Deformations and Stresses in Structures

We have learnt the basic concepts of stress and strain in materials under loadings. Material behaviors
in the elastic regime are essentially governed by the three basic material constants, namely Young’s and
shear moduli and Poisson’s ratio. When the stress-level within the material exceeds its yield strength,
it starts to experience plastic (permanent) deformation, which can be viewed as the failure criterion of
the material.

In this chapter, we will look into how stresses and strains are induced in two common structures in
engineering applications, namely the rod/truss and beam members. You will learn the relationships
between the applied loads and the stress/strain induced in the structures, which in turn can be used to
determine the nominal loads that the structures can withstand without failing. Such stress analyses are
important in order to assess the structural integrity of the structures under specified loading conditions.

Learning outcomes
For this session, you should be able to
• Calculate the stresses/strains and deformations in rod/truss that is under axial loads
• Recognize that a beam under lateral load will induce bending moment which manifests as
axial/normal stresses
• Recognize the key parameters that affect the strength of rod/truss and beam
• Design and size components for nominal load and safety factor for rod/truss and beam

Initial thoughts
In many engineering applications, slender structures are used in their designs due to their superior load-
bearing capacities. Two such basic structures are the truss (or rod) and beam. Geometrically, they look
identical, but the main difference lies on how the structure is being loaded.

For the rod/truss, it is loaded axially in the direction along the length of the member, whereas for the
beam, it is usually bearing lateral loads (forces that are perpendicular to the beam length). In the
following sections, we will study the load and deformation relationships for these two types of
structures.

Rod under Axial Loading

Consider a homogeneous rod BC of length L and uniform cross-section area

of A. When it is subjected to an axial load of P at the free end, the stress
and strain induced in the rod can be assumed to be uniform along the rod,
which are given by
𝑃
𝜎= , 𝑎𝑛𝑑 (1𝑎)
𝐴
𝛿
𝜖= (1𝑏)
𝐿

where 𝛿 denotes the change in the length of the rod.

Figure 1: Deformation of axially loaded rod.

1
Now using the stress/strain relation in the elastic regime, one can derive the force/deformation relation
as follows
𝑃 𝛿
𝜎 = 𝐸𝜖 ⇒ =𝐸 (2)
𝐴 𝐿
which can be rearranged as
𝐸𝐴
𝑃 = ( ) 𝛿, 𝑜𝑟 (3𝑎)
𝐿
𝑃𝐿
𝛿=( ) (3𝑏)
𝐸𝐴

One should notice that the first expression in Eq. (3a) resembles the Hooke’s law for an elastic spring,
in which the spring stiffness is then given by
𝐸𝐴
𝑘𝑠 = (4)
𝐿

From Eq. (4), it is obvious that a rod made with higher Young’s modulus, such as steel over aluminium,
is stiffer. And a rod with larger cross-section and/or shorter length will also be more resistant to
deformation when subjected to identical load.

𝑘𝑠

Figure 2: Equivalent of rod under axial load to linear spring.

Suppose the rod is loaded at multiple points, and/or if it is a composite rod that is made up of several
portions of various cross-sections, lengths, and may be different materials, as shown in Fig. 3(a).

Figure 3: Composite rod under multiple axial loads.

2
In this case, we have to divide the rod into separate parts, such that each portion will have identical
cross-section, same material, and exerted by the same load. By doing so, Eq. (3b) can be used to
calculate the elongation of each portion, and the total deformation is simply their sums, i.e.
𝑃𝑖 𝐿𝑖
𝛿𝑡𝑜𝑡𝑎𝑙 = ∑ ( ) (5)
𝐸𝑖 𝐴𝑖
𝑖=𝑛

where n is number of sub-sections identified. For the example in Fig. 3, we can identify three separate
portions, namely: AB, BC, and CD.
Note that the axial load in each portion is determined by their respective free-body diagram, as shown
in Fig. 3(c). And they are given by
−𝑃1 + 𝑃𝐵 − 𝑃𝐶 + 𝑃𝐷 = 0 ⇒ 𝑃1 = 𝑃𝐵 − 𝑃𝐶 + 𝑃𝐷
−𝑃2 − 𝑃𝐶 + 𝑃𝐷 = 0 ⇒ 𝑃2 = 𝑃𝐷 − 𝑃𝐶
−𝑃3 + 𝑃𝐷 = 0 ⇒ 𝑃3 = 𝑃𝐷

Truss under Axial Loading

A truss is an assembly of straight members connected by pin (revolute) joints to form a rigid structure.
There are several important applications of truss structures in engineering, especially in the
constructions of large structures like the bridge, roof-top, and offshore jack-up platform.

Figure 4: Truss structures in the constructions of bridge, rooftop design, and offshore jack-up platform.

The individual truss member is allowed to rotate about the pin joints, and as such the joints will not
resist moments (torques). Another important feature of truss is that it is a two-force member, where the
loadings only occur at the two ends of the member at its joints. Hence, due to such loading and/or
supporting conditions, truss members only experience axial load along its length. To illustrate this fact,
let us consider a truss member that is subjected to arbitrary loadings of 𝐹1 and 𝐹2 at its two ends, as
shown in Fig. 5 (a).

𝐹1 𝐹 𝐹

𝐹2 𝐹 𝐹
(a) (b)
Figure 5: A truss member under loading at their two-ends.

3
Now in order for this truss member to be in static equilibrium state, the moment equations about the
two pivot ends must be equal to zeros. And these conditions can only be satisfied if and only if the two
forces are pointing in the direction along the truss member. In addition, the sum of the forces along the
axial direction must also be zeros in order to satisfy the force equilibrium condition. Hence in summary,
the resultant loads at the two ends of the truss member can either be in tensile or compressive forces, as
shown in Fig. 5(b).

Since truss structures are also subjected uniaxial loading condition, hence Eqs. (1) to (4) that were
derived for the rod can also be used in the truss analysis. Note also that there are also other structures
that resemble this mode of loading, such as a column (which can take compressive load only, and
provided the column does not buckle), and a cable (which can take tensile load only).

Sample Problem 1

The rigid bar BDE is supported by two links AB and CD. Link
AB is made of aluminium (E = 70 GPa) and has a cross-
sectional area of 500 mm2; link CD is made of steel (E = 200
GPa) and has a cross-sectional area of 600 mm2. For the 30 kN
force applied at E, determine the deflections of B, D, and E.

SOLUTION:

First, it should be recognized that links AB and CD are truss members. This suggests that the formulas
derived above can be used to analyze this problem. Of particular relevant is Eq. 3(b), which solves for
the deformation of a rod under uniaxial load. This in turns asks for the axial loads to be determined.

To solve for the axial forces, we can use the static equilibrium approach. Consider the FBD of the rigid
bar, where 𝐹𝐴𝐵 and 𝐹𝐶𝐷 are the internal forces due to the links, which must be acting in the directions
of the axes of the links. In the following FBD, they are drawn pointing away from the joints and into
the members. With this definition, the two links are assumed to be in tension. We may be wrong about
the force directions, but they will be corrected by solving the static equations. We need to write two
independent static equations to solve for the two unknown forces, either from force equilibrium in the
vertical direction, or moment equations about two convenient points.

Here, we use the following two moment equations, i.e.

+↺ ∑ 𝑀𝐵 = 0: − (30𝑘𝑁)(0.6) + 𝐹𝐶𝐷 (0.2) = 0

⇒ 𝐹𝐶𝐷 = 90 𝑘𝑁 (𝑡𝑒𝑛𝑠𝑖𝑜𝑛)

+↺ ∑ 𝑀𝐷 = 0: − (30𝑘𝑁)(0.4) − 𝐹𝐴𝐵 (0.2) = 0

⇒ 𝐹𝐴𝐵 = −60 𝑘𝑁 (𝑐𝑜𝑚𝑝𝑟𝑒𝑠𝑠𝑖𝑜𝑛)

With the internal forces in the links and their properties given in the question, the deformations of the
links can be calculated using Eq. 3(b) as follows:

4
 𝐹𝐴𝐵 𝐿𝐴𝐵 (−60 × 103 )(0.3)
𝛿𝐵 = =
𝐴𝐴𝐵 𝐸𝐴𝑙𝑢𝑚 (500 × 10−6 )(70 × 109 )
⇒ 𝜹𝑩 = −𝟓𝟏𝟒 × 𝟏𝟎−𝟔 𝒎

𝐹𝐶𝐷 𝐿𝐶𝐷 (90 × 103 )(0.4)

𝛿𝐷 = =
𝐴𝐶𝐷 𝐸𝑆𝑡𝑒𝑒𝑙 (600 × 10−6 )(200 × 109 )
⇒ 𝜹𝑫 = 𝟑𝟎𝟎 × 𝟏𝟎−𝟔 𝒎

The deflection at E is determined from the geometrical viewpoint. Recall that rod BDE is said to be
rigid, which means that it does not deform, and hence remains straight upon loading. Hence, the
deformations profile at the three points forms similar triangles as shown in the following diagram.

Hence, using the similar triangle concept, we have

𝐵𝐵′ 𝐵𝐻 0.514 200 − 𝑥
= ⇒ = ⇒ 𝑥 = 73.7 𝑚𝑚
𝐷𝐷′ 𝐻𝐷 0.300 𝑥

And hence the deflection at E is

𝐸𝐸′ 𝐻𝐸 𝛿𝐸 400 + 73.7
= ⇒ = ⇒ 𝜹𝑬 = 𝟏. 𝟗𝟑 𝒎𝒎
𝐷𝐷′ 𝐻𝐷 0.300 73.7

Beam under Lateral loads

Now let us move onto another scenario where the slender rod is loaded laterally, such as the cantilever
beam shown in Fig. 6. In this case, this structure is usually referred to as a beam. A beam is a structural
element that primarily resists loads applied laterally to the beam’s axis (also referred to as the
longitudinal axis). The external forces acting on the beam produces only internal shear force and
bending moment at any cut section along the beam (see FBD in Fig. 6 (b)), but zero internal force in
the longitudinal direction (normal to the cut section).

Figure 6: A simply supported beam bends under a uniformly distributed load.

For a slender beam structure, its load-carrying and deflection characteristics are known to be governed
by the classical beam theory, also known as the Euler-Bernoulli beam theory, which is most appropriate
for thin slender beam. The following sections show the derivation of the core equation in this topic on
beam, which is the relationship between the bending stress induced by the bending moment that comes
from the external loadings on the beam. Note that the important equations are highlighted in the red
brackets, which will be tested in this module.

5
Sign convention

Let us begin by defining the sign convention adopted in this study, as shown in Fig. 7. This sign
convention is consistent with the right-hand rule, which states that z-direction is defined by the cross-
product of the x-axis into y-axis. And positive shear force and moment on the positive face of the
isolated element (face with normal direction in the x-direction) are in the positive y- and z- directions,
respectively. [Link]

(Lateral)

(Longitudinal)

Figure 7: Sign convention for shear force and moment.

Based on this sign convection, positive moment will cause the beam to bend with concavity on the top
surface, and negative moment otherwise. A simple way to remember this convection is that positive
moment makes the beam “smile”, while negative moment makes it frown.

Note however that different sign conventions may be used by different authors, especially for the shear
force direction. This is simply because the engineering communities have not come to an agreement on
the “standard sign” convention on beam bending. So beware of such differences. Nevertheless, as long
as you stick to one sign convection throughout the entire analysis, you should get consistent results.

Neutral axis and surface

An important observation associated with this sign convection is that for positive bending moment, the
beam will experience contraction and elongation at the top and bottom surfaces, respectively. This
means that the lower part of the beam is elongated while the upper part is being compressed, as shown
in Fig. 8. And there exist a surface along the beam axis (line DE in Fig 8 (a)) that divides the beam into
tensile and compressive states. This surface is called the neutral surface. This surface cuts the transverse
cross-section of the beam along a straight line called the neutral axis, as illustrated in Fig. 8(b). This
line divides the cross-section into the compressive and tensile zones above and below it, respectively.
For convenience, it is common to let the origin of the coordinates to fall on the neutral axis. With this,
the y- coordinate denotes the distant from the neutral axis.

(a)

(b)

Figure 8: Bending profile of a small section of the beam.

6
Linear strain and stress variation over the thickness

As noted in Fig. 8(a), the arc length of the neutral axis DE and arc JK are given respectively by

𝐿 = 𝜌𝜃, 𝑎𝑛𝑑 𝐿′ = (𝜌 − 𝑦)𝜃 (6.1)

The difference in the two lengths corresponds to the deformation of arc JK, whose original length is
identical to the neutral axis. Hence, the longitudinal strain (along the axial direction of the beam) is

𝐿′ − 𝐿 (𝜌 − 𝑦)𝜃 − 𝜌𝜃 𝑦
𝜀𝑥 = = =− (6.2)
𝐿 𝜌𝜃 𝜌

Equation (6.2) clearly shows that the longitudinal normal strain 𝜀𝑥 varies linearly with the distance y
from the neutral surface.

Suppose the beam is made of material with Young’s modulus

E, then the corresponding longitudinal normal stresses will
also takes on a linear profile as shown in Fig. (9), which is
given by

𝐸
𝜎𝑥 = 𝐸𝜀𝑥 = − ( ) 𝑦 (7.1)
𝜌

At the top surface where y = c, it attains the maximum absolute

stress value of 𝜎𝑚 , that is
Figure 9: Linear variation of bending
stresses across the beam cross-section. 𝐸
𝜎𝑚 = ( ) 𝑐 (7.2)
𝜌
Finally by taking the ratio of Eq (7.1) and (7.2), we get
𝜎𝑥 𝑦 𝜎𝑥 𝜎𝑚
=− ⇒ =− (7.3)
𝜎𝑚 𝑐 𝑦 𝑐

And hence the linear stress profile can be defined as

𝑦
𝜎𝑥 = − 𝜎𝑚 (7.4)
𝑐

Neutral axis passes through the centroid of the section

Recall that there is zero internal longitudinal force acting on the cut-section of the beam simply because
it is only subjected to lateral loads. This also means that the integration of the normal stresses over the
cross-sectional area must be equal to zero, that is

𝑦 𝜎𝑚
∫ 𝜎𝑥 𝑑𝐴 = ∫ − 𝜎𝑚 𝑑𝐴 = − ∫ 𝑦𝑑𝐴 = 0 (8.1)
𝑐 𝑐
𝐴 𝐴 𝐴

𝜎𝑚
which leads to the relation (since 𝑐
cannot be zero)

∫ 𝑦𝑑𝐴 = 0 (8.2)
𝐴

7
Equation (8.2) essentially means that the neutral axis must pass through the centroid of the cross-section
(center of the area). To be more explicit, the centroid of an area with respect to an arbitrary coordinate
system is given by

𝑥̅ = ∫ 𝑥𝑑𝐴 ; 𝑎𝑛𝑑 𝑦̅ = ∫ 𝑦𝑑𝐴 (8.3)

𝐴 𝐴

Now if the coordinate system is placed at the centroid of the area,

then 𝑥̅ = 𝑦̅ = 0. This means that

∫ 𝑥𝑑𝐴 = ∫ 𝑦𝑑𝐴 = 0
𝐴 𝐴

which proofs the statement given in Eq. (8.2). Note also that for area
that is symmetrical about the y-axis, ∫𝐴 𝑥𝑑𝐴 = 0 is automatically
satisfied.

Bending stress and moment relation

As mentioned above, it is the normal stress distributed across the thickness that gives rise to the bending
moment at cut-section. Hence, the bending moment M is related to the bending stress by the following
equivalence, as shown in Fig. 10.

Figure 10: Bending stress give rise to an effective bending moment M at the cross-section surface.

Taking moment about the neutral axis, we have

𝑀 = ∫ −𝑦𝜎𝑥 𝑑𝐴 (9.1)
𝐴

Then substituting the linear stress profile derived in Eq. (7.4) into (9.1) gives
𝑦 𝜎𝑚
𝑀 = ∫ −𝑦 (− 𝜎𝑚 ) 𝑑𝐴 = ∫ 𝑦 2 𝑑𝐴 (9.2)
𝐴 𝑐 𝑐 𝐴

where the integral 𝐼 = ∫𝐴 𝑦 2 𝑑𝐴 is called the centroidal moment of inertia (or second moment) of the
cross section with respect to the neutral axis. We will discuss more about this geometrical quantity later.

Having to define the centroidal moment of inertia and using Eq. (7.3), Eq. (9.2) can be rewritten as

𝜎𝑚 𝜎𝑥 𝑀
𝑀= 𝐼 = − 𝐼 ⇒ 𝜎𝑥 = − 𝑦 (9.3)
𝑐 𝑦 𝐼
which expresses the normal stress 𝜎𝑥 at any distance y from the neutral axis in terms of the centroidal
moment of inertia I and to the internal bending moment M induced on the given cut-section.

8
Centroidal moment of inertia of area

From Eq. (9.3), it is clear that the moment of inertia of the cross-section area plays a vital role in the
structural performance of beam. It is a geometrical property of the beam cross-section area that provides
insight into a beam’s resistance to bending.

Let us evaluate this quantity for the rectangular area shown in Fig. 11.
Consider an elementary area 𝑑𝐴 = 𝑏𝑑𝑦 that is at y distance from the z-
axis. The moment of inertia of this strip about that axis is
𝑑𝐼 = 𝑦 2 𝑑𝐴 = 𝑦 2 𝑏𝑑𝑦

Integrating from 𝑦 = −ℎ/2 to 𝑦 = +ℎ/2, we have

+ℎ/2 +ℎ/2 +ℎ/2

2
𝑦3 1
𝐼=∫ 𝑑𝐼 = ∫ 𝑦 𝑏𝑑𝑦 = 𝑏 [ ] = 𝑏ℎ3
−ℎ/2 −ℎ/2 3 −ℎ/2 12

Figure 11: Moment inertia of a rectangular cross-section.

It should be emphasized that the moment of inertia is not only dependent on the dimensions of the cross-
section area, its orientation also matters. Take for example the rectangle area in Fig. 11. If it is rotated
by 90 degree, such that the width and thickness become b and h, respectively. In this case, the centroidal
1
moment of inertia becomes 𝐼 ′ = 12 ℎ𝑏 3, and the ratio of the centroidal moment of areas for the two
different orientations is 𝐼 ⁄𝐼′ = (ℎ⁄𝑏)2 . And for the case when h = 2b, this means that 𝐼 = 4𝐼′, indicating
that a rectangular beam is more resistant to bending moment if it is oriented with larger thickness than
its width.

Note that you are not required to perform the second moment of area integration. For several of the
simple geometrical shapes, the formulas can be found here and some of the common ones are tabulated
in the following figure.

1 1
𝐼𝑥 = 𝑏ℎ3 𝐼𝑥 = 𝑏ℎ3
12 36

1 1
𝐼𝑥 = 𝜋𝑟 4 𝐼𝑥 = 𝜋𝑎𝑏 3
4 4

Figure 12: Centroidal moment inertia for some of the common geometrical shapes.

9
Structural analysis of a cantilever beam with rectangular cross-section
Now let us consider a beam with a rectangular profile given by Fig. 11. Using Eq. (9.3), the maximum
bending stress can be calculated as

𝜎𝑚 1 6𝑀
𝑀= ( 𝑏ℎ3 ) ⇒ 𝜎𝑚 = 2 (10)
ℎ⁄2 12 𝑏ℎ

From Eq. (10), it can be seen that for a rectangular profiled beam, the maximum stress varies inverse
linearly with the width of the beam, but it scales down with a power of two with increasing thickness.
This suggests that changing the thickness of the beam will be significantly more effective in reducing
the stresses in the beam, as compared to changing its width.

Yield and failure strength of materials

Recall from the previous session when we look at the uniaxial tensile test. We identify then that material
starts to suffer permanent damage when the stress level goes beyond its yield strength 𝜎𝑦 . Similarly, a
beam under bending starts to fail when 𝜎𝑚 > 𝜎𝑦 , which is expected to be at either the top or bottom
surface of the beam, and at the section where the internal moment is the largest.

Let us revisit the cantilever beam given in Fig. 6. For this example, the maximum bending moment
occurs at the fixed end, which is given by M = PL, with P being the lateral load at the free end, and L
is the length of the beam. Substituting this moment into Eq. (10), we can determine the maximum
normal stress to be

6𝑀 6𝑃𝐿
𝜎𝑚 = = (11)
𝑏ℎ2 𝑏ℎ2

Notice that this expression is independent on the material

used, that is an aluminum or steel beam will experience the
same stress level. Suppose we let the beam dimensions to
be L = 100 mm, h = 5 mm, and b = 2 mm, and the load P =
10 N. Here, we get 𝜎𝑚 = 120 𝑀𝑃𝑎. Comparing this value
with some of the yield strength of the materials tabulated
in the table, we may conclude that only the steel and
aluminum alloys should be selected for this purpose.

On the other hand, if the material has been decided, say copper, then the question poses may become:
what is the maximum lateral load the beam can withstand without yielding? Or how can we alter the
beam dimensions so that the beam can support the load?

Safety factor in product designs

In many engineering applications, the material in use cannot afford to suffer any form of permanent
damage during operations. Hence, it is common to use the yield strength of the material as a design
specification for stress failure analysis. However, due to the uncertainty in the loading conditions, you
may have to incorporate a safety factor (denoted by S.F.) in your stress calculations. Safety factor
essentially reduces the maximum allowable stress level that can be induced in the material due to the
nominal applied loads. In other words, the maximum allowable stress level is scaled by the safety factor
of the yield strength of the materials as follows
𝜎𝑦
𝜎𝑚 < 𝜎𝑎𝑙𝑙𝑜𝑤𝑎𝑏𝑙𝑒 = (12)
𝑆. 𝐹.

10
It can also be viewed as the margin at which the stress level in the material is from its yield strength.
However, it should be highlighted that the safety factor is a double sword. High safety factor can usually
give you safer products, but may add towards the products cost and weigh.

Now going back to the cantilever example above, where the maximum bending stress is calculated to
be 120 MPa. Suppose a safety factor of 3 is required for this product. In this case, ASTM A36 steel no
longer can be regarded as a candidate for this product since the maximum allowable stress is reduced
to 83 MPa.

Sample Problem 2
The rectangular hollow tube shown is extruded from an aluminium alloy for which 𝜎𝑦 = 400 𝑀𝑃𝑎,
and E = 70 GPa. Neglecting the effect of fillets, determine the allowable bending moment M for which
the factor of safety will be 2.0.

SOLUTION:

First, for a safety factor of S.F. = 2.0 and using the yield strength 𝜎𝑦 = 400 𝑀𝑃𝑎 as the failure criterion,
the maximum allowable stress is calculated as
𝜎𝑦
𝜎𝑎𝑙𝑙𝑜𝑤𝑎𝑏𝑙𝑒 = = 200 𝑀𝑃𝑎
𝑆. 𝐹

For this beam cross-section, it is easy to see that the neutral axis is at the mid-plane. Hence, the
magnitude of the bending stresses at either the top and bottom surfaces is given by

𝑀 ℎ
𝜎𝑚 = ( )
𝐼𝑧 2

where the centroidal moment of inertia for this hollow tube can be derived from the difference of
moment inertias of the two rectangular profiles, that is

1 1
𝐼𝑧 = (3.25)(5)3 − (2.75)(4.5)3
12 12
= 12.97 𝑐𝑚4 = 12.97 × 10−8 𝑚4

Note that this direct subtraction (or summation) of moment of inertias is possible if and only if the
neutral axes of the areas involved coincide. For cases in which this is not satisfied, you will have to use
the parallel-axis theorem (to be taught in ME2112) to perform the calculations.

11
Finally substituting 𝜎𝑚 = 𝜎𝑎𝑙𝑙𝑜𝑤𝑎𝑏𝑙𝑒 = 200 𝑀𝑃𝑎, ℎ = 0.05 𝑚, and 𝐼𝑧 = 12.97 × 10−8 𝑚4 into the
above expression, we have the allowable bending moment as

2𝐼𝑧 2(12.97 × 10−8 )

𝑀= 𝜎𝑚 = (200 × 106 ) = 𝟏𝟎𝟒𝟎 𝑵𝒎
ℎ (0.05)

Summary
This chapter begins with deriving Eq.(3) that expresses the deformation of a rod under axial loadings
in terms of the loads, geometry of the rod (length and cross-sectional area), and Young’s modulus of
the material. Based on this expression, a rod under axial loading behaves like a linear spring whose
spring constant is given by Eq. (4).

Next, we introduce the beam structure which is a slender rod that takes on lateral loads. Based on the
static analysis, it is found that there will only be internal shear force and moment induced at a given
cut-section of the beam. It is important to realize that the internal moment is essentially due to the sum
of the moment of the internal normal stresses induced on the cross-sectional area. And for slender beam,
the normal stress distribution can be assumed to be varying linearly across the thickness of the beam,
with one side in tensile and the other in compression. The line that divides the cross-sectional area into
the two different stress states is called the neutral axis, where material along this line does not deform.

Now due to the fact that the internal force on the cross-section is zero, this also means that the neutral
axis must pass through the centroid of the area. And by equating the internal moment to the sum of
moments due to the normal stresses, we derived an important relation, i.e. Eq. (9.3), that relates the
normal bending stress to the internal moment, which also depends on the lateral distant of the stress
point from the neutral axis, and a geometrical parameter called the centroidal moment of inertia (or 2nd
moment of area about the neutral axis). Based on this expression, it is clear that the maximum normal
stress occurs at the point on the cross-section, which experiences the largest internal bending moment,
and is furthest away from the neutral axis, either at the top or bottom surface of the beam.

Finally having to identify the maximum bending stress, we can use this value to quantify whether the
beam is strong enough to withstand the applied loads. However, to enhance the safety of the structure,
a factor of safety is often used in the design specification. It essentially reduces the maximum allowable
stress that can be induced in the structure, which in turns limit the external loads that can be applied on
the structure.

Supplementary Reading
F.P. Beer and E.R. Johnston, “Mechanics of Materials” Chapters 2.8, and 4.1-4.4.

12