You are on page 1of 11

Int J Fract (2014) 187:187197

DOI 10.1007/s10704-013-9924-2

BRIEF NOTE

On characteristic lengths used in notch fracture mechanics


G. Pluvinage J. Capelle

Received: 20 November 2013 / Accepted: 17 December 2013 / Published online: 29 January 2014
Springer Science+Business Media Dordrecht 2014

Abstract In this paper, four kinds of characteristic volume to the defect promoting fracture. The majority
length parameters used in a local notch fracture crite- are based on two local parameters: the first in terms
rion are presented: (1) a characteristic length c gener- of fracture stress, strain or strain energy density, and
ally connecting to the notch radius, (2) a characteris- the second as a characteristic length parameter. New
tic distance Xc considered as intrinsic to material and trends incorporate a third parameter like constraint for
connected to the microstructure, (3) a critical distance example (Meliani et al. 2011).
d0 considered as intrinsic to material and connected to Local stress fracture criterion was first suggested by
the fracture process zone, (4) an effective distance Xef Neuber (1937), but later there appeared local fracture
considered as a characteristic of the stress distribution. criteria based on strain energy density or local fracture
Each approach is discussed. The paper ends with the strain in addition to local stress.
authors opinion about the different methods. Guillemot (1965) postulated that the strain energy
density for fracture Wc in a local volume is a material
Keywords Characteristic length Effective distance constant, which can be evaluated as the area of the stress
Notch radius Notch plastic zone Constraint Loading strain curve for a smooth specimen until fracture. This
mode local volume is a small cylinder of diameter Xc and
height l0 according to the concept of micro-sample. The
micro-sample length l0 is related to the critical strain
1 Introduction energy release rate Gc by the following relationship:
Gc
Fracture criteria are divided into two families: global l0 = (1)
and local. Global criteria are related to external loads Wc
or work done, while local criteria are related to stress, Later Sih (1974) proposed a similar criterion.
strain or strain energy density distribution at defect Assuming that the stress distribution at crack tip is
tip (crack, crack like defects or notch like defects). given by Irwins solution Irwin (1948), he computes
Local fracture criteria are based on an accurate descrip- the so-called strain energy density factor S,
 which
 is a
tion of stress, strain or strain energy density in a close function of stress intensity factors S = f K i2 , where
Ki are stress intensity factors of one of the three frac-
ture modes (i = 1, 2, 3). Notably, this strain energy
G. Pluvinage (B) J. Capelle
LaBPS - Ecole Nationale dIngnieurs de Metz, 1 route
density factor is independent of r distance and partic-
dArs Laquenexy, 57078 Metz, France ularly characteristic distance Xc . Therefore, the local
e-mail: pluvina@univ-metz.fr fracture criteria needs only one parameter, the critical

123
188 G. Pluvinage, J. Capelle

strain energy density factor Sc . The same approach was lf at a critical distance equal to local strain fracture in
also used by Irwin in order to avoid the problem of the plane strain.
critical distance by introducing the concept of stress The concept of micro sample at the notch tip was
intensity factor. used in order to connect critical COD c and local frac-
In this case, the product of the stress and the square ture strain Pluvinage and Montariol (1971) by the fol-
root of the corresponding distance is a constant inde- lowing relationship:
pendent of distance. However, to transform a local two
c = l0 lf (5)
parameters fracture criterion into a single parameter is
only possible if the defect tip distribution is ruled by a The micro sample length is often taken as l0 = 2.
single parameter. Effective stress based on average stress over a char-
More recently, Lazzarin and Berto (2005) proposed acteristic length was earlier considered by Neuber
the averaged strain energy density criterion (SED), (1937) and related to the notch radius. The failure crite-
which states that brittle failure occurs when the aver- rion, proposed by Novozhilov (1969) and expanded by
age value of the strain energy density W over a control Seweryn (1994), suggests considering the mean normal
volume (which becomes an area in two dimensional stress along the anticipated path of the failure. Thus,
cases) is equal to a critical value. failure occurs when the average stress equals a mater-
They distinguish two cases: a sharp and a blunt ial dependent value, denoted by c , which is the failure
notch. stress. The average stress is considered along the effec-
For a sharp notch loaded in mode I where the notch tive distance Xef , the critical distance d0 or the char-
radius is zero, W , is a function of the notch stress inten- acteristic distance Xc , according to the authors. This
sity factor K . method is sometimes called the line method (Taylor
  2004).
I1 K 2 The fracture criterion (when using effective distance
W = . (2)
4E1 ( ) do11 Xef ) can be written in the following generalised form
where E is the Young modulus, 1 the stress singularity including mixed mode of fracture:
exponent, I1 a geometrical function, half of the notch 1 Xe f
angle and d0 the critical distance related to material max (r ) dr = c (6)
Xef < 0
characteristic.
For a blunt notch W is a function of the maximum where c is the failure stress, the circumferential
stress max . stress, r and polar coordinates.
  Local fracture criterion based on a characteristic
d0 2 stress corresponding to a characteristic length on stress
W = F (2) H 2, . max (3)
E distribution was introduced by Peterson (1959). Whit-
H and F two geometrical functions. Strain energy den- ney and Nuismer (1974) proposed this criterion with
sity for the notch is obtained through the stress distrib- the following form:
 
ution solution proposed by Filippi et al. (2002), where
the origin is at distance ro from the notch tip: max min (r ) = c (7)
< 0r X e f
( 2) where c is another characteristic stress. This appro-
r0 = (4)
(2 2) ach is called the point method (PM) (Taylor 2004).
is the notch radius. We note that when 2 = 0 (U Pluvinage (2001) proposed averaging the stress dis-
notch) r0 = /2, which is the value introduced in the tribution over the entire process volume Vef . Then, the
Creager and Paris solution (Creager and Paris 1967). fracture criterion has the following form:

In two dimensional cases, the fracture process volume 1
yy (x, y, z) . (x, y, z) d V = c (8)
is the part of a circle area centred at distance r0 to the Ve f
notch tip and outside of the notch.

Osborne and Embury (1973) considering that strain where (x) and yy (x) are relative stress gradient and
distribution at the notch tip exhibits a 1/r singularity, opening stress or maximum principal stress along the
assumed that fracture occurs when local fracture strain notch tip, respectively, Vef the fracture process volume

123
On characteristic lengths used in notch fracture mechanics 189

and a weight function. This method is called volu- The intrinsic character of the local fracture stress is
metric method (VM). also an open question. It is not the object of the paper,
We note that according to different approaches, four but Fig. 1 indicates the dependence of critical stress and
kinds of characteristic length parameters can be used constraint through the notch radius of axisymmetric
in a local fracture criterion: tensile specimens made from steel. These results have
been obtained by Kaechele and Tetelman (1969), who
a characteristic length c generally connecting to consider that critical stress can be identified with the
the notch radius, maximum of the notch tip stress distribution.
a characteristic distance Xc considered as intrinsic
to material and connected to the microstructure,
a critical distance d0 considered as intrinsic to mate-
rial and connected to the fracture process zone, 2 Characteristic length connecting to notch radius
an effective distance Xef considered as a character-
istic of the stress distribution. It is obvious that the fracture process zone is the high
stressed area. Hence, the question is to know the limit
These characteristic length parameters play exactly of this zone. The high stressed area incorporates the
the same role in point, line or area methods, and the maximum stress for physical reason mainly connected
choice of this characteristic length parameter differs to the idea of a high stressed area.
according to different models. In addition, in each When we consider the distribution at the notch tip,
model, formulae and definition of the length parameter the position of the maximum stress rmax depends on
can be different. the stress strain behaviour: the maximum stress is pre-
The objective of this paper is to examine these dif- cisely at the notch tip for an elastic behaviour, at a cer-
ferent length parameters and to show their advantages tain distance behind the notch for an elastic plastic one.
and their limitations. It can be considered a contribution Creager and Paris (1967) postulated that this maximum
given according to research works and opinion of the is constant over a distance r /2. The stress distribu-
authors which are more sensitive to the use of effective tion is governed by the (crack) stress intensity factor at
distance. However, an important part of research in this distance r > /2. The associated characteristic length
field is based on characteristic distance. is therefore c = /2. Experimental results obtained
More precisely, this paper examines the relation- by Saghafi et al. (2010) indicate that the relationship
ships between characteristic length and the notch between effective distance and half of the notch radius
radius, grain size, plastic zone and the minimum of is satisfied for the brittle material PMMA at 60 C
the relative stress gradient. (Fig. 2).
If we consider an elastic perfectly plastic behaviour
and apply the slip lines theory, the distance of maximum
stress from notch tip is given by this equation
    
rmax 2
ln 1 + = 1+ (9)
2 2
For U notch with parallel sides, distance rmax is equal to
3.81 with the notch radius and 2 the notch angle.
In this case, the characteristic length is several times
the notch radius according to the slip lines model.
For elastic-plastic material, finite element results
provide the relationship between the notch radius and
the effective distance given by VM (Pluvinage 2001).
This relationship deviates from the Creager and Paris
relationship Xef = /2 and is replaced by the follow-
Fig. 1 Influence of the notch radius on critical stress of axisym- ing:
metric tensile specimens made from steel (Kaechele and Tetel-
man 1969) X e f =C + B (10)

123
190 G. Pluvinage, J. Capelle

the distance equal to half of the notch radius is less than


the distance where the maximum stress occurs. In this
case, it does not take into account the most stressed
region at the crack tip and lose its physical meaning.
Finally, in this category one considers only the charac-
teristic length equal to half the notch radius.

3 Characteristic distance related to material


properties

3.1 Characteristic distance connected to


microstructural unit

Fig. 2 Effective distance versus the notch radius for PMMA at For very brittle material, the fracture process zone
60 C (Saghafi 2010)
is related to the microstructural unit. A model of
cleavage triggering indicates that dislocations pile-up
on grain boundaries explain the mechanism of brittle
fracture (ZenerStroh). Other mechanisms (Almond
Smith) associate cleavage to carbides size and inter
inclusions distance.
Characteristic distance connected to grain size is the
basis of the Ritchie et al. (1973) local stress fracture
criterion (RKR). This criterion belongs to PM as we
can see in Fig. 4. The characteristic distance Xc is equal
to grain size dg :Xc = dg (or: Xc = 2dg Ritchie et al.
1973).
This criterion is devoted not to the notch but to the
crack and is used for very brittle material where the
characteristic length is very small.
Fig. 3 Relationship between the effective distance and notch A good correlation between characteristic distance
root radius for Al 7175-T74 (Kim et al. 2004)
and cleavage initiation site was demonstrated by

where C and B are constants. Experimental results from


Akourri et al. (2005) on ferritic steel XC38, Kim et al.
(2004) on aluminum alloy Al 7175-T74 (Fig. 3) and
austenitic steel 17-4 PH and El Minor et al. (2003) on
high strength steel 45SCD5 confirm that the effective
distance Xef increases linearly with the notch radius.
According to these experimental results, the character-
istic length is of the same order of magnitude as the
notch radius. The Creagers value characteristic length
equal to half the notch radius seems a very good approx-
imation for very brittle materials. For ductile material,
Akourri et al. (2005) have found a linear relationship
between effective distance and effective stress and the
notch fracture criterion can be reduced to one parame-
ter. This relationship is not general and for more ductile Fig. 4 Description of the Ritchie, Knott and Rice local stress
material, it loses its validity. For very ductile material, fracture criterion (Ritchie et al. 1973)

123
On characteristic lengths used in notch fracture mechanics 191

Fig. 5 a Origin of cleavage


determined by the centre of
the river pattern. b Fracture
appearance of specimen. FC
material

Pluvinage et al. (1999) on a CrMoV rotor steel. Two


types of microstructures were selected: ferrite with
fine carbides (FC) and ferrite with coarse carbides.
For the FC microstructure, no distinct initiation sites
were found on one cleavage facet; the fracture sur-
face exhibits macroscopic river patterns. This pattern
originates from a small spot on the fracture surface,
as seen in Fig. 5a, which contains one or a few cleav-
age facets. For microstructure with coarse carbides, the
river patterns are traced to the initiation region. In both
cases, the centre of the river pattern was assumed to
coincide with the origin of cleavage initiation. Dis-
location pile-up micro mechanisms of cleavage were Fig. 6 Relationship between effective distance Xef and charac-
assumed. teristic distance XC (CrMoV rotor steel)
Figure 6 indicates that the agreement between effec-
tive distance Xef obtained from VM and the lowest dis-
Eq. (11) is modified in order to provide the critical dis-
tance for cleavage initiation XC,min is good.
tance d0
 
KIc 2
d0 = (12)
3.2 Critical distance connected to fracture process c
zone KIc is the plane strain material fracture toughness,
whereas the inherent material fracture strength under
We consider for physical reasons that the fracture static loading. c is upper but close to ultimate strength
process zone is associated with the plastic or dam- and is sometimes replaced by the ultimate strength u .
aged zone. Crack critical plastic zone Ry is related to is a parameter which is different from that found by
two material characteristics: fracture toughness KIc and other authors.
yield stress y . In the literature several values of can be found
  and they are all relative to brittle material with elastic
KIc 2
Ry = (11) behaviour until fracture. Some are function of notch
y
angle and type of notch (U or V).
is a parameter, which depends on the state of stress Seweryn (1994) using the Novozhilov (1969) local
(plane stress or plane strain), and is only valid for an failure criterion proposed using the value of = 2.
elastic perfectly plastic behaviour. For a strain harden- Following this idea, Taylor (2004), in the theory of
ing material and to take into account constraint effect, critical distance, suggested using a definition of the crit-

123
192 G. Pluvinage, J. Capelle

ical distance close to the value of the plane stress plas-


tic zone size and = 1.
Yosibash et al. (2004) postulated that for fracture
SED, W* is independent of the notch opening angle
2 then for 2 = 0 (a crack) and for 2 = (a flat
specimen). If this assumption holds for a specific inte-
gration radius then the characteristic distance is given
by:
 
(1 + ) (5 8 ) KIc 2
d0 = in plane strain (13)
4 c
 
(5 3 ) KIc 2
d0 = in plane stress (14) Fig. 7 Evolution of A constant versus elastic stress concentra-
4 c tion. Low strength steel (Pluvinage 2003). Comparison of A con-
and = 0, 845 ( Poissons ratio). Tovo and Livieri stant (Eq. 16) with (Eq. 12)
(2008) proposed that constant is equal to 0.545, and
= 3.14 for Gmez and Elices (2004). Leguillon function describes A evolution. The notch plastic zone
(2001) proposed a criterion for failure initiation at a has been computed by FE method using the Von Mises
sharp V-notch based on a combination of the Griffith plasticity criterion and increasing notch fracture tough-
energy criterion for a crack, and the strength criterion ness with notch radius having been taken into account.
for a straight edge. This approach is based on the change The material is a low strength steel. In the same fig-
of the potential energy in a notched specimen due to ure, the values proposed by Gmez and Elices (2004)
the creation of a small crack in the direction 0 , which ( = 3.14), Seweryn (1994) ( = 2), Taylor (2004)
generates maximum change in potential energy. Here, ( = 1) and Yosibash et al. (2004) ( = 0.845) or Tovo
as well, a critical distance is introduced, which is the and Livieri (2008) = 0.545) are also reported. The A
length of the virtual created crack. constant for a crack, trends to value ( = 1). Two values
  ( = 0.845 or = 0.545) are lower than the asymptotic
( (0 ))2 K I c 2 value and the ratio of the two extreme values ( = 3.14
d0 = . (15)
K (2 ) c and = 0.545) is about 6.
K(2) is a parameter depending on notch angle 2. For very brittle materials, the characteristic distance
In Leguillons approach, the characteristic distance is physically related to the grain size (but not neces-
depends on the notch angle, but its value is close to sarily one, two and sometimes 10 grain size) and con-
the Yoshibash solution. sequently intrinsic to material. Other materials, such
Previous values of critical distance are used for pre- as composites, polymers, wood, etc., are not formed by
diction of the fracture of brittle materials, where frac- conventional grains. Therefore, the characteristic dis-
ture stress is close to yield stress. Then, a compari- tance value is related to other internal microstructural
son between notch plastic zone and critical distance barriers.
can be made. It has been shown (Pluvinage 2003) that For less brittle material, the grain size is a distance
the notch plastic zone depends on the notch radius. It too small to be greater than the maximum stress dis-
decreases when the notch radius decreases. tance. Its use as characteristic distance leads to a loss of
The size of the notch plastic zone R may also be physical meaning. Then, it is preferable to use a critical
expressed by the following relationship: distance connected to the fracture process zone.
  It is also generally assumed that this length is intrin-
K ,c 2
R = A (16) sic to material and independent from notch geome-
y try, loading mode, etc. In procedure describe in Taylor
where A is a constant, K,c the notch fracture toughness (2004), this assumption is made and critical distance
and y the yield stress. is the same for a sharp or a blunt notch (Fig. 8). This
In Fig. 7, the evolution of A constant is reported assumption is also valid for the effective stress at fail-
not versus the notch radius, but versus the elastic stress ure. In the next section, this assumption is discussed in
concentration factor kt . We note that a negative power detail.

123
On characteristic lengths used in notch fracture mechanics 193

Fig. 9 Stress distribution at the notch tip from Lazzarin and


Fig. 8 Procedure to determine the characteristic length and the Tovo (1996). Range of do values indicates that they are inside
inherent material strength according to Taylor (2004). We note the region where Eq. 14 is valid
that this procedure is only valid for brittle materials

specimen. Distance do is in the range [0.010.1] for


The use of different values of , as reported in Fig. 7, PMMMA according to different authors.
affects the intrinsic character of critical distance and The maximum stress is precisely at the notch tip
failure stress c . For brittle material, stress distribution and is then included in the fracture process zone and
at the notch tip and for a distance x > Xef is governed satisfies the concept fracture process zone as the high
by a singular stress distribution with a r dependence stressed region. We note in Fig. 9 that the condition
and d0 > Xef . In this part, the product: d0 < Xef is only fulfilled for brittle material.

c . (d0 ) = cst (17)


4 Effective distance related to stress distribution
For a blunt notch Lazzarin and Tovo (1996) eval-
uated the linear elastic stress fields in the neighbor- In the VM (Pluvinage 2003), the effective distance is
hood of cracks and notches with an analytical solution considered as a characteristic of the stress distribution.
according to Muskhelishvilis method based on com- The VM is a local fracture criterion, which postulates
plex stress functions. They indicated that stress distri- that fracture process requires a certain volume. This
bution recovers the r dependence for a distance x > volume is a cylindrical volume with effective distance
0.1 mm. as its diameter and thickness as its height. The physical
Table 1 gives some values of critical distance accord- meaning of this fracture process volume is the high
ing to materials. Values of d0 are relatively small stressed region where the necessary fracture energy is
because the validity of theory of critical distance has stored. The difficulty is to find the limit of this high
been checked only for brittle material. In this case do stressed region. The limit is a priori not a material con-
is of the same order as characteristic distance Xc and stant but depends on loading mode, structure geometry
the behaviour of the material is practically elastic until and load level. The size of the fracture process volume
fracture. reduces to the effective distance according to the above
Figure 9 gives the stress elastic distribution from mentioned assumptions and is obtained by analysis of
Lazzarin and Tovo (1996), at the notch tip of a PMMA the stress distribution.

Table 1 Critical distance for different materials


Material References d0 (mm) Material References d0 (mm)

PMMA (at 60 C) Gmez and Elices (2004) 0.074 PVC Livieri (2008) 0.066
PMMA Livieri (2008) 0.090 Y-PSZ Yosibash et al. (2004) 0.022
PMMA Saghafi (2010) 0.086 Si3N4 Yosibash et al. (2004) 0.018
Alumina-7 % Zirconia Gogotsi (2003) 0.034 Martensitic steel Mehran et al. (2012) 0.090

123
194 G. Pluvinage, J. Capelle

stress gradient associated the effective distance with


the minimum of .
The relative stress gradient is given by:
1 yy(r )
(r ) = . (18)
yy(r ) r
where (r) and yy (r) are the relative stress gradient
and maximum principal stress or crack opening stress,
respectively. The effective stress for fracture is then
considered as the average volume of the stress distrib-
ution over the effective distance.
Effective distances listed in Table 2 are greater than the
critical distances listed in Table 1.
Fig. 10 Stress distribution at notch tip (Elastic plastic material). They are relative to more ductile materials than those
Determination of effective stress Xef by relative stress gradient listed in Table 1. However effective distance is always
procedure. Comparison of effective distance with notch plastic greater than the maximum stress distance, no matter
zone for a brittle material Ry (1) and a ductile one Ry (2) the ductility of the material. For very brittle material,
A graphical or numerical method associated with critical distance is less than the maximum stress dis-
relative stress gradient is used to determine effec- tance.
tive distance as inflexion point on opening stress dis- For these reasons, the length parameters are not com-
tribution. The minimum of relative stress gradient parable.
coincides with this inflexion point. Effective distance is associated with the stress dis-
The effective stress for fracture is then considered tribution which is sensitive to constraint.
as the average volume of the stress distribution over the Consequently, effective distance is also sensitive
effective distance. However, stresses are multiplied by to constraint. Several parameters have been used to
a weight function in order to take into account stress describe it: the constraint parameter L (Dowling and
gradient due to geometry and loading mode. Towley 1975), the stress triaxiality (Clausmeyer et al.
The bi-logarithmic elastic-plastic stress distribution 1991), the Q parameter (ODowd and Shih 1991) and
(Fig. 10) along the ligament exhibits three distinct T Stress (Betegon and Hancock 1991).
zones which can be easily distinguished. The elasto- T stress has been used as a constraint parameter
plastic stress primarily increases and attains a peak because it represents the modification of lateral con-
value, and then gradually drops to the effective dis- traction owing to diminution of Poissons effect. T
tance (zone I) and after to the elastic-plastic regime stress is one of the characteristics of stress distribu-
(zone II). Zone III represents linear behaviour in the tion at the crack or notch tip. For example, the elastic
bi-logarithmic diagram. It has been proven by exami- stress fields in a region surrounding the crack tip can
nation of fracture initiation sites that the effective dis- be characterised by the following solution (Williams
tance corresponds to the beginning of zone III, which 1957)
is in fact an inflexion point on this bi-logarithmic stress KI
distribution. A graphical method based on the relative i j = f i j ( )+T xi x j + A3 2r + O (r ) (19)
2r

Table 2 Effective distance for different materials


Material Yield stress (MPa) Notch radius (mm) Effective distance (mm)

XC 38 Osborne and Embury (1973) 304 0.152.0 15


CrMoV Rotor steel Pluvinage et al. (1999) 771 0.25 0.38
Al-Zn-Mg-cu Vratnica et al. (2010) 283.3 0.151.0 0.382.64
Al 7175-T74 Kim et al. (2004) 499 0.04.0 0.81.2
17-4PH Kim et al. (2004) 1170 0.02.0 1.21.45

123
On characteristic lengths used in notch fracture mechanics 195

Fig. 12 Evolution of the effective distance versus effective T


stress for a X52 steel pipe. Values obtained from four specimen
Fig. 11 Evolution of T stress versus distance over ligament. types (CT, SENT, DCB and RT)
Charpy V specimen made in X65 pipe steel

where K I is the stress intensity factor, f i j ( ) is the (DCB), single edge notch tensile (SENT) and roman
angular function, i j is the symbol of Kroneckers tile (RT). These specimens provide a large range of Tef
determinant. A polar coordinate system (r, ) with an stress from negative to positive values. We note that
origin at the crack tip is used. The second term is called DCB specimen exhibits only positive values of T stress.
the T -stress. T-stress is constant stress acting parallel The same kind of specimens have different ligament
to the crack line in the directionxx of the crack exten- size and then exhibit different Tef values. Effective dis-
sion with a magnitude proportional to the gross stress. tance has been plotted versus effective T stress, accord-
The non-singular term T may be a tensile or a com- ing to specimen type and ligament size (Fig. 12). We
pressive stress. Positive T -stress strengthens the level note that Xef increases when the constraint increases
of crack tip stress triaxiality and leads to high crack from negative to positive values. Effective distance is
tip constraint, while negative T -stress leads to the loss therefore not intrinsic to material but sensitive to con-
of constraint. Several methods have been proposed in straint through stress distribution.
literature to determine the T-stress for cracked speci-
men. The stress difference method has been proposed
by Yang and Ravi-Chandar (1999). In this method, the
5 Conclusion
T-stress is evaluated from stress distribution on the line
of crack extension. Generally computed by finite ele-
Local stress fracture criteria are associated with two
ment method, it is the difference between stress xx
parameters: scale length of the fracture volume process
parallel to crack plane and opening stress yy .
and a local fracture stress. Four kinds of approaches
For a notch, T stress is not constant along the lig-
of this scale length can be found and given different
ament as we can see in Fig. 11. T stress results have
names, but their use in a local stress fracture criterion
been obtained by finite element method and are rela-
is the same. The stress distribution is averaged over this
tive to a Charpy V specimen at critical load and for a
scale length to give the failure stress (line method) or
temperature below transition temperature.
the failure stress corresponding to the value of the stress
The material behaviour is assumed to be elastic plas-
distribution for this length (PM). One distinguishes:
tic. After some distance, T stress increases slowly with
distance. Effective T stress is defined as the T stress a characteristic length c generally equal to half the
value at distance Xef given by VM Tef = T(Xef ). T- notch radius,
stress distribution along the ligament (r) in the xx direc- a characteristic distance Xc considered as intrinsic
tion is obtained by stress different method (Yang
 and to material and connected to microstructure,
Ravi-Chandar 1999) T = Tx x = x x yy =0 . a critical distance d0 considered as intrinsic to mate-
Effective distance has been determined for four rial and connected to damage zone,
kinds of specimens used to determine fracture tough- an effective distance Xef considered as a character-
ness compact tension (CT), double cantilever beam istic of the stress distribution.

123
196 G. Pluvinage, J. Capelle

These different approaches increase the complexity Guillemot LF (1965) Brittle fracture on welded materials. Second
of the choice of the right scale parameter. Commonwealth Welding Conference London, pp 353382,
C7
This choice can be guided by the following consid- Irwin GR (1948) Analysis of stresses and strains near the end of
erations: a crack traversing a plate. Trans J Appl Mech 24:361364
The fracture process zone incorporates the maxi- Kaechele LE, Tetelman AS (1969) A statistical investigation of
mum stress according to the concept fracture process microcrack formation. Acta Metall 17:463475
Kim JH, Kim DH, Moon SI (2004) Evaluation of static and
zone is the highest stressed region. dynamic fracture toughness using apparent fracture toughness
The scale length is not intrinsic to material but of notched specimen. Mater Sci Eng A 11:387389
depends on notch geometry, loading mode, constraint Lazzarin L, Tovo R (1996) A unified approach to the evaluation
etc. of linear elastic stress fields in the neighborhood of cracks and
notches cracks and notches. Int J Fract 78:319
The first consideration is generally fulfilled if the Lazzarin P, Berto F (2005) Some expressions for the strain energy
material is very brittle or brittle and every approach in a finite volume surrounding the root of blunt V-notches. Int
can be used. However, they do not provide the same J Fract 135:16185
value of fracture stress. Then, transferability problem Leguillon D (2001) A criterion for crack nucleation at a notch in
homogeneous materials. C R Acad Sci Paris Ser IIb 329:97
occurs when measuring failure stress on a specimen 102
and applying it to predict the fracture of a structure or Livieri P (2008) Use of J-integral to predict static failures in sharp
component. V-notches and rounded U-notches. Eng Fract Mech 75:1779
If the material is ductile, characteristic length or crit- 1793
Mehran S, Rouhi B, Ramzani B, Barati E (2012) Fracture analy-
ical distance does not satisfy the concept of high stress sis of functionally graded materials with U- and V-notches
region and cannot be applied. The high stressed region under mode I loading using the averaged strain-energy den-
is too small in the first case and too large in the sec- sity criterion. Fatigue Fract Eng Mater Struct 35(7):614627
ond case. This fulfils the concept of effective distance, Meliani Hadj M, Matvienko Y, Pluvinage G (2011) Two-
parameter fracture criterion (K rho, c-Tef, c) based on notch
which is sensitive to constraint. The local fracture stress fracture mechanics. Int J Fract 167(2):173182
is therefore sensitive to constraint. However, to sepa- Minor HE, Kifani A, Louah M, Azari Z, Pluvinage G (2003)
rate effect of constraint on effective distance and local Struct Safety 25
fracture stress remains an open question. Neuber H (1937) Kerbspannungslehre. Springer, Berlin
Novozhilov VV (1969) On necessary and sufficient criterion of
brittle strength. Appl Math Mech (PMM) 33:212222
ODowd NP, Shih CF (1991) Family of crack tip fields character-
References ized by a triaxiality parameter. J Mech Phys Solids 39(8):989
1015
Akourri O, Elayachi I, Pluvinage G, Elayachi I, Bensalah MO, Osborne DE, Embury JD (1973) The influence of warm rolling on
Lebienvenu M, Dlouhy I (2005) To joint effect of tempera- the fracture toughness of bainitic steel. Metall Trans 4:2051
ture and notch root radius on fracture toughness. Eng Mech 2061
12(1):1122 Peterson RE (1959) Notch sensitivity. In: Sines G, Waisman JL
Betegon C, Hancock JW (1991) Two parameter characteristics (eds) Metal fatigue. MacGraw Hill, New York
of elastic-plastic crack-tip fields. J Appl Mech 58:104101 Pluvinage G, Azari Z, Kadi N, Dlouhy I, Kozak V (1999)
Clausmeyer H, Kussmaul K, Roos E (1991) Influence of stress Effect of ferritic microstructure on local damage zone dis-
state on the failure behaviour of cracked components made of tance associated with fracture near notch. Theor Appl Fract
steel. ASME Appl Mech Rev 44(2):7792 Mech 31:149156
Creager M, Paris PC (1967) Elastic field equations for blunt Pluvinage G (2001) Notch effects in fatigue and fracture. Kluwer,
cracks with reference to stress corrosion cracking. Int J Fract Dordrecht
3:247252 Pluvinage G (2003) Fracture and fatigue emanating from stress
Dowling AR, Towley CHA (1975) The effect of defect on struc- concentrators. Kluwer, Dordrecht
tural failures: a two-criteria approach. Int J Press Vessel Pip Pluvinage G, Montariol F (1971) Influence de la taille de grain
3(2):77107 sur les courbes COD-Temprature dessai dun acier doux.
Filippi S, Lazzarin P, Tovo R (2002) Developments of some Revue Metallurgie 68(4):7381
explicit formulas useful to describe elastic stress fields ahead Ritchie RO, Knott JF, Rice JR (1973) On the relationship between
of notches in plates. Int J Solid Struct 39:45434565 critical tensile stress and fracture toughness in mild steel. J
Gogotsi GA (2003) Fracture toughness of ceramics and ceramic Mech Phys Solids 21:359410
composites. Ceram Int 29(7):777784 Saghafi H, Ayatollahi MR, Sistaninia M (2010) A modified MTS
Gmez FJ, Elices MA (2004) Fracture criterion for blunted V- criterion (MMTS) for mixed-mode fracture toughness assess-
notched samples. Int J Fract 127:239264 ment of brittle materials. Mater Sci Eng A 527(2122):5624
Gmez FJ, Elices MA (2004) Fracture criterion for blunted V- 5630
notched samples. Int J Fract 127:239264

123
On characteristic lengths used in notch fracture mechanics 197

Seweryn A (1994) Brittle fracture criterion for structures with Whitney JM, Nuismer RJ (1974) Stress fracture criteria for lam-
sharp notches. Eng Fract Mech 47(4):673681 inated composites containing stress concentrations. J Comp
Sih GC (1974) Strain energy density factor applied to mixed Mater 8:253265
mode crack problems. Int J Fract 10(3):305-321 Williams ML (1957) On the stress distribution at the base of
Taylor D (2004) Predicting the fracture strength of ceramic mate- stationary crack. ASME J Appl Mech 24:1091421
rials using the theory of critical distances. Eng Fract Mech Yang B, Ravi-Chandar K (1999) Evaluation of elastic T-stress by
71:24072416 the stress difference method. Eng Fract Mech 64:589605
Tovo R, Livieri P (2008) An implicit gradient application to Yosibash Z, Bussiba A, Gilad I (2004) Fracture criteria for brittle
fatigue of complex structures. Eng Fract Mech 75:18041814 elastic materials. Int J Fract 125:307333
Vratnica M, Pluvinage G, Jodin P, Cvijovic Z, Rakin M, Burzic
Z (2010) Influence of notch radius and microstructure on
the fracture behaviour of AlZnMgCu alloys of different
purity. Mater Des 31:17901798

123

You might also like