.8b that the angular difference between the
neighboring systems is constant and is equal to the value K . The angle
K of the last K-th system must lie in the interval
K
,
2 K 2 ,
(7.89)
as shown in Fig. 7.8b, and the following equation is valid for each of
angle i
i
K i
K
i
, i 1, 2, , K .
K
(7.90)
So, if we know the angle K of the last yarn system then we can calculate
each of angle i according to Eq. (7.90).
The equations in Section 7.1 are written in terms of the non-oriented
angles -i in the form of goniometrical functions of cos-i, sin2-i and tan -i.
Now, the following equalities are valid according to Eq. (7.88)
cos - i cos
, sin 2 - i sin 2
, tan - i tan
(7.91)
Applying the symbols mentioned in Table 7.3 and equations from (7.85)
to (7.88), and (7.91), we can obtain a set of equations for the regular multiaxial textiles from the previous chapter.
Forces in regular multi-axial textile. The nominal (vertical) force SS
follows Eq. (7.55) and takes the following expression
K G
7
SS &
1 N Hcos 2
i 1 7
I
1 N
Fl N l ,i
i
l ,i
67
, -b 8 ,
79
(7.92)
N l ,i 1 N 2 2N cos 2
M
, -b 1
M
tan i
tan - b
, - b 0
1,
6
< -b , 7
8 especially, M
0, - b M
- b , 79
(7.93)
, 1.
2
(7.94)
276
The maximum of nominal (vertical) force SS,-=0 follows Eq. (7.59) and
takes the following expression
K
SS ,- 0 & Fl
N &K Fl
N .
(7.95)
i 1
The jaw displacement Ni at the time of rupture of the i-th yarn system
follows Eq. (7.60) and takes the following expression
N i 1
a 2 2a
1.
cos 2 i
(7.96)
Forces in regular multi-axial textile for simplest case. The nominal (vertical)
force SS follows Eq. (7.61) in this case and takes the following expression
SS
&P K
N : N, Ni cos4
a i 1
, -b
(7.97)
and the maximum of nominal (vertical) force SS,-=0 follows Eq. (7.62)
and takes the following expression
SS ,- 0 :
N, a
&P K
&P
N1 K
N :
N, a .
a i 1
a
(7.98)
1 N
K
'cos
i 1
1 N l ,i
l N l ,i
'
l N l ,i
i
1 N l ,i
, -b (
(
, -b (
(
(7.99)
and the maximum specific stress S,-=0 follows Eq. (7.70) and we get
K
S ,-0 g l
N 1
i 1
1
l
N K l
N .
K
(7.100)
S g
l K
N : N, Ni cos 4
a i 1
l K
N : N, Ni cos 4
Ka i 1
277
, - b
, - b .
(7.101)
The maximum specific stress S,-=0 follows Eq. (7.72) and takes the
following expression
S ,-0 g
K
l
N :
N, a 1 l N :
N, a K l N :
N, a .
a
Ka
a
i 1
(7.102)
0a
1 N K 'cos 2
K l
N i 1 '
1 N
l Nl , i
i
l,i
, - b (,
(
(7.103)
where l (N) > 0. Especially, the angular utilization in the simplest case
follows Eqs. (7.101) and (7.102). Then, it is valid to write that
l K
N : N, N i cos 4 i M
i , - b
Ka i 1
0a
l
N :
N, a
a
K
1
: N, N i cos 4 i M
i , - b ,
K :
N, a i 1
(7.104)
where N a.
Note: In accordance with the commentary stated below Eq. (7.73), this
angular utilization is independent of the jaw displacement N.
Special alternative to the simplest case of regular multi-axial textile.
We often use regular textiles prepared from a very few expensive fibers or
yarns, which have roughly linear forcestrain relation (i.e. corresponding
to our simplest case glass, kevlar, etc.) in different types of technical
textiles. Nevertheless, these structures are not primarily developed for
application in a tensile testing machine, but for use in specific technical
applications, where
278
(a) All yarns are fixed (sewn, glued, etc.) in the structure of the product;
the free ends of yarns do not exist.
(b) The applications do not demand high deformation of the yarns that
the breakage of the yarns would occur. It is assumed that all yarns
will be strained without breakage during applications.
We will call such a simplest case of regular multi-axial textile as special
alternative in short.
The assumption (a) can be interpreted with a view to in an (imaginary)
tensile testing machine as a case shown in Fig. 7.9. Here the gauge length
is much smaller than the width of multi-axial textile under the jaws, h << c.
Then the angle -b approaches to its value 2 , then the value tan-b A
and the value
M
-i , - b M
, -b 1
(7.105)
according to Eq. (7.57). (The free ends of the yarns near the edge of
the jaws can be neglected).
1 Eq. (7.96) is
valid for each yarn system and then the assumption (b) is also valid.
Note: Evidently, this inequality is always valid, when N a.
Moreover, if N N i then
: N, Ni 1 .
(7.106)
0a
1
K
cos
4
i
i 1
1
K
neglected
A
h
B
7.9 Neglecting margins.
cos
i 1
i
.
K
neglected
-b
(7.107)
279
known formulas
cos 2 )
1 1
cos
2) , and
2 2
cos
) E cos ) cos E sin ) sin E , we can write the following expressions
cos 4
i
K
i
' cos K (
K
i
cos 4 K 'cos 2
K
2
1 1
i
' cos 2 K 2 (
K
2 2
1 1
cos 2
4 2
1 1
cos 2
4 2
3 1
cos 2
8 2
3 1
cos
2
8 2 '
1
' cos
4
8
3 1
cos
2
8 2
1
cos
4
8
i
(
K
i
1
2 cos 2 2
K 4
i
1 1 1
2 ' cos 4
K 4 2 2
i
1
2 cos 4
K
8
cos
i
2 sin
2
K
i
4 sin
4
K
i
1
2 sin
2
K cos
K 2
K
cos
cos
i
1
4 sin
4
K 8
i
4 (
K
i
4
K
i
2
K
sin
i
2 (
K
i
4 (
K
i
2
K sin
K
sin
sin
i
4 .
K
(7.108)
Substituting Eq. (7.108) in Eq. (7.107) we obtain the angular utilization
in the case of special alternative to the simplest case of regular multi-axial
textile in the form as follows
0a
3 1
cos
2
8 2K
1
cos
4
8K
cos
i
1
2
sin
2
K 2K
cos
i
1
4
sin
4
K 8K
i 1
K
i 1
sin
i
2
K
sin
i
4 .
K
i 1
K
i 1
(7.109)
280
Note: We used
1 K i 1
3 8
1 K K
3 8 3 8 .
K
There are four sums in Eq. (7.109). However, they can be solved
analytically as shown below.
K
Evaluation of sums
cos K 2
and
i 1
(7.110)
where j is the imaginary number (j2 = 1). Especially, the first complex
number (i = 1) is
2
2
w1 cos j sin .
K
K
(7.111)
i
i
2
2
cos 2 j sin 2 'cos j sin ( ,
K
K K
K
wi w1i .
(7.112)
sK i 1 wi w1
K
1 w1K
. Using Eqs. (7.110), (7.111) and (7.112) in the
1 w1
K
K
1 cos 2 j sin 2
K
1 w
1 wK
K
sK wi w1
w1
w1
1 w1
1 w1
1 cos
2 K j sin
2 K
i 1
K
1
w1
1 cos
2 j sin
2
.
2
2
1 cos j sin
K
K
(7.113)
10. Generally, it is valid that cos
i) j sin
i)
cos ) j sin ) ; this is known as
i
Moivres theorem.
11. Generally, the sum of the first K members of a geometrical progression + ai , is
a1 1 q K
1 q , where q is the quotient of this geometrical progression.
281
(7.114)
A complex number is equal zero when the real part and the imaginary
part are equal to zero. Thus
K
Evaluation of sums
(7.115)
i 1
cos K 4
and
i 1
(7.116)
4
4
v1 cos j sin .
K
K
(7.117)
i
i
4
4
cos 4 j sin 4 'cos j sin ( .
K
K K
K
i
vi v1 .
(7.118)
K
K
1 cos 4 j sin 4
K
K
1 v1K
1 vK
sK; vi v1
v1
v1
1 v1
1 v1
1 cos
4 K j sin
4 K
i 1
K
v1
1 cos
4 j sin
4
.
4
4
1 cos j sin
K
K
(7.119)
282
The numerator of the last quotient is also equal to zero as in the previous
case. But, the denominator is different from zero for each K 3. Then,
K
K
i
i
0 sK; vi 'cos 4 j sin 4 (
K
K
i 1
i 1
K
K
i
i
cos 4 j sin 4 ,
K
K
i 1
i 1
i
cos 4 0,
K
i 1
K
sin K 4 0,
K 3.
(7.120)
i 1
(7.121)
i
cos 4 2,
2
i 1
2
sin 2 4 0,
K 2.
(7.122)
i 1
i 1
i
cos 4 2
K 2,
K
i 1
K
i 1
i
cos 4 0
K 3.
K
i 1
K
8
7
79
(7.123)
283
3 1
cos
4
0a
8 8* 2
3 1
cos 2
2
8 8
3 1
cos 2
2
8 8
3 1
cos
4
8 8
2
2 sin
2 2
2
*2
1 1
cos 2
2
2 1 cos
2 2
4 4
2
(7.124)
For K 3:
3
0a .
8
(7.125)
12. Many years ago, Mr. Vladimr Svat, who is the originator of the principle and
construction of air-jet looms, passed this information to one of authors of this
book.
284
1
2
0.5
0a
K 2
38
K 3
0.25
1
4
6
2
6
2
woven fabrics developed too high stress concentration which finally led
to the destruction of the parachutes during usage. This encouraged them
to initiate the development of a weaving machine for the production of
three-axial woven fabric and the parachutes prepared from the three-axial
woven fabric were then found as fully successful.
The regular multi-axial textiles are often used also as reinforcement in
textile composites and/or similar products where the mechanical isotropy is
required. The regular four-axial textile ( 1 = 45, 2 = 0,
3 = 45,
4 = 90 ) is most frequently used for such usage, because the
longitudinal and the transversal directions (0 and 90) can be technologically
realized more easily than the other two directions. Similarly, in a three-axial
textile, the longitudinal direction 2 = 0 is technologically easy to realize,
but the other two directions ( 1 = 60, 3 = 60) are difficult to realize.
7.3
The linear textiles (fibers and yarns) do have a finite number of directions
as discussed in the previous chapters. However, the products created from
continuously distributed directions of suitable linear textiles most
frequently fibers are also often used (for example, fibrous webs required
for preparing non-woven textiles, etc.). Therefore, in spite of the previous
chapters, we call the linear textiles which are creating products that are
solved in this chapter as fibers.
285
Simple planar bundle. Let us think of a simple planar bundle that satisfies
the following assumptions.
1.
2.
3.
4.
Note: Try to compare this set of assumptions with the definition of simple
multi-axial textiles as mentioned in Section 7.1.
Basic idea of solution. Let us imagine a multi-axial textile as described
in the previous chapters which is created from a very high number K of
systems so that the individual values gi of the mass fractions [Eq. (7.68)]
are very small. Such a multi-axial textile is approaching to a planar yarn
and/or fiber bundle made up of fibers showing continuously distributed
directions. So, we can perceive that each planar bundle as a multi-axial
textile composed of infinite number of systems (K A), which are oriented
to all directions and have very small mass fractions. The angular
distribution of such fibers was studied in detail in Chapter 3.
Mass fraction. Let us imagine the strip width c of a planar bundle situated
in-between jaws A and B of (real or imaginary) a tensile testing machine
as shown in Fig. 7.11. The (mean) number of fibers present in the unitary
width of jaw B (marked by symbol in Fig. 7.11) was called & in Section 3.5
and according to Eq. (3.56) it is valid that &
G t k n , where G is the
areal mass density of the whole planar bundle and t is the fineness of the
fibers. The quantity kn is generally expressed according to Eq. (3.18) as
follows kn 5
2
cos - u
- d- , where u
- , - 0, 2 is the probability
1
A
h
B
c
&
286
planar bundle. (The angle - is measured from the vertical jaw axis see
this angle made by one general fiber in Fig. 7.11).
The horizontal gripping line of jaw B can be considered as a sectional
line, too. The probability density function of direction of fibers protruding
from the gripping line of jaw B (sectional line), was called u*(-) in
Chapter 3 and it was derived in Eq. (3.20) that u
- cos - u
- k n .
Thus, the relative frequency of fibers protruding from the gripping line of
jaw B and having an angle - is u*(-)d-. The number of such fibers, going
out from the unit width of jaw B, is very small and it then takes the
following expression
d& = &u*(-) d-U
(7.126)
h
h
h s
s d&
s & u
- d- &
u
- d- . (7.127)
cos cos cos -
dmh &
cos - u
-
h
ht
d- & u
- d- .
t
kn
kn
cos -
(7.128)
The total mass of fibers present in the rectangle of area 1*h as shown in
Fig. 7.11 is then
- 2
mh
- 0
dmh &
ht
kn
2
ht
5 u
- d- & k
0
(7.129)
287
dg
dmh
mh
&
ht
u
- dkn
u
- d- .
ht
&
kn
(7.130)
Specific stress of simple planar bundle. The specific stress of multiaxial textiles is expressed in general by Eq. (7.69). In order to apply our
basic idea of solution to this, we must make the following changes:
We use the general angle - in place of -i.
We use the general fiber strain Nl according to Eq. (7.6) in place of
Nl,i.
We use the general specific stress in fiber l(Nl), defined by Eq. (7.46),
in place of l,i (Nl,i).
We use the general form of the function M (-, -b) influence of
borders or margins of the jaws according to Eq. (7.28) in place of
M (-i, -b).
We use the mass fraction dg according to Eq. (7.130) in place of gi.
Finally, we use the definite integral
- 2
- 0
in place of
K
i1
S
- 0
dg l
Nl cos 2 -
2
5 u
-
N cos
l
1 N
M
-, - b
1 Nl
1 N
M
-, - b d-,
1 Nl
(7.131)
288
S , - 0
- 0
dg l
N l
N
2
5 u
- d-
N .
l
(7.132)
in place of : N, N i .
S N
dg : N , N
- 0
N
l
a
l
cos 4 - M
-, - b
a
2
5 u
- :
N, N cos
- M
- , - b d-
(7.133)
S , - 0 N
- 0
2
dg :
N, a l :
N, a l N 5 u
- d- :
N, a l N (7.134)
a
a 0
a
0a
S
S ,-0
2
u
-
N cos
N 5
l
1 N
M
-, - b d-, l
N " 0 ,
1 Nl
(7.135)
289
0a
S
S ,- 0
1
:
N, a
l
a
2
5 u
- :
N, N cos
- M
- , - b d-
:
N, a
2
5 u
- :
N, N cos
l
N
a
(7.136)
- M
-, - b d-, N a.
l
N
2
5 u
-
N cos
l
1 N
d- ,
1 Nl
(7.137)
where Nl is given by Eq. (7.6), and Eq. (7.136) simplest case can be
written by
2
0a
5 u
- cos
- d- .
(7.138)
290
0a
2
2
5 cos
- d- .
(7.139)
However, it is valid 14
2
2
1 1
1
1 1
cos - cos 2 - ' cos
2- ( cos
2- cos 2
2-
4 2
4
2 2
1 1
1 1 1
1
3 1
cos
2- ' cos
4- ( cos
2- cos
4- ,
4 2
4 2 2
8
2
8
(7.140)
4
2 3 1
1
cos
2- cos
4- ( d'
8
8 2
0
2
2
2
G
67
27 3
1
1
H 5 d- 5 cos
2- d- 5 cos
4- d- 8
I7 0 8
2
8
0
0
97
2
2
G
2 73 2 1 1
1 1
67
H - 0 ' sin
2- ( ' sin
4- ( 8
I7 8
2 2
8 4
0
0 79
2 G 1
1 6 3
H 0 08 .
I2 4
32 9 8
0a
(7.141)
0a
1
5 '' C
C
2
C
C
1
( cos 4 - d-,
2
2
2
2
1 cos
- ) C C 1 cos
- ) (
(7.142)
1 1
14. The formula cos ) cos
2) is used.
2 2
2
291
0a
C 2 4
4
2
2
'
2C 1 cos ) 6 cos ) sin ) C sin ) ( . (7.143)
2
C 1
C
1
0a 3cos 4 ) 6 cos 2 ) sin 2 ) 3sin 4 )
8
2
3
3
3
cos 4 ) 2 cos 2 ) sin 2 ) sin 4 ) cos 2 ) sin 2 ) ,
8
8
8
(7.144)
which is identical to Eq. (7.141), as expected. On the other hand, if the
value C A then the orientation of fibers is limited to parallel fiber bundle
and the angular utilization coefficient according to Eq. (7.143) is expressed
by the following form
C 2 sin 4 )
2C 1 cos 4 ) 6 cos 2 ) sin 2 )
'
(
C A
C
2
C 1
0a lim
C
2C 1
6C
C 2 sin 4 )
cos 4 )
cos 2 ) sin 2 )
lim '
(
2
2
2
C A
2
C 1
2
C 1
' 2
C 1
(
1
2
1
6
' 2
(
2
C
C C
4
2
2
4
C
(
cos
cos
sin
si
lim '
)
)
)
n
)
2
2
2
C A '
(
1
1
1
2 1
2 1
' 2 1
(
C
C
C
cos 4 ).
(7.145)
(The result of the last equation can be derived also directly from the
geometry of the oblique parallel fiber bundle as shown in Fig. 7.12).
The graphical interpretation of the resulting Eq. (7.143) is shown in
Fig. 7.13.
15. A more general solution considering the influence of functions :
N, N and
M
-, - b , i.e., an analytical solution of Eq. (7.136) is reported in Reference [20].
292
Parallel
0a
1
0.8
0.6
Isotropic
0.4
0.2
2
2
6
6
C 500 (C A)
32
16
8
4
2
1
38
0
0
) rad 6
2
6
2
Note: It is evident from Fig. 7.13 that the maximum value of the angular
utilization coefficient 0a is obtained if the direction vertical jaw axis
corresponds to the preferential direction of fibers, i.e. ) = 0. On the other
hand, the minimum value of the angular utilization coefficient 0a is obtained
if the direction of the vertical jaw axis is perpendicular to the preferential
direction of fibers, i.e. ) @ 2 . This phenomenon is often used during
the testing of fiber orientation of non-woven textiles. The strength of a nonwoven textile, say produced from a carded web, is measured in longitudinal
(preferential direction) and transversal directions and the ratio of both the
values is then used as an empirical characteristic of fiber orientation.
7.4
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
293
References
Dow, N., and Tranfield, G. (1969). Preliminary investigations o flexibility of
weaving triaxial fabrics, Textile Research Journal, 40, 986998.
Skelton, J. (1971). Triaxially woven fabrics: Their structures and properties,
Textile Research Journal, 41, 637647.
Sugun, B. S., and Rao, R. M. V. G. K. (2000). Mechanical behaviour of woven
and multiaxial fabric composites, Journal of Reinforced Plastics and
Composites, 19, 743753.
Pastore, C. M. (2000). Opportunities and challenges for textile reinforced
composites, Mechanics of Composite Materials, 36, 97116.
Araujo, M., Lima, M., and Costa, N. (2001). New weaving concept for multiaxial
fabrics, Proceedings of the 1 st Autex Conference, Portugal, June 2629, 2001.
Cybulska, M., and Frontczak-Wasiak, I. (2002). Modelling and properties of
multiaxial woven structures, Proceedings of the 2 nd Autex Conference, Belgium,
2002.
Wang, Y. (2002). Mechanical properties of stiched multiaxial fabric reinforced
composites from manual layup process, Applied Composite Materials, 9, 8197.
Jiao, W. H., and Chen, N. L. (2004). Performance advantages of biaxial warp
knitted fabrics used as coated substrates, Journal of Donghua University, 30,
91104.
Frontezak-Wasiak, I., and Snycerski, M. (2005). Characteristics of multi-axial
woven structures, Fibers and Textiles in Eastern Europe, 13, 2733.
Baozhong, S., Hong, H., and Bohong, G. (2005). Compressive behavior of multiaxial multi-layer warp knitted (MMWK) fabric composite at various strain rates,
Composite Structures, 78, 8490.
Robinson, P., Iannucci, L., and Oakeshott, J. L. (2007). Development of a
representative unit cell model for bi-axial NCF composites, Journal of Composite
Materials, 41, 801835.
Luo, Y., and Hong, H. (2008). Mechanical properties of PVC coated bi-axial
warp knitted fabric with and without initial cracks under multi-axial tensile
loads, Composite Structures, 89, 536542.
Truong, T. C., Ivanov, D. S., Klimshin, D. V., Lomov, S. V., and Verpoest, I.
(2008). Carbon composites based on multi-axial multi-ply stitched preforms,
Part 7: Mechanical properties and damage observations in composites with
sheared reinforcement, Composites Part A: Applied Science and Manufacturing,
39, 13801393.
Hong, H., and Ying, X. (2008). Tearing properties of coated multi-axial warp
knitted fabric, Autex Research Journal, 8, 1316.
Luo, Y., Hong, H., and Fangueiro, R. (2008). Tensile and tearing properties of
bi-axial warp knitted coated fabrics, Autex Research Journal, 8, 1720.
Luo, Y., Hong, H., and Fangueiro, R. (2008). Tensile and tearing properties of
PVC coated bi-axial warp knitted fabrics under bi-axial loads, Indian Journal
of Fiber and Textile Research, 33, 146150.
Beil, N. B., and Roberts, W. W. Jr. (2009). Modeling and computer simulation
of the compressional behavior of fiber assemblies, Part I: Energy-absorption
capability of multi-axial warp-knitted FRP tubes, Journal of Crashworthiness,14,
407418.
294
18. Tomohiko, S., Asami, N., and Hiroyuki, H. (2009). Effect of CF/GF fiber hybrid
on impact properties of multi-axial warp knitted fabric composite materials,
Composites Part A: Applied Science and Manufacturing, 40, 19821990.
19. Delisee, C., Pierre-Jacques, L., and Vautrin, A. (2010). Characterization of 3D
morphology and microcracks in composites reinforced by multi-axial multi-ply
stitched preforms, Composites Part A: Applied Science and Manufacturing,
41, 653662.
20. Neck, B. (2001). Morphology and Structural Mechanics of General Fibrous
Assemblies, Technical University of Liberec, 1st edn. 1998 and 2nd edn. 2001
(Czech).
Appendices
295
Appendix 1
Evaluation of the integral
2
50 cos - u
- d-
for fiber
orientation
2
cos - u
- d- , where the probability
1
C
1
C
cos
'
( d50
2
2
2
2
2
2
' C C 1 cos
- ) C C 1 cos
- ) (
2
2
cos
- dcos - dC
'5 2
(
5
2
2
2
2
2
' 0 C C 1 cos
- ) 0 C C 1 cos
- ) (
- , d- d
2
kn
C
'
'
2
5
0
cos d
2
2
C C 1 cos 2
)
2
5 C
C
2
cos - d(.
2
1 cos
- ) (
(A1.1)
(We use cos- = cos(-)).
The integral variable in both integrations is renamed to x, we thus obtain
C
cos x dx
' 5 2
2
' 2 C C 1 cos 2
x )
0
kn
2
5
0
cos x dx
(
C 2 C 2 1 cos 2
x ) (
2
2
cos
x ) ) dx
C
cos x dx
C
5
2
2
2
2 C C 1 cos
x ) 5 2 C 2 C 2 1 cos 2
x )
cos
x ) cos ) sin
x ) sin )
C
dx
5
2
C 2 C 2 1 cos 2
x )
2
295
296
cos
x )
C cos )
dx
5 2 C 2 C 2 1 1 sin 2
x )
2
sin
x )
C sin )
dx
5
2
2 C C 2 1 cos 2
x )
2
cos
x )
sin
x )
C cos )
C sin )
dx
dx.
5 2 1 C 2 1 sin 2
x )
5 2 C 2 C 2 1 cos 2
x )
2
2
(A1.2)
We solve the above integration by applying the rule of substitution. We
apply the following substitutions in the first integral expression
sin
x )
y
C 1
2
cos
x ) dx
dy
C2 1
(A1.3)
Cz
C 1
2
sin
x ) dx
C dz
C 2 1
(A1.4)
Thus, we find
kn
C cos )
C 2 1 sin )
2
dy
C sin )
2
y
1
C 1 C 2 1 sin
)
C2 1
2
2
C cos )
C 2 1 sin )
2
dy
sin )
2
y
1
C 1 C 2 1 sin
)
C2 1
2
2
C 2 1
cos )
2
C
C 2 1
cos )
2
C
C 2 1
cos )
2
C
C 2 1
cos )
2
C
dz
C 2 C2 z2
dz
.
1 z2
(A1.5)
kn
C cos )
C2 1
+ arctan y,
C 2 1 sin )
2
C 2 1 sin )
2
sin )
G1 z 1 6
H ln
8
2
C 1 I 2 z 1 9
C 2 1
cos )
2
C
C 2 1
cos )
2
C
Appendices
kn
297
C cos ) G
2
2
6
Harctan ' C 1 sin ) ( arctan ' C 1 sin ) ( 8
2
2
2
9
C 1 I
G
7
sin ) 7 1
H ln
C2 1 7 2
7
I
C2 1
cos ) 1
2
C
C 1
cos ) 1
C
2
2
1
ln
2
6
C2 1
cos ) 1 7
2
7
C
8.
2
C 1
cos ) 1 7
7
2
C
9
(A1.6)
By rearranging, we find
kn
C cos ) G
2
2
6
Harctan ' C 1 sin ) ( arctan ' C 1 sin ) ( 8
2
2
2
9
C 1 I
G
sin ) 77
Hln
2 C 2 1 7
I7
C 2 1 cos ) C
C 2 1 cos ) C
2
2
ln
C 2 1 cos ) C
C 2 1 cos ) C
2
2
6
77
8
7
97
2C cos )
arctan ' C 2 1 sin ) (
2
C 1
2
2
2
' C 1 cos 2 ) C ( ' C 1cos 2 ) C (
.
ln
2
2 C 2 1
2
' C 1 cos 2 ) C ( ' C 1 cos 2 ) C (
sin )
(A1.7)
(We use sin
2 sin 2, cos
2 cos 2, arctan
u arctan u ).
Furthermore,
kn
2C cos )
arctan ' C 2 1 sin ) (
2
C 1
2
2
2
' C 1 cos 2 ) C ( ' C 1 cos 2 ) C (
ln
2
2
2 C 2 1
' C 1 cos 2 ) C ( ' C 1cos 2 ) C (
sin )
2C cos )
sin )
ln
arctan ' C 2 1 sin ) (
2
2 C 2 1
C 1
2
C 2 1 cos ) C
2
C 2 1 cos ) C
2
(A1.8)
298
From the previous equation, the coefficient kn takes the following form
kn
2C cos )
sin )
arctan C 2 1 cos )
ln
2
C 1
C2 1
C 2 1sin ) C
C 2 1sin ) C
(A1.9)
(We apply cos ) sin ) and sin ) cos ) ).
2
2
In case of ) = 0, and from Eq. (A1.9), the following expression is found
kn
2C arctan
C2 1
C 1
2
(A1.10)
Moreover, if C = 1, then
kn lim
2C arctan
C 1
C2 1
C 1
2
2.
(A1.11)
kn
1
C2 1
ln
C2 1 C
C2 1 C
(A1.12)
Note: We can also obtain Eq. (A1.12) from the limit of Eq. (A1.11) as
C 1.
Finally, we introduce the integral of kn as C1 (isotropic orientation in
plane). From Eq. (A1.9), we find
G
77 2cos )
sin )
kn lim H
arctan C 2 1 cos )
ln
C A
C2 1
7 1 1
7I
C2
kn
2 cos )
sin ) 1
0 * ln
cos ).
2
sin ) 1
1
sin ) 1
C2
1
1 2 sin ) 1
C
6
7
7
8.
7
97
(A1.13)
Appendices
299
Appendix 2
Function (ex 1)n
a @ b n
@1 k
k 0
a 1 n
(A2.2)
a 1 n
(A2.3)
Note: The last expression shows the combination numbers, which are
popularly known as the so-called binomial coefficients. We may remind
that it is valid to write the following expressions
n
n
0
n!
n!
0!
k
n k ! k ! , 0! 1, 0 n ! 0! 1, 0 0!0! 1 .
(A2.4)
(A2.5)
Function (ex 1)n. We can write the following expression from Eq. (A2.1)
considering a = ex, b = 1, and the negative sign as follows
300
n
n
n
k
n
k
n
e x 1
1 e
n k x1k
1 e
n k x , n 1, 2 . (A2.6)
k
k
k 0
k 0
By applying Eqs. (A2.1) and (A2.5), Eq. (A2.6) can be written as follows
n
k
n
1 '
1
k
k 0
n
n 1
k
n
'
1
k
k 0
n 1
k
n
'
1
k
k 0
0
n
n
k
n
k
n
n k x
n k x
e
1 (
( '
1 ( '
1 e
k k 0
k
k 0
0
n
n
e
n k x 1 (
1 e
n n x 1
n
e
n k x 1 ( .
(A2.7)
According to the expansion of Taylor series, we find the following
expression
e
n k x 1
n k x
n k 2 x 2
n k 3 x3
1!
2!
3!
n k i xi
i 1
i!
. (A2.8)
n 1
A
n 1
n
n k i xi
k
n
k
n
e x 1 '
1 e
n k x 1 ( '
1
(
i!
k
k i 1
k 0
(
k 0 '
A G i n 1
6
n
k
i 7
7x
H '
1
n k ( 8, n 1, 2 .
k
i
!
i 1 I
7 k 0
79
(A2.9)
Function S ni . We apply the following summation symbol in the previous
expression
Sni
1 n 1
k
n
i
'
1
n k (, n 1, 2, , i 1, 2, .
n ! k 0
k
(A2.10)
Appendices
A
n
G xi
6
1 H n ! S ni 8 .
i
!
i 1 I
9
301
(A2.11)
(A2.12)
1 1
1
nu
n i
S '
1
u (
n! u n
n!
n u
i
n
n
1
n!
u 1
'
1
1 u
n i
u(
u (
u 1 '
n
' 1
n i
u u (.
(A2.13)
(A2.14)
Sni
1 n Qi
n!
(A2.15)
(A2.16)
n 1 !
u
u
Qni '
1
u i ( '
1
u(
n u ! u ! u 1 '
n 1 u 1!
u 1 ! u (
u 1
n
n
u
n 1 i 1
u
n 1 i 1
n '
1
u
n
u (.
(
'
1
u 1
u 1
u 1
u 1
(A2.17)
302
v 1i 1 (
(
'
1
v
v
v0
v0
n 1
u
n 1
n '
1
u 1 i 1 (, i 1, 2, , n 2,3, .
u
u 0
(A2.18)
For i = 1, Qni can be calculated directly from Eqs. (A2.18) and (A2.5)
as follows
0
n 1
n 1
n
1
n
1
u
u
1
1
Qn1"1 n '
1
u 1 ( n '
1 ( 0, n 2,3, .
u
u
u 0
u 0
(A2.19)
For i = 2, 3, and n = 2, 3, we rearrange Eq. (A2.18) by applying
Eq. (A2.3)
n 1
6
i 1 i 2
i 1
u
n 1 G
i 1 i 1
Qni n '
1
u
u
u 18(
H
u I 0
1
i 2
u 0
'
97(
i 1 n 1
i 1 n 1
u
n 1 i 1
u
n 1 i 2
u
n
u (
n
1
'
(
'
1
0 u 0
u
1 u 0
u
n 1
i 1 n 1
u
n 1
u
n 1
n
u
n
1
'
( '
1
(.
u
i 2 u 0
u u 0
(A2.20)
i 1 n 1
i 1 n 1
u
n 1 i 1
u
n 1 i 2
Qni n
u ( n
u (
'
1
'
1
0 u 1
u
1 u 1
u
i 1 n 1
u
n 1
n
u( n *0
'
1
i 2 u 1
u
i 1 i 1
i 1 i 2
i 1 1
n
Qn 1 n
Qn 1 n
Qn 1 , i 2, 3, , n 2,3, .
0
1
i 2
(A2.21)
Q by one unit.
Appendices
303
(A2.22)
6
7
7
0
7
1
1
n
n
7
n 1
n2
Qnn n
Q
n
Q
Qnn
1 n Qnn11 , i n 2,3, .
8
n
n
1
1
1
0
7
0
7
n 1
1
7
Q
i
n
,
3,
4,
,
n
n 1
n 2
79
(A2.23)
1
Q22 2 Q11 , (i n 2),
0
n 1
n!
1
n
n
n 1
2 Q11
1 n.!
(A2.24)
It is valid Q11 1 from the Eq. (A2.16) by i = 1 so that we can extend
the previous equation as follows
Qnn
1 n !, n 1, 2, .
n
(A2.25)
304
Q i 0,
0
i
n "i
n
n
i
n 1
i 1, 2, , n i 1, i 2, .
n "i
n!
n
n
1 n
1
Q
n!
n!
1
1 i
Q 1,
1 n n! 1,
n 1, 2, .
(A2.26)
(A2.27)
1
1!
i 1, 2, , n 1 .
(A2.28)
Function B inn1 and its properties. We define B inn1 in the following manner
for n = 1, 2,, , i = n, n + 1,
6
7
in
8 i n, n 1, .
Binn1 nt Bin n t , n 2,3, , 7
t 0
9
Binn1 1, n 1
(A2.29)
B0n 1 1, n 1 and
(A2.31)
(A2.32)
Appendices
305
in
(A2.33)
n0B
in11
1n 1 n1B
in21
1n 1 ni n B
nn 11 1
n 1 .
(A2.34)
By applying Eq. (A2.15) and then Eq. (A2.14) into the last equation,
we find
B
in
n 1
n 1
n 1
1
1 7G 0 n 1
u
n 1 i 1
i n n 1
0 i 1
1 i2
n Qn 1 n Qn1 n Qn 1 ,
u (
+
H n '
1
n
n
u
1
!
1
!
7I u 1
n 1
n 1
u
n 1 i 2
u
n 1 n 1 6
7
n1 '
1
u ( ni n '
1
u ( 8
u
u
u 1
u 1
97
n 1 n 1
n
1
i 1
1
u
i2
i n n 1
'
1
u nu n u ( .
u
n 1 !
u 1
(A2.35)
u i 1 nu i 2 n i nu n 1 u i 1
n u i n1 1 u i 1
ni n1 u i n1 u
n u 1
u i n 1
n u
1 n1 i n1 i n1
ni
u n 1 u i
u
n
u
.
nu
n u n n 1 ni
(A2.36)
n u
ui (
u
n 1
in
B n 1
(
'
1
u n u n n 1 ni
n 1 !
u 1
n 1 n 1
1 n1 n 1 1 u 1
n 1 u i
1 n 1 n1 1 u 1
n 1 ni n1 u n 1 .
'
(
'
u n u (
n 1 !
n 1 ! u 1
u n u
u 1
A
(A2.37)
306
1 n1 n1 1 u 1
n 1 ! u i
1 n1 n1 1 u 1 n ! u i
'
'
n u ! u ! (
n 1 !
n 1 u ! u ! n u (
n 1 ! n
u 1
u 1
Sn
n n 1
n n
n
1
n
n
1
1
ni
u
u
n
n i
i
i
i
1
1
1
u
u
n
S
.
'
' u ( n !
n
u (
n
n ! u 1
n!
n!
u 1
i
Qni
(A2.38)
Now, we rearrange the second element B of the right-hand side of Eq.
(A2.37). Applying Eqs. (A2.14) and (A2.22), we find
1 n1 ni n1 1 u 1
n 1 ! n
u n 1 (
n 1 '
n 1 ! n n u 1
n 1 u ! u !
n u
n
ni
1 n 1
u
n
n 1
'
1 u n1 (
B
n!
u 1
u
Qnn 1 0
ni
1
n 1
n
n!
n
n
n
ni
1
ni
u
n
n n 1
n 1
1
u
1
n
.
'
(
n 1
u
n
n!
n!
u 1
n
(A2.39)
By applying Eqs. (A2.38) and (A2.39) into Eq. (A2.37), the following
expression is obtained
ni
n i
Binn1 Sni S ni .
n ! n !
(A2.40)
(A2.41)
Appendices
307
Appendix 3
Useful terms and quantities in mechanics
308
; F s .
(A3.1)
N /l l ,
(A3.2)
309
Appendices
Table A3.1 Commonly used terms.
Before rupture
State Type
of behavior
Term of quantity
Elongation
Strain
(+)
At rupture
Deformation
Physical
dimensions
Examples of
symbols
Chapter 1
Chapter 5
F
*
*
p, *
;
[mm]
[1]
/l, %
N
[N]
[Pa]
[N]
[Pa]
[N tex1]
[Pa]
Chapter 6
F*, P,P
(Engineering) Specific
Strength(+), or Tenacity(+)
[N tex ]
;
*
Breaking Elongation
Breaking Strain(+),
eventually Extension(+)
[mm]
[1]
/l *
A
(+) It is used also with suitable adjective, e.g. tensile, compression, normal, tangential,
etc.
References
1.
2.
3.
310
Appendix 4
Derivative of deformation energy dE
:, B d
dE
:, B k FV
=
d
q d 2
6
d G7
1 h0 : 0
7
f
H
28
5
d
2
7I
h0 : 0 : Bd 97
2
1 d
(A4.1)
F
1 h0 : 0
h0 : 0 : Bd
(A4.2)
dF
OF
d: OF
dB
.
d
O: d
OB d
(A4.3)
6
OF
O 7G
2
h0 : 0
1 h0 : 0
7
1 h0 : 0
H
28
3 .
O:
O: 7
h0 : 0 : Bd 7
h0 : 0 : Bd
9
I
(A4.4)
From Eq. (5.55) we obtain
d: d '
d
8
J
d d '
1 e
( d
1 (
8
J
8
( ' J
2
8
J
1
1 e
1 'e
(
311
Appendices
d
8
d
J
1 e
2
e
2
8
J
8
J
8
J
d
: 2 1
d
.
:
(A4.5)
2
h0 : 0 8 J
d
1 h0 : 0
1 : 2
1
3
O: d
:
h0 : 0 : Bd d
16 J
d
1 h0 : 0
h0 : 0
d
1
2
3
h0 : 0
h : B d : :
' 0 0
: (
3
d
1
16 J
1
1 h0 : 0
d
:
.
2
3
2
d
h0 : 0
1 B d : 0 :
: h0
The following expression is evident from Eq. (A4.2)
(A4.6)
6
OF
O G7
2
d
1 h0 : 0
7
1 h0 : 0
H
28
3
OB
OB 7
h0 : 0 : Bd 7
h0 : 0 : Bd
9
I
d
1 h0 : 0 d
d
2 1 h0 : 0 1
:
2
2
,
3
2
3
d
h0 : 0 : h0
d : 0
d : 0 :
3
3
:
h0 : 0 1 B
d : 0 1 B : h
: h0
0
(A4.7)
so that
OF
dB
2
OB d
1 h0
:
h0 : 0
3
: 0 d
d : 0
1 B : h
0
2 1 h0 : 0 1
2
d
h0 : 0 2 : h0
d :0
d
d dB .
:
3
d : 0 : d
1
B
: h0
(A4.8)
312
dF
16 J
d
d 2
d
1 h0 : 0 1 :
d 2 1 h0 : 0
h0 : 0 2
1 B d : 0 3 : d 2
h0 : 0 2
: h0
8J
2
1 h0 : 0 d
1
3 '
d 2
h0 : 0 2 :
d : 0 '
1 B : h '
0
16 J
d 2
1 d
1 h0 : 0 d
h0 : 0 2 :
: h0 '
d
1
8J
:
1 dB
: h0 d
d :0
d : 0
1 B : h
0
d
d dB
:
3
d : 0 : d
1 B : h
0
1 d dB
: h0 : d (
(
d :0
(
:
1 h0 : 0 d
1
d
1
2
3 '
h0 : 0 :
1 B d : 0 ' : 8J
1
1
16 J
d 2
1
: h0
d :0
1 d dB
: h0 : d (
(
d :0
(
(A4.9)
Derivative of deformation energy. Applying Eqs. (A4.2) and (A4.9) into
Eq. (A4.1), we can write
dE
:, B k FV
d
q d 2
dF
f
d
2
2
G
7
2 7
716 J
k V
F 2 5 H
qd 2 7 d 2
7
I7
G
7
2 7
7
dE
:, B 16k FV
2 4 5 HJ
d
q d 2 7
7
7I
1 d
1
1
1 h0 : 0 d
h0 : 0 2 :
1
1
1 h0 : 0 d
h0 : 0 2 :
d
1
: 8J
1 dB 6
h0 d 7
:
1
77
d :0
8f
3
7
d : 0
7
1 B : h
0
97
d
1
: 8J
1 dB 6
: h0 d 7
77
d :0
8f
3
7
d : 0
7
1 B : h
79
0
1
1 d
1 d
(A4.10)
313
Appendices
Appendix 5
Angular utilization coefficient of special alternative
to simple planar bundle
0a
1
5 '' C
C
2
2
C
C
1
( cos 4 - d2
2
2
2
1 cos
- ) C C 1 cos
- ) (
C cos 4 - d1
2
C C 2 1 cos 2
- )
2
5
0
C cos 4 - d1
2
C C 2 1 cos 2
- )
(Substitution: x - ), dx d- ) (Substitution: y - ), dy d- )
) 2
) 2
5
5
)
1 C cos
x ) dx
C 2 C 2 1 cos 2 x
) 2
) 2
) 2
4
1 C cos
x ) dx
C 2 C 2 1 cos 2 x
5
)
4
1 C cos
y )
dy
C 2 C 2 1 cos 2
y
1 C cos
y ) dy
C 2 C 2 1 cos2 y
4
1 C cos
x ) dx
1 C cos
y ) dy
.
5
C 2 C 2 1 cos 2 x )
C 2 C 2 1 cos 2 y
2
)
(A5.1)
We rename the integrating variables in both the integrals by the
symbol . Thus
) 2
0a
5
)
4
1 C cos
) d
C 2 C 2 1 cos2
4
1 C cos
) d
5 C 2 C 2 1 cos2
) 2
) 2
4
1 C cos
) d
5 C 2 C 2 1 cos 2
) 2
)
(A5.2)
314
4
1 C cos
) d
5 C 2 C 2 1 cos 2
2
) 2
0a
2
. (A5.3)
cos4
)
cos
sin ) cos 4
4
cos ) sin
4 cos3
cos3 ) sin
6 cos
cos ) sin
4 cos
cos4 ) cos 4
sin )
cos ) sin 3
sin 2 )
sin 3 ) sin 4
sin
sin
sin 2
cos 4 )
sin 4 )
(A5.4)
).
C
cos 4 d
0a cos 4 ) 5
2
2
2
2 C C 1 cos
2
C
cos3 sin d
4 cos3 ) sin ) 5
2
2
2
2 C C 1 cos
2
C
cos 2 sin 2 d
6 cos 2 ) sin 2 ) 5
2
2
2
2 C C 1 cos
2
cos sin 3 d
C
4 cos ) sin 3 ) 5
2
2
2
2 C C 1 cos
(A5.5)
2
sin 4 d
C 4
sin ) 5
2
2
2
2 C C 1 cos
C 2 C 2 1 cos 2
C 2 1 sin 2
1 C 2 sin 2
cos 2
(A5.6)
315
Appendices
2
2
f
2 f
2 .
I1 5
d
C C 1 cos 2
5
d
2
C sin
cos
5
1
C tan 2
2
d
1 cos 2
1
lim ' arctan
C tan
2 C
.
2C 2C C
I1 2 2
I2 5
cos 2 d
C 2 C 2 1 cos 2
(
1
lim ' arctan
C tan
2 C
(
(A5.7b)
5
cos 2 d
C 2 sin 2 cos 2
C
1 sin cos
C 1 cos d
5
5
d
C 1
C sin cos
C 1
C sin cos
C
C sin
cos
5
d
C 1
C sin cos
2
d
C2
2
2
2
5
C 1 C sin
cos 2
1
C2
d
I1 2 .
C2 1 5
C2 1
C 1
(A5.8a)
316
I 2 2
2
2
'
( ' (
C2
C2
2
2
2 I1 2 ' 2 (
2 I1 2 ' 22 ( ' 2 2 (
C 1
C 1 C 1
C 1 2 C 1
C2
C2
C 1
2
I
.
2
1
C 2 1 2 C 2 1 C 2 1 C C 2 1
C 1 C 1
(A5.8b)
cos sin d
I3 5 2
C C 2 1 cos 2
1
cos sin d
1 1 dt
1
ln t
2
2
5
5
C 1 C
C 1 2 t 2 C2 1
2
cos
C2 1
C2
dt
cos 2 t , cos sin d .)
(Substitution: 2
C 1
2
2
C
1
cos 2 .
ln 2
2
C
1
2 C 1
(A5.9a) 2
2
I
2
3 2
C2
1
'
cos 2 (
ln
2
2
(
' 2 C 1 C 1
2
1
2 C2 1
C2
C2
2
cos 2
ln
cos
ln
' 2
(
2
2
2
C 1
C 1
1
2 C2 1
C2
C2
ln
ln
' 2
C 2 1 ( 0.
C 1
(A5.9b)
I4 5
cos3 sin d
C 2 C 2 1 cos 2
1
cos sin
C2 1 5 C2
cos 2
C2 1
1
cos sin
cos 2 d
5
C 1 C2
2
cos
C2 1
C2
C2
cos 2 ( d
' 2
2
C 1 C 1
2
2. Because C > 1 it must be valid that t > 0 so that a symbol for an absolute value is
not necessary.
317
Appendices
cos sin d
C2
2
5
2
C 1 C C 2 1 cos 2
1
C2
I3
2
C 1
2 C2 1
1
cos sin
C 1 5
2
5
2 cos
sin
cos 2
C2
I
.
3
C2 1
2 C2 1
(A5.10a)
I 4 2
2
2
cos 2
C2
2
2 I 3 2 '
2
C 1
' 2 C 1
(
( 2
2
' cos 2
C2
2
2 I 3 2 '
2
C 1
' 2 C 1
2
C
2 0 0.
C 1
2
( ' cos 2
('
2
( ' 2 C 1
(
C2
2
( 2 I 3 2
1
C
(
(A5.10b)
cos sin 3 d
I5 5 2
C C 2 1 cos 2
5
sin cos3 d
C 2 C 2 1 cos2
I3 I 4 .
I 3 2 I 4 2 0 0 0.
2
5 2
2
2
cos 2 sin 2 d
C 2 C 2 1 cos 2
cos
1
2
5
C 1
cos
1
2
5
C 1
cos
1
2
5
C 1
sin cos d
C 2 C 2 1 cos 2
I
I6 5
sin 2
C sin
sin 2
cos
cos 2
cos2
cos
(A5.12)
1
cos 2
2
5
C 1 C sin 2 cos 2
2
1 cos 2
sin 2
C 2 sin 2
sin
(A5.11b)
C 2 1 cos 2 sin 2
1
C 2 1 5 C 2 sin 2 cos 2
cos
C sin
1
cos 2
C 1 5
2
(A5.11a)
d
d .
318
5 cos
d sin
5 sin 2
cos
d sin
cos
5 1 cos 2
25 cos 2
sin
cos
5 d 5 cos 2
sin
cos
d sin
cos
,
5 cos 2
d
d ,
5 cos
d
sin
cos
2
(A5.13)
I
2
6 2
sin cos
1
2
I2
2
2
2 C 1
2 C 1 C 1 .
sin cos
'
2
2 C2 1
' 2 C 1
(A5.14a)
2
1
2
2 I 2 2
(
1
C
( 2
sin
cos
' sin 2 cos 2
(
'
(
2
2
2
2
'
('
(
2
2
2
2
2 C 1 ( ' 2 C 1
2 C 1 (
' 2 C 1
1
1
2
2
2 I 2 2
2 I 2 2
C 1
2 C 2 1 C 1
1
2
2
2 C 1 C 1 C 1 2 C2 1
C 1 .
2
2
2 C 1
C 1 2
C 1
2
'1 C 1 (
(A5.14b)
319
Appendices
Further
I7 5
5
5
I
2
7 2
cos 4 d
C 2 C 2 1 cos 2
cos 2
cos 4 d
C 2 sin 2 cos 2
5
1 sin d
2
(A5.15a)
cos 2
C 2 sin 2
cos2 d
C 2 sin 2 cos 2
I 2 2 I 6 2
2
2
cos 2 sin 2 d
C 2 sin 2 cos 2
I 2 I6 .
2
C 1 1
2C 1
.
2
2
2
C 1 2
C 1
2
C 1
2
C 1
(A5.15b)
5
d
C 2 sin 2 cos 2
I1 2 I 2 I 7 .
5
1 cos
2
sin 4 d
I8 5 2
C C 2 1 cos2
25
C sin
cos
5
cos
d
C 2 sin 2 cos 2
5
1 2 cos
2
C sin
cos 4
2
2
cos
cos
d
C 2 sin 2 cos 2
(A5.16a)
I
2
8 2
I1 2 2 I 2 2 I 7 2
2
2
2
2C 1
2
C
C 1 2
C 1 2
2
C 1 4C
C 1 C
2C 1
2
2C
C 1
2C 2 4C 2 4C 2 4C 2C 2 C
2C
C 1
C 2
2C
C 1
.
(A5.16b)
0a
C
C
C
2
2
2
cos 4 ) I 7 2 4 cos3 ) sin ) I 4 2 6 cos2 ) sin 2 ) I 6 2
C
C
2
2
4 cos ) sin 3 ) I 5 2 sin 4 ) I 8 2
320
2C 1 C
C
C
cos 4 )
4 cos 3 ) sin ) * 0 6 cos 2 ) sin 2 )
2
2
2
C 1
2
C 1
C 2
C
C
4 cos ) sin 3 ) * 0 sin 4 )
2
2C
C 1
C 2
2C 1 C
C
C
cos 4 )
sin 4 )
6 cos 2 ) sin 2 )
2
2
2
2
C 1
2
C 1
2C
C 1
C 2 4
4
2
2
'
2C 1 cos ) 6 cos ) sin ) C sin ) ( .
2
C 1
C
(A5.17)
Index
Contact place, 97
binomial distribution of, 108
modified parameters of, 130
number of, 102
relative size of, 129
Continuum mechanics, 307
Conventional pores, 34, 35
Cotton fibers, fineness of, 3
Crimped fiber, linear density of, 83
Deformation energy, 140, 142, 155, 156,
174, 310
Denier, 3, 10
Direct system, 3
Distribution function, 236
Engineering (nominal) stress, 308
Engineering/nominal (fictive) stresses, 173,
177, 180
Equivalent fiber diameter, 4
Fiber, 1
bending deformation of, 139
characteristics of, 1
evaluation of, 7
non-oriented angles of, 66
perimeter of, 6
properties of, 1
scheme of, 1f
shape factor of, 6, 8
volume of, 4, 5
Fiber arrangement, 136
Fiber aspect ratio, 6
Fiber blend, characteristics of, 1927
Fiber break, 9
Fiber breaking forces, 202
Fiber breaking strain, 247
Fiber compactness, 10
322
Index
Joint probability density function, 102, 103,
111
Laminar flow, 4143
Length utilization coefficient, 243, 245
Limit state, 149, 157
Limit structure, 15
Lindslays apparatus, 76, 77f
Lindslays methodology, 76, 77, 78
Linear density, 2. See also Titre.
Linear textiles, 284
Local packing density, 13
Loose structure, 17
Mass fraction, 19, 25, 26, 285
expression of, 19, 26
Mean equivalent fiber diameter, 21, 22
expression of, 22
Mean fiber aspect ratio, 23
expression of, 23
Mean fiber cross-sectional area, 21
expression of, 21
Mean fiber density, 19, 20
expression of, 20
Mean fiber fineness, 20, 21
expression of, 21
Mean specific surface area, 24
expression of, 24
Mean specific surface area per unit volume,
24
expression of, 24
Mechanical (engineering) stress 10
Mechanical isotropy, 284
Mechanical stress, 9
SI unit of, 9
Micronaire values, 3
Microscopy technique, 7
Modeling concept, 51
Modeling scheme, 78, 79f, 80f, 81f, 82
Modified probability, 116
Modified van Wyks model, 147
application of, 180
Modularly isotropic, 283
Multi-axial textile, 249, 262, 285
mechanical behavior of, 250
specific strength (tenacity) of, 269
Multi-component parallel fiber bundle, 201,
246
Neglecting margins, 278f
323
324