MACE31302: Aerospace Structures 3 Lecturer: Dr.

KB Katnam

Tutorial Solutions: Fracture Mechanics

Question 1

A large plate in a structure is subjected to a uniform uni-axial tensile stress, σ, at the

boundaries as shown in Figure Q1. If a through-thickness centre-crack with length 2a,
perpendicular to the loading direction, exists in the plate, then calculate the following by
using linear elastic fracture mechanics:

Figure Q1: A large plate with a centre-crack in mode-I loading

(i) The mode-I stress intensity factor at the crack tip assuming a tensile stress of 50 MPa
and a total crack length of 20 mm;

Solution:

First, you need to go through ‘slides 9 & 10’ in the handout and understand the concept of
mode-I and corresponding stress intensity factor (KI).

For this particular case, the mode-I stress intensity factor can be calculated when you
know the remotely applied stresses (i.e. 𝜎) and half of the crack length (i.e. a) by using
the following equation,

𝐾𝐼 = 𝑌𝜎√𝜋𝑎 = 1.0 × 50 × √𝜋 × 10/1000 = 8.86 𝑀𝑃𝑎√𝑚

You should pay attention to the following points:

 Y is the geometric correction factor, and Y≈ 1.0 for a large plate with a centre-crack;
 The total crack length (i.e. measured from the first tip to the second tip) is 2a;
 The units of the crack length, a, is converted from ‘mm to ‘m’ to be consistent with the
units used for the stress intensity factor (i.e. 𝑀𝑃𝑎√𝑚).

(ii) The stress tensor at a point with polar coordinates (r, θ) = (0.5 mm, 45o) assuming the
crack plate as (a) thin, and (b) thick;

Page 1 of 20
 MACE31302: Aerospace Structures 3 Lecturer: Dr. KB Katnam

Solution:

You already know what we mean by stress tensor at a material point (if not, stop here
and go through the tutorial-1 questions/solutions). Assuming you have mastered the
concepts of ‘stress tensor’ and ‘von Mises yield criterion’, let us calculate the individual
stress components and arrange them in a tensor form. Here, it is important to note that
the mode-I stress intensity factor (i.e. KI) defines the near-crack-tip stress state.
Therefore, the stress state, and thus the corresponding stress tensor, can be calculated
by using the following equations (see the figure in ‘slide 9’ and pay attention to the sign
convention; and the set of equations given for the stress components in ‘slide 10’),
𝐾𝐼 𝜃 𝜃 3𝜃 8.86 𝜋 𝜋 3𝜋
𝜎𝑥 = cos (1 − sin sin ) = cos (1 − sin sin ) = 94.4 𝑀𝑃𝑎
√2𝜋𝑟 2 2 2 √2𝜋 × 0.5 × 10−3 8 8 8
𝐾𝐼 𝜃 𝜃 3𝜃
𝜎𝑦 = cos (1 − sin sin ) = 197.6 𝑀𝑃𝑎
√2𝜋𝑟 2 2 2
𝐾𝐼 𝜃 𝜃 3𝜃
𝜏𝑥𝑦 = sin cos cos = 21.4 𝑀𝑃𝑎
√2𝜋𝑟 2 2 2

So, the in-plane normal stresses (i.e. 𝜎𝑥 and 𝜎𝑦 ) and in-plane shear stresses (i.e. 𝜏𝑥𝑦 ) are
calculated. The stresses that are acting on the z-plane, which will depend on the stress
conditions (i.e. either plane stress or plane strain), defines the complete stress tensor at
this material point, i.e. (r, θ) = (0.5 mm, 45o).

(a)

If the cracked plate is thin, then we have ‘plane stress’ conditions (see ‘slides 11, 18, 19 &
20’). Hence the stress components acting on the z-plane are zero. The stress tensor is,
94.4 21.4 0
𝝈 = [21.4 197.6 0] 𝑀𝑃𝑎
0 0 0
(b)

If the plate is thick, then we assume that ‘plane strain’ conditions dominate (see ‘slides 11,
18, 19 & 20’). Therefore the shear stress components acting on the z-plane are zero and
the normal stress component, 𝜎𝑧 = 𝜈(𝜎𝑥 + 𝜎𝑦 ) = 87.6 𝑀𝑃𝑎. The stress tensor is,
94.4 21.4 0
𝝈 = [21.4 197.6 0 ] 𝑀𝑃𝑎
0 0 87.61
(iii) If the material yield strength and Poisson’s ration are 300 MPa and 0.3, respectively,
then check whether yielding occurs at this point based on the von Mises criterion.

Solution:

As the complete stress tensor at this material point is known, you can calculate the in-
plane principal stresses (i.e. 𝜎1 and 𝜎2 ) and then apply the von Mises yield criterion (here
the third principal stress component 𝜎3 = 𝜎𝑧 ) to calculate the effective stress (i.e.

Page 2 of 20
 MACE31302: Aerospace Structures 3 Lecturer: Dr. KB Katnam

1
𝜎̅ = √(𝜎1 − 𝜎2 )2 + (𝜎2 − 𝜎3 )2 +(𝜎1 − 𝜎3 )2 ). If the effective stress is less than the yield
√2
strength (i.e. 𝜎̅ < 𝜎𝑌 ), then you don’t expect yielding at this point.

For the plane stress condition (i.e. thin plate),

𝜎̅ = 144.8 𝑀𝑃𝑎 < 𝜎𝑌 = 300 𝑀𝑃𝑎 ⟹ No yielding

For the plane strain condition (i.e. thick plate),

𝜎̅ = 113.0 𝑀𝑃𝑎 < 𝜎𝑌 = 300 𝑀𝑃𝑎 ⟹ No yielding

Important Note:
For the material point considered, which is at a distance of 0.5 mm from the crack tip, you
don’t expect the material to yield. This means that you have a very small plastic zone
(obviously < 0.5 mm in size!) at the crack tip (well, at both the crack tips) for this
combination of ‘remotely applied stress’, ‘crack size’ and ‘yield strength’. This makes the
whole analysis valid (i.e. small scale yielding and LEFM condition).

Question 2

Considering a large thin plate, with a through-thickness centre-crack, as shown in Figure Q1,
subject to a remotely applied uniform tensile stress acting perpendicular to the crack length,

(i) Calculate the mode-I stress intensity factor at the crack tip for: (a) a tensile stress of
100 MPa and a crack length of 20 mm; and (b) a tensile stress of 50 MPa and a crack
length of 80 mm;

Solution:

As the direction of the remotely applied stresses (tensile and uniform) is perpendicular to
the crack length, you have mode-I (i.e. opening mode; see ‘slide 9’) loading conditions.
The mode-I stress intensity factor can be calculated when the remotely applied stress and
crack length are known.

(a)

The remotely applied stress is 100 MPa, and the crack length (remember, it is a centre-
crack with two crack tips, i.e. the length is 2a) 2a = 20 mm. As it is a large plate (as no
dimensions are given, assume it as infinitely large), the geometric correction factor Y=1.0.

𝐾𝐼 = 𝑌𝜎√𝜋𝑎 = 1.0 × 100 × √𝜋 × 10/1000 = 17.70 𝑀𝑃𝑎√𝑚

(b)

In this case, the remotely applied stress is 50 MPa, and the crack length 2a = 80 mm.
Again, it can be considered as infinitely large, the geometric correction factor Y=1.0.

𝐾𝐼 = 𝑌𝜎√𝜋𝑎 = 1.0 × 50 × √𝜋 × 40/1000 = 17.70 𝑀𝑃𝑎√𝑚

Important Note:
You see that, in (a) as well as (b), the mode-I stress intensity factor is 17.70 𝑀𝑃𝑎√𝑚. The
combination of ‘remotely applied load’ and ‘crack size’ is what defines the stress intensity
factor, not one of them alone. Moreover, the near-crack-tip stresses (seen in Question 1)
are defined only by the stress intensity factor (i.e. the combination of ‘remotely applied

Page 3 of 20
 MACE31302: Aerospace Structures 3 Lecturer: Dr. KB Katnam

stress’ and ‘crack length’, not one of them alone). This further means that the stress
distribution near the crack tip in both (a) and (b) are the same. This is called ‘similitude’
(i.e. similar conditions). Therefore, you need to keep in mind that, for a given geometry
and material fracture toughness, the combination of ‘remotely applied stress’ and ‘crack
length’ is what drives fracture! Don’t just think that large cracks are bad and relatively
small cracks are not!

(ii) The material fracture toughness, which is experimentally measured using a thick
specimen, is equal to 537.6 MPa √mm. Check whether or not catastrophic fracture
occurs for the two conditions (a) and (b) given in (i).

Solution:

You need to understand the concept of ‘fracture toughness’ (see ‘slide 13’). It is the
‘critical level’ of stress intensity factor at which the material fracture toughness (i.e. KIc
for mode-I loading) is completely used up and catastrophic failure occurs (i.e. unstable
crack growth). For a given ‘crack length’, ‘component geometry’ and ‘loading condition
(e.g. mode-I)’, if the remotely applied load is increased gradually, then the stress intensity
factor at the crack tip will be increased and eventually reach a stage where the stress
intensity factor is equal to the material fracture toughness. Or you can imagine a similar
situation where the remotely applied stress is fixed, but the crack length gradually
reaches a ‘critical size’ (e.g. fatigue crack growth) and then causes catastrophic fracture.

In this case, the material fracture toughness is measured using a thick specimen (i.e. plane
strain conditions exist for fracture).

The plane strain fracture toughness (note that the units are converted to 𝑀𝑃𝑎 √𝑚 ),

𝐾𝐼𝑐 = 537.6 𝑀𝑃𝑎√𝑚𝑚 = 17.0 𝑀𝑃𝑎√𝑚

(a)

From Part (i)-(a), the remotely applied stress is 100 MPa, and the crack length is 20 mm.
The stress intensity factor is 17.70 𝑀𝑃𝑎√𝑚. This is greater than the ‘plane strain’
fracture toughness measured experimentally. Hence,

𝐾𝐼𝑐 = 17.00 𝑀𝑃𝑎√𝑚 < 𝐾𝐼 = 17.70 𝑀𝑃𝑎√𝑚

Catastrophic fracture is very likely.

(Note: you are predicting this by using the ‘plane strain’ fracture toughness)

(b)

From Part (i)-(b), the remotely applied stress is 50 MPa, and the crack length is 80 mm.
The stress intensity factor is 17.70 𝑀𝑃𝑎√𝑚. This is greater than the ‘plane strain’
fracture toughness measured experimentally. Hence,

𝐾𝐼𝑐 = 17.00 𝑀𝑃𝑎√𝑚 < 𝐾𝐼 = 17.70 𝑀𝑃𝑎√𝑚

Catastrophic fracture is very likely.

(Note: you are again predicting this by using the ‘plane strain’ fracture toughness)

Page 4 of 20
 MACE31302: Aerospace Structures 3 Lecturer: Dr. KB Katnam

This is what you expect because of the concept of ‘similitude’.

(iii) Comment on the material property used and its influence on the two conditions
analysed.

Solution:

It is important to note that the ‘plane strain’ fracture toughness is less than the ‘plane
stress’ fracture toughness (see ‘slides 19 & 20’) because of the ‘constraint imposed’ and
‘the size of plastic zone’ at the crack tip.

Moreover, it is important to notice here that the cracked plate is ‘thin’ and thus have
‘plane stress conditions’. Using the ‘plane strain’ fracture toughness in the analysis makes
it ‘a conservative approach’. Therefore, the ‘plane stress’ fracture toughness ‘might just
be enough’ to prevent fracture when the stress intensity factor is 17.70 𝑀𝑃𝑎√𝑚 (i.e. Part
(i)-(a) & (i)-(b)). But always use the plane strain fracture toughness!

Question 3

In a steel plate (6 m width x 12 m height x 4 mm thick) in a large structure, a centre-crack

parallel to the width of the plate was detected during inspection. Use Brown’s equation for
estimating the geometric correction factor when necessary.

(i) For a remotely applied tensile stress of 50 MPa, calculate the stress intensity factor
when the crack length is (a) 20 mm and (b) 300 mm?

Solution:

You can see that the plate, with a centre-crack, is large but with finite dimensions. The
stress intensity factor will have a geometric correction factor, which can either be equal
to 1.0 or greater than 1.0, depending on the ratio ‘a/W’ (go through ‘slide 14’). The
geometric factor, Y, can be calculated using Brown’s or Feddersen’s equations.

(a)

The remotely applied tensile stress is 50 MPa (i.e. mode-I condition), and the crack length
2a = 20 mm. The width of the plate W = 6 m = 6000 mm.

From Brown’s equation, Y ≈ 1.0004 for (a/W) = 10/6000.

The stress intensity factor is,

𝐾𝐼 = 1.0004 × 50 × √𝜋 × 10/1000 = 8.86 𝑀𝑃𝑎√𝑚

(b)

In this case, the crack length 2a = 300 mm. From Brown’s equation, Y ≈ 1.006 for (a/W) =
150/6000. The stress intensity factor is,

𝐾𝐼 = 1.006 × 50 × √𝜋 × 150/1000 = 34.3 𝑀𝑃𝑎√𝑚

Page 5 of 20
 MACE31302: Aerospace Structures 3 Lecturer: Dr. KB Katnam

(ii) What is the critical crack length if it is subjected to a load that introduces a uniform
nominal tensile stress of 100 MPa? Plane-strain fracture tests performed on a set of
lab specimens suggest that the fracture toughness is approximately 80 MPa √𝒎.

Solution:

You need to calculate the critical crack length, i.e. the crack length that leads to a stage
where the mode-I stress intensity factor is equal to the mode-I fracture toughness.

Note that, if you have components with mode-II (or mode-III) loading condition, then you
need to use the mode-II (or mode-II) stress intensity factor and mode-II (or mode-III)
fracture toughness value to calculate the corresponding mode-II (or mode-III) critical
crack length. Hope it is clear! But, because mode-I is often the critical mode, the mode-I
critical crack length is the critical ‘critical crack length’ for ensuring structural integrity!

The plane strain fracture toughness measured from lab specimens is 80 𝑀𝑃𝑎√𝑚.

The stress intensity factor for the critical crack length ac is,

𝐾𝐼 = 𝑌𝜎√𝜋𝑎𝑐 = 𝐾𝐼𝑐

1 𝐾𝐼𝑐 2
⟹ 𝑎𝑐 = ( )
𝜋 𝑌𝜎
To solve this equation, you can try two things: (a) substitute Brown’s equation for Y in
terms of (ac /W) and solve the equation (hard work!), or (b) assume an initial value for Y
and solve iteratively until it converges (relatively simple). If you follow the path (b), then
you can assume that the critical crack length is small and Y≈1.0 (this can be verified away
once you calculate ac).

If Y≈1.0 and the remotely applied stress is 100 MPa, then

2
1 𝐾𝐼𝑐 2 1 80
⟹ 𝑎𝑐 = ( ) = ( ) = 0.203 𝑚 = 203 𝑚𝑚
𝜋 𝑌𝜎 𝜋 1.0 × 100
Now, you can verify whether Y≈1.0 or not. For a = 203 mm, the geometric correction
factor Y=1.0059 from Brown’s equation. So the assumption that you made (i.e. Y≈1.0) is
valid. The critical crack length 2a = 406 mm.

In case Y > 1.0, it is not valid (e.g. this could be possible when the remotely applied stress
is low or fracture toughness is very high), then you could repeat this and iteratively solve
such a problem.

(iii) When a centre-crack of 300 mm was present, it failed catastrophically under a

nominal tensile stress of 120 MPa. Estimate the fracture toughness of the material?
Compare the estimated value with the fracture test data given in (ii).

Solution:

As the plate had failed catastrophically for a remotely applied stress of 120 MPa and a
crack length of 300 mm, you can calculate the stress intensity factor for this combination
of ‘σ’ and ‘a’. Moreover, as the stress intensity factor had led to failure, it must be equal

Page 6 of 20
 MACE31302: Aerospace Structures 3 Lecturer: Dr. KB Katnam

to the material fracture toughness. Note that the geometric correction factor Y≈1.0 for a
crack length 2a=300 mm from Brown’s equation.

The fracture toughness is,

𝐾𝐼𝑐 = 𝑌𝜎√𝜋𝑎𝑐 = 1.0 × 120√𝜋 × 150/100 = 82.4 𝑀𝑃𝑎√𝑚

As you would expect (I hope!), the plane strain fracture toughness, which was measured
experimentally using lab specimens (i.e. CT or SENB geometries as shown in ‘slide 22’), is
very close (80.0 𝑀𝑃𝑎 √𝑚) to the estimated fracture toughness (82.4 𝑀𝑃𝑎√𝑚) from the
plate fractured catastrophically for a combination of ‘centre-crack length’ and ‘remotely
applied stress’.

Question 4

A titanium 6Al-4V plate (40 mm width x 120 mm height x t mm thick) used as a structural
member is loaded in tension. Non-destructive tests had identified an edge crack parallel to
the width of the plate as shown in Figure Q4. If the mode-I tensile load is 55 kN and the edge
crack is 6 mm, what is the minimum plate thickness required to provide a safety factor of 3.0
on fracture? Assume that the fracture toughness of titanium 6Al-4V is 66 MPa √𝒎.

Figure Q4: A plate with an edge crack in mode-I loading

Solution:

You can see that the plate, with an edge-crack, has finite dimensions. Moreover, the plate
‘thickness is not given’. First, you need to understand the concept of ‘safety factor on
fracture’. As you already know, based on LEFM assumptions, catastrophic fracture occurs
when the stress intensity factor (for a given combination of σ, a and Y) is equal to the
material fracture toughness. In this case, as you have an edge-crack under mode-I
condition, the safety factor on fracture can be defined as ‘the ratio of the mode-I fracture
toughness to the edge-crack mode-I stress intensity factor’ (i.e. 𝐾𝐼𝑐 = 𝑛𝐾𝐼 , where 𝑛 is the
safety factor).

Page 7 of 20
 MACE31302: Aerospace Structures 3 Lecturer: Dr. KB Katnam

The edge-crack mode-I stress intensity factor will have a geometric correction factor (Y),
which can either be equal to 1.12 or greater than 1.12, depending on the ratio ‘a/W’ (go
through ‘slide 15’). This can be calculated using Brown’s equations.

You need to calculate the remotely applied stress for the tensile load P = 55 kN.

𝑃 55 × 103 1.375 × 103

𝜎= = = 𝑀𝑃𝑎
𝐴 40 × 𝑡 𝑡
From non-destructive inspection, you have an edge-crack for 6 mm (‘only one crack tip’,
therefore ‘a = 6 mm’). The ratio a/W = 6/40 = 0.15; and the geometric correction factor Y ≈
1.26 from Brown’s equation.

The mode-I stress intensity factor,

1.375 × 103
𝐾𝐼 = 𝑌𝜎√𝜋𝑎 = 1.26 × × √𝜋 × 6/1000 𝑀𝑃𝑎√𝑚
𝑡

For providing a safety factor n=3.0 on fracture,

𝐾𝐼𝑐 = 𝑛𝐾𝐼
1.375 × 103
⟹ 𝐾𝐼𝑐 = 𝑛 × 1.26 × × √𝜋 × 6/1000 𝑀𝑃𝑎√𝑚
𝑡
Solving this equation for 𝐾𝐼𝑐 = 66.0 𝑀𝑃𝑎√𝑚 and 𝑛 = 3.0, you see that the minimum plate
thickness required is ‘approximately 11.0 mm’.

Note: This safety factor is only for this combination of ‘edge-crack size’ and ‘remotely
applied stress levels’ and ‘thickness calculated’. If any of these parameters change, then
the safety factor on fracture also changes accordingly.

Question 5

A 2014-T651 aluminium plate (100 mm width x 200 mm height x 5 mm thick) with a centre-
crack is subjected to a mode-I service load of 50 kN. Assume that the fracture toughness of
the material is 24 MPa √𝒎.

(i) What is the largest crack length that can be permitted for a safety factor of 3.0 on
stress in order to avoid fracture?

Solution:

First, you need to go through the solution provided for Question 4 and understand what
you mean by ‘a safety factor on fracture’. In this case, you need a safety factor of 3.0 ‘on
stress’. As the stress intensity factor is proportional to the remotely applied stress, the
safety factor ‘on stress’ is equal to the safety factor ‘on fracture’.

For a safety factor n=3.0 on fracture, the stress intensity factor must be,
𝐾𝐼𝑐 24
𝐾𝐼 = = = 8.0 𝑀𝑃𝑎√𝑚
𝑛 3
The corresponding crack length a is,

Page 8 of 20
 MACE31302: Aerospace Structures 3 Lecturer: Dr. KB Katnam

𝐾𝐼 = 𝑌𝜎√𝜋𝑎 =
2
1 𝐾𝐼 2 1 8.0
⟹𝑎= ( ) = ( ) ≈ 2.0 𝑚𝑚
𝜋 𝑌𝜎 𝜋 1.0 × [50 × 1000/(100 × 5)]

Here, the geometric correction factor Y ≈ 1.0 for the given plate dimensions. The largest
crack permitted is 4 mm (i.e. 2a, the total crack length).

(ii) What is the safety factor on crack length for the safety factor used on stress in (i)?

Solution:

The idea here is that the safety factor can also be defined using crack size. If the current
crack length is a and the critical crack length for fracture is ac, then the ratio ‘ac/a’ is
defined as the safety factor on crack length. Note that this ratio is calculated for the same
remotely applied stress.

For the safety factor on fracture used in (i), the largest crack length permitted is 4 mm
(2a). But the critical crack length is (you can solve it iteratively by assuming Y=1.o initially),

2
1 𝐾𝐼𝑐 2 1 24
⟹ 𝑎𝑐 = ( ) = ( ) = 16.2 𝑚𝑚
𝜋 𝑌𝜎 𝜋 1.06 × [50 × 1000/(100 × 5)]

Here, you can check that the geometric correction factor Y is approximately 1.06 for the
given plate dimensions and the critical crack length estimated.

The safety factor on crack length is,

ac ⁄a = 16.2/2.0 = 8.10

Note: The safety factor on fracture is not equal to the safety factor on crack length.

(iii) What is the safety factor on yielding? Assume that the yield strength of 2014-T651 is
415 MPa. When compared with (i), do you expect brittle fracture or yielding?

Note:
Safety factor on fracture = fracture toughness / stress intensity factor = KIc ⁄KI
Safety factor on crack length = critical crack length / crack length = ac ⁄a
Solution:
Here, you need to assume two possible failure conditions, i.e. yielding and brittle fracture,
when you have a centre-crack of 4 mm. First, assume (just neglect the crack for a
moment!) that the plate would fail due to yielding and estimate the safety factor on
yielding (the ratio of the yield strength to the remotely applied stress, i.e. σY/σ).

The safety factor on yielding is, σY/σ = 415/100 = 4.15.

Now, assume that the plate would fail due to brittle fracture and estimate the safety
factor on fracture (you have it already in (i)). The safety factor on fracture is 3.0. As you

Page 9 of 20
 MACE31302: Aerospace Structures 3 Lecturer: Dr. KB Katnam

can see, the safety factor on fracture is less than the safety factor on yielding. Therefore,
one would expect ‘brittle fracture’.

Question 6*

A spherical pressure vessel, made of ASTM A517-F steel, operates at room temperature. The
inner diameter and wall thickness are 1.5 m and 10 mm, respectively. It can be subjected to a
maximum pressure of 6 MPa. Moreover, the fracture toughness and yield strength are known
to be 187 MPa √𝒎 and 760 MPa, respectively.

(i) If a crack exists on the surface (i.e. not a through-thickness crack, as shown in Figure
Q6) and propagates over a period of time in service, do you expect a leak before full
fracture (i.e. leak-before-break condition)?

Figure Q6: A spherical pressure vessel with an existing internal surface crack

Solution:

First, you need to go through the concept of ‘hoop stresses in thin spherical vessels’ (refer
to PP Benham, RJ Crawford and CG Armstrong: Mechanics of Engineering Materials) and
understand that the in-plane stresses (with no in-plane shear stresses) are the principal
stresses and equal to ‘pr/2t’ (p is the internal pressure, r is the radius, and t is the
thickness).

For the given dimensions of the spherical vessel and internal pressure,

The hoop stress is,

𝑝𝑟 6 × 0.75 × 100
𝜎= = = 225 𝑀𝑃𝑎
2𝑡 2 × 10
Now, it is important to note that there is a small internal surface crack (not a through-
thickness crack). Assuming that this crack can propagate in service due to fluctuating
internal pressure over a period of time, you can imagine the crack tip propagating in all
possible directions (i.e. towards the bulk of the wall and external surface as shown in
Figure Q6(a).

The initial crack starts to radially propagate from the Point O and eventually reach the
lines P-Q and P*-Q*, and becomes a through-thickness crack with a crack length of 2t
(when measured from the point P to the point P*). When it becomes a through-thickness

Page 10 of 20
 MACE31302: Aerospace Structures 3 Lecturer: Dr. KB Katnam

crack 2a = 2t, you can consider it as a centre-crack in a large plate (here the internal radius
of the vessel is very large compared to the wall thickness; therefore the radius of
curvature is also quite large and thus this assumption is quite reasonable to make!). Note
that accurate analysis can be performed using computational methods (i.e. Abaqus or
Ansys software for LEFM).

Figure Q6(a): A radially propagating internal surface crack

When the initial crack becomes a through-thickness crack, there are two possible
phenomena can be observed for a given internal pressure and material type: (a) leak or
(b) complete fracture. As the crack length (i.e. 2a = 2t) is known, the stress intensity
factor can be calculated for an internal pressure of 6 MPa (or in other words, when the
hoop stress is 225 MPa).

The stress intensity factor is,

𝐾𝐼 = 𝑌𝜎√𝜋𝑎 = 𝑌𝜎√𝜋𝑡 = 1.0 × 225 × √𝜋 × 10/100 ≈ 39.9 𝑀𝑃𝑎√𝑚

However, the fracture toughness of ASTM A517-F steel is 187.0 𝑀𝑃𝑎√𝑚, which is greater
than the stress intensity factor for this crack length (i.e. a = t). This means it cannot cause
catastrophic fracture and thus you would expect it to ‘leak’.

(ii) What is the safety factor on (a) fracture and (b) yielding?

Solution:

Similar to Question 5 (iii), here, again you need to assume two possible failure conditions,
i.e. yielding and brittle fracture, when you have a centre-crack of 20 mm (because 2a = 2t
= 20 mm). First, assume that it would fail due to yielding and estimate the safety factor
on yielding.

The safety factor on yielding is, σY/σ = 760/225 = 3.38. Notice here that σ is also the
effective stress as you have a bi-axial stress state!

Now, assume that it would fail due to brittle fracture and estimate the safety factor on
fracture. Then, the safety factor on fracture is,

KIc/KI = 187.0/39.9 =4.69.

As you can see, the safety factor on fracture is greater than the safety factor on yielding.
Therefore, one would expect ‘yielding’.

Note: If you estimate the critical crack length required for catastrophic fracture, you
would see that it is equal to 440 mm!

Page 11 of 20
 MACE31302: Aerospace Structures 3 Lecturer: Dr. KB Katnam

Question 7

A centre-cracked plate (100 mm width x 200 mm height x 5 mm thick) is subjected to a tensile

load of 50 kN. Assuming the yield strength of the material as 415 MPa,

(i) For a centre-crack of 20 mm, check whether plane strain conditions exist?

Solution:

First, you need to go through ‘slides 17 to 21’ and understand the crack tip plastic zone
and its size, and how it is affected by plane stress and strain conditions. For a given crack
length and remotely applied stress, you can calculate the stress intensity factor and then
use it to estimate the size of the plastic zone at the crack tip.

The stress intensity factor (using Brown’s equation for Y; approximately 1.0) is,

𝐾𝐼 = 𝑌𝜎√𝜋𝑎 ≈ 1.0 × 100 × √𝜋 × 10/1000 ≈ 17.72 𝑀𝑃𝑎√𝑚

For ensuring the plane strain and LEFM conditions, the plastic zone size needs to be very
small when compared to the following geometric parameters ‘see the figures in slide 21’:
(a) the distance between the crack tip and the plate boundaries (i.e. the width-wise
ligament, which is equal to (100-20)/2 = 40 mm; and the height-wise ligament, which is
equal to 200/2 = 100 mm), (b) the thickness of the specimen, i.e. 5 mm, and (c) the crack
length, i.e. a = 10 mm. Moreover, satisfy the following condition ‘see the figure in slide 21’,

𝐾𝐼 2 17.72 2
𝑏 − 𝑎, ℎ, 𝑎, 𝑡 ≥ 2.5 ( ) = 2.5 ( ) = 4.5 𝑚𝑚
𝜎𝑌 415

As you can see, all the geometric parameters (i.e. 𝑏 − 𝑎, ℎ, 𝑎, and 𝑡) satisfy (notice that the
thickness t = 5 mm, thus is the critical parameter) the plain strain and LEFM conditions.

(ii) Based on (i), estimate the plastic zone size near the crack tip using the Irwin’s model.

Solution:

You need to go through ‘slides 17’. The Irwin model for estimating the size of the crack tip
plastic zone is,

∗
1 𝐾𝐼 2 1 17.72 2
𝑟 𝑝 = ( ) = ( ) = 0.19 𝑚𝑚
3𝜋 𝜎𝑌 3𝜋 415

This is very small (approximately 200 microns, which is very close to the diameter of
human hair!). This is the size of the plastic zone for the plane strain region of the crack
front (see ‘slide 19’). You would expect a slightly bigger plastic zone near the surfaces
where plane stress conditions exist. Moreover, this is what you mean by small scale
yielding (compare it with all the other dimensions!).

Question 8

Compact tension specimens are prepared to determine the fracture toughness of the 7178
aluminium alloy (i.e. 7178-T651 and 7178-T7651), subjected to two different heat treatments.

Page 12 of 20
 MACE31302: Aerospace Structures 3 Lecturer: Dr. KB Katnam

The crack length and thickness used for the test specimens are 40 mm and 10 mm,
respectively.

(i) The critical stress intensity factor and yield strength values are 23.10 MPa √𝒎 and 570
MPa for 7178-T651. Check whether plane strain test conditions exist for the specimen
dimensions used?

Solution:

You need to go through ‘slides 22 & 23’ and understand what you mean by a compact
tension specimen as well as the valid test conditions. For plane strain conditions, the crack
length, a, and the specimen thickness t (well, B in ‘slide 23’ as that figure is taken from a
textbook) must be

𝐾𝐼𝑐 2 23.10 2
𝑎, 𝑡 ≥ 2.5 ( ) = 2.5 ( ) = 4.1 𝑚𝑚
𝜎𝑌 570

As the crack length and specimen thickness used for the tests are 40 mm and 10 mm,
respectively, the tested geometry thus satisfies plain strain fracture conditions. It is
important to note that the validity of tests not only depend on the specimen geometry but
also on the material properties (i.e. the combination of fracture energy and yield strength).

(ii) Similarly, for 7178-T7651, check whether plane strain test conditions exist? Consider
that the critical stress intensity factor and yield strength values are 33.0 MPa √𝒎 and
490 MPa, respectively. Comment of the specimen geometries used in the tests.

Solution:

Similar to Part (i), for plane strain conditions, the crack length, a, and the specimen
thickness, t, must be

𝐾𝐼𝑐 2 33.0 2
𝑎, 𝑡 ≥ 2.5 ( ) = 2.5 ( ) = 11.4 𝑚𝑚
𝜎𝑌 490

As the crack length and specimen thickness used for the tests are 40 mm and 10 mm,
respectively, the tested geometry thus does not satisfy plain strain fracture conditions. As
mentioned in Part (i), the validity of these tests not only depend on the specimen
geometry but also on the material properties (i.e. the combination of fracture energy and
yield strength).

For the combination of 33.o MPa √𝑚 and 490 MPa, the plastic zone is relatively large
(when compared to the plastic zone for 7178-T651) and therefore the specimen thickness
must be increased (at least > 11.4 mm) to obtain the plane strain fracture toughness of
7178-T7651.

Question 9*

A structural component in the form of a large flat sheet needs to be fabricated. A specific
grade of steel, i.e. 0.45C-Ni-Cr-Mo, is shortlisted during material selection process. For
ensuring safety and robust structural maintenance conditions, the critical crack length in such
a component must be greater than 3 mm (driven by a specific non-destructive inspection

Page 13 of 20
 MACE31302: Aerospace Structures 3 Lecturer: Dr. KB Katnam

technique and its resolution). A design stress level of one-half of the tensile strength is
considered. Moreover, to save weight, an increase in the tensile strength from 1520 MPa to
2070 MPa is also suggested. The plane strain fracture toughness of 0.45C-Ni-Cr-Mo is 66.0
MPa √𝒎 when the tensile strength is 1520 MPa; and is 33.0 MPa √𝒎 when 2070 MPa.

(i) Assuming plane strain conditions, check whether or not such an increase in strength is
allowable for reducing structural weight?

Solution:

In order to solve this question, you need to think from a ‘materials selection’ as well as
‘structural design’ viewpoint. If you would like to reduce the cross-sectional area of a
component (and thus reduce its weight) for a given design load, it is often the case that
you would first think of a high strength material for designing and manufacturing such a
component. But, the problem with such an approach is that high strength materials are
likely to offer limited ductility (i.e. could lead to brittle fracture when the components
contain manufacturing defects). Now, you think from a ‘fracture mechanics’ viewpoint!

For this case, you have 0.45C-Ni-Cr-Mo with two different grades: (a) 66.0 MPa √𝑚 fracture
toughness and 1520 MPa tensile strength, and (b) 33.0 MPa √𝑚 fracture toughness and
2070 MPa tensile strength. Moreover, for ensuring safe service conditions and appropriate
inspection, the NDT technique selected is only sensitive to cracks larger than 3 mm. It
means a crack that is larger than 3 mm can be detectable.

For the combination of 66.0 MPa √𝑚 fracture toughness and 1520 MPa tensile strength,
the critical crack length can be estimated by assuming a very large plate with a centre-
crack (i.e. Y = 1.0) and remotely applied stress equal to half of the tensile strength (i.e.
0.5σu). Then, the critical centre-crack length is,
2
1 𝐾𝐼𝑐 2 1 66.0
𝑎𝑐 = ( ) = ( ) = 2.40 𝑚𝑚
𝜋 𝑌𝜎 𝜋 1.0 × 0.5 × 1520
So the critical crack length (from tip-tip) is 4.8 mm (i.e. > 3 mm). Therefore the NDT system
‘can detect’ a crack before it reaches a critical size (but the margin is very small; ideally you
would like to have a critical crack length that is much greater than the smallest crack the
NDT system can detect!). This material is ‘just OK’ for this design load level and the selected
NDT system.

In a similar way, for the combination of 33.0 MPa √𝑚 fracture toughness and 2070 MPa
tensile strength, the critical crack length can be estimated by assuming a very large plate
with a centre-crack (i.e. Y = 1.0) and remotely applied stress equal to half of the tensile
strength (i.e. 0.5σu). Then, the critical centre-crack length is,
2
1 𝐾𝐼𝑐 2 1 33.0
𝑎𝑐 = ( ) = ( ) = 0.324 𝑚𝑚
𝜋 𝑌𝜎 𝜋 1.0 × 0.5 × 2070
So the critical crack length (from tip-tip) is 0.648 mm (i.e. < 3 mm). Therefore the NDT
system ‘cannot detect’ critical cracks and thus not appropriate for ensuring safety and
avoiding catastrophic fracture. This material is ‘not suitable’ for this design load level and
the selected NDT system.

Page 14 of 20
 MACE31302: Aerospace Structures 3 Lecturer: Dr. KB Katnam

(ii) If 0.45C-Ni-Cr-Mo with a tensile strength of 2070 MPa is used, then what must be the
design stress level for detecting in-service cracks in the component using the same
inspection technique?

Solution:

If you have selected 0.45C-Ni-Cr-Mo with 2070 MPa tensile strength, then you can design
such a component by reducing the allowable stress levels that are less than 0.5σu (this is
the design stress level used in Part (i)). In order to calculate the maximum allowable stress
level, you need to calculate the remotely applied stress value that gives a critical crack
length that is detectable by the NDT system (i.e. 2ac = 3 mm). This means,

1 𝐾𝐼𝑐 2
𝑎𝑐 = ( ) ≥ 1.5 𝑚𝑚
𝜋 𝑌𝜎
33.0
⟹𝜎≤ = 480 𝑀𝑃𝑎
√𝜋 × (1.5/1000)

Therefore, the maximum allowable stress level is 480 MPa (which is less than 25% of the
tensile strength!).

Question 10*

Using the superposition principle, determine the stress intensity factors for a large plate with
a centre-crack subjected to:

Figure Q10: A large plate with a centre-crack subjected to: (a) internal pressure (b) rivet load

(i) Internal pressure acting on the crack faces as shown in Figure Q10 (a);

Solution:

First, you need to go through ‘slide 16’ and understand the concept of ‘superposition’. In
general, for a linear elastic structure, the combined load effects (i.e. stress tensor at a
material point for a set of external loads) caused by two or more loads are the sum (i.e.
superposition) of the individual load effects (i.e. the sum of stress tensors at the same
material point for the individual external loads). As crack tip stress state is governed by the

Page 15 of 20
MACE31302: Aerospace Structures 3 Lecturer: Dr. KB Katnam

stress intensity factor, you can apply the principle of superposition to obtain the stress
intensity factors for combined loading conditions. But, you need to be consistent with your
analysis (i.e. for the same mode and geometry). For example, you cannot superpose mode-I
and mode-II stress intensity factors (which is kind of adding a shear stress component to a
normal stress component!).

Figure Q10*: A large plate: (a) without a crack under remotely applied uniform tensile
stresses; (b) with a centre-crack under remotely applied uniform stress and uniform
compressive stresses closing the crack faces; (c) with a centre-crack under remotely
applied uniform tensile stresses; and (d) with a centre-crack under uniform compressive
stresses closing the crack faces.

Now, by applying the principle of superposition to the problem in Figure Q10 (a), you can in
a logical way define a combination of mode-I loading conditions to solve the problem. As
shown in Figure Q10*, first imagine a plate without a crack that is subjected to remotely
applied tensile stresses (i.e. Figure Q10*(a)); there is no crack and thus the stress intensity
factor, say KIa, is zero. The loading condition in Figure Q10*(a) is equivalent to Figure

Page 16 of 20
MACE31302: Aerospace Structures 3 Lecturer: Dr. KB Katnam

Q10*(b), where you have a centre-crack but with a uniform compressive stress to
completely close the two crack faces (i.e. no crack opening, thus the stress intensity factor
at the two crack tips, say KIb, is also zero). Now, you can split the loading condition in
Figure Q10*(b) as a sum of mode-I loading conditions as shown in Figure Q10*(c) and Figure
Q10*(d); and call the stress intensity factors for these mode-I conditions as KIc and KId.

From the principle of superposition (see that you have the same geometry and mode
conditions),

𝐾𝐼𝑏 = 𝐾𝐼𝑎 = 𝐾𝐼𝑐 + 𝐾𝐼𝑑 = 0

⟹ 𝐾𝐼𝑑 = −𝐾𝐼𝑐 = −𝜎√𝜋𝑎

Here, the stress intensity factor for a centre-crack in a very large plate (i.e. Y=1.0) is
employed. If you have a uniform tensile stress is acting on the crack faces, then you can
change the sign of stress and calculate the stress intensity factor. Hence, the stress
intensity factor for the mode-I loading shown in Figure Q10(a) is 𝜎√𝜋𝑎.

Figure Q10**: A large plate with a hole subjected to: (a) remotely applied uniform tensile
stresses at both ends; (b) remotely applied uniform stress at one end and a point load at

Page 17 of 20
 MACE31302: Aerospace Structures 3 Lecturer: Dr. KB Katnam

the hole; (c) remotely applied uniform tensile stress at one end and a point load at the
hole; and (d) a pair of point loads.

(ii) Rivet load as shown in Figure Q10 (b).

Solution:

Similar to the approach used in Part (i), you can apply the principle of superposition as
shown in Figure Q10**. The stress intensity factor for the loading condition shown in Figure
Q10**(a) is 𝜎√𝜋𝑎 (assuming Y=1.0); and the stress intensity factor for the loading
condition shown in Figure Q10**(d) can be obtained from the solution given for the
loading condition shown in Figure Q10(c) by setting x=0 (i.e. KIA = KIB = 𝑃/√𝜋𝑎, where ‘A’, ‘B’
are crack tip points as shown in Figure Q10(c)). The stress intensity factors for the loading
conditions shown in Figure Q10**(b) and (c) are equal, i.e. KIb = KIc).

Therefore, the stress intensity factor for the loading condition shown in Figure Q10(b) is,

𝐾𝐼𝑎 = 𝐾𝐼𝑏 + 𝐾𝐼𝑐 − 𝐾𝐼𝑑 = 𝜎√𝜋𝑎

𝑃
⟹ 𝐾𝐼𝑏 + 𝐾𝐼𝑐 = 𝜎√𝜋𝑎 + 𝐾𝐼𝑑 = 𝜎√𝜋𝑎 +
√𝜋𝑎
1 𝑃
⟹ 𝐾𝐼𝑏 = 𝐾𝐼𝑐 = (𝜎 √𝜋𝑎 + )
2 √𝜋𝑎
Here, the condition, KIb = KIc, is used.

Note:

Figure Q10 (c): A large plate with a centre-crack subjected to a pair of wedge loads

Use the following expressions available for the stress intensity factors for a large plate
with a centre-crack subjected to a pair of wedge forces as shown in Figure 10 (c). Where,
KIA is the stress intensity factor at the crack tip A; KIB is the stress intensity factor at the
crack tip B; x is the distance between the centre of the crack and the position of the
wedge loads; and P is the force per unit plate thickness.

𝑃 (𝑎 + 𝑥) 𝑃 (𝑎 − 𝑥)
𝐾𝐼𝐴 = √ 𝐾𝐼𝐵 = √
√𝜋𝑎 (𝑎 − 𝑥) √𝜋𝑎 (𝑎 + 𝑥)

Page 18 of 20
 MACE31302: Aerospace Structures 3 Lecturer: Dr. KB Katnam

Question 11*

Derive expressions for the plastic zone shapes (i.e. small scale yielding) by using the von
Mises stress criterion for the following crack conditions:

(i) Assume a mode-I crack tip with: (a) plane stress, and (b) plane strain conditions;

Solution:

First, you need to go through the solution strategy used in Question 1. The near-crack-tip
stresses are calculated by using the following equation for any point (r, θ):
𝐾𝐼 𝜃 𝜃 3𝜃
𝜎𝑥 = 𝑐𝑜𝑠 (1 − 𝑠𝑖𝑛 𝑠𝑖𝑛 )
√2𝜋𝑟 2 2 2
𝐾𝐼 𝜃 𝜃 3𝜃
𝜎𝑦 = 𝑐𝑜𝑠 (1 − 𝑠𝑖𝑛 𝑠𝑖𝑛 )
√2𝜋𝑟 2 2 2
𝐾𝐼 𝜃 𝜃 3𝜃
𝜏𝑥𝑦 = 𝑠𝑖𝑛 𝑐𝑜𝑠 𝑐𝑜𝑠
√2𝜋𝑟 2 2 2

Moreover, the shear stresses on the z-plane, i.e. τxz and τyz, are zero; and the normal stress,
i.e. σz, is zero for plane stress conditions and is 𝜈(𝜎𝑥 + 𝜎𝑦 ) for plane strain conditions.

(a)

Using the above equations for the near-crack-tip stress components and applying the
stress transformations (i.e. Mohr’s circle), you can show that the principal stresses for
plane stress conditions are,

𝐾𝐼 𝜃 𝜃
𝜎1 = 𝑐𝑜𝑠 [1 + 𝑠𝑖𝑛 ]
√2𝜋𝑟 2 2

𝐾𝐼 𝜃 𝜃
𝜎2 = 𝑐𝑜𝑠 [1 − 𝑠𝑖𝑛 ]
√2𝜋𝑟 2 2

𝜎3 = 0

Now, you can apply the von Mises criterion in terms of the three principal stresses, and
derive the expression for the shape of the plastic zone. The expression is,
𝐾𝐼2 3 2
𝑟𝑝 = 2 [1 + 2 𝑠𝑖𝑛 𝜃 + 𝑐𝑜𝑠 𝜃]
4𝜋𝜎𝑌

Here, 𝑟𝑝 is the yield zone for any given 𝜃, and σY is the material yield strength. If you plot
this function in Matlab, then you can see a shape similar to the one shown in ‘slide 19’.

(b)

Similar to Part (i)-(a), you can use the near-crack-tip stress components and applying the
stress transformations (i.e. Mohr’s circle), you can show that the principal stresses for
plane strain conditions are,

Page 19 of 20
 MACE31302: Aerospace Structures 3 Lecturer: Dr. KB Katnam

𝐾𝐼 𝜃 𝜃
𝜎1 = 𝑐𝑜𝑠 [1 + 𝑠𝑖𝑛 ]
√2𝜋𝑟 2 2

𝐾𝐼 𝜃 𝜃
𝜎2 = 𝑐𝑜𝑠 [1 − 𝑠𝑖𝑛 ]
√2𝜋𝑟 2 2

2𝜈𝐾𝐼 𝜃
𝜎3 = 𝑐𝑜𝑠
√2𝜋𝑟 2

Now, you can apply the von Mises criterion in terms of the three principal stresses, and
derive the expression for the shape of the plastic zone. The expression is,
𝐾𝐼2 3
𝑟𝑝 = [ 𝑠𝑖𝑛2 𝜃 +(1 − 2𝜈)2 (1 + cos 𝜃)]
4𝜋𝜎𝑌2 2

Here, 𝑟𝑝 is the yield zone for any given 𝜃, and σY is the material yield strength. If you plot
this function in Matlab, then you can see a shape similar to the one shown in ‘slide 19’.

Note: The plastic zone near the surface (i.e. plane stress conditions) is larger than the size
in the bulk (i.e. plane strain conditions).

(ii) Assume a mode-II crack tip with: (a) plane stress, and (b) plane strain conditions.

Solution:

Similar to Part (i), you can use the mode-II stress intensity factor and the expressions for
the near-crack-tip stresses and derive the expressions for the plastic zone shape when (a)
plane stress and (b) plane strain conditions exist. Do it yourself!

Page 20 of 20