Mechanics of Solids

Chapter 6
Pure Bending and Bending with Axial Forces
• When a segment of a beam is in equilibrium under the action of
bending moments alone, such a condition is referred to as ‘pure
bending’ or ‘flexure’.
• A cantilever loaded with a concentrated moment at the end, or a
segment of a beam between the concentrated forces, are examples of
pure bending.
• Usually the bending stresses in slender beams are dominant.
• Some beams by virtue of their slenderness or lack of lateral support
may become unstable under an applied load and may buckle laterally
and collapse.
 Part A- Bending of Beams with Symmetric Cross
Sections
The Basic Kinematic Assumption
1. A plausible deformation assumption reduce the internally statically
indeterminate problem to a determinate one.
2. The deformations causing strains be related to stresses through the
appropriate stress-strain relations.
3. The equilibrium requirements of external and internal forces must
be met.
• Consider a horizontal prismatic beam having a cross section with a
vertical axis of symmetry, Fig. 1(a). Consider a typical element of the
beam between two planes perpendicular to the beam axis.
• In side view, such an element is identified in the figure as 𝑎𝑏𝑐𝑑.

Fig. 1: Assumed behavior of elastic beam in bending

• When such a beam is subjected to equal end moments 𝑀𝑧 acting
around the 𝑧 axis, Fig. 1(b), this beam bends in the plane of
symmetry, and the planes initially perpendicular to the beam axis
slightly tilt. Nevertheless, the lines such as 𝑎𝑑 and 𝑏𝑐 becoming 𝑎′𝑑′
and 𝑏′𝑐′ remain straight.
• The fundamental hypothesis of the flexure theory: plane sections
through a beam taken normal to its axis remain plane after the beam
is subjected to bending.
• This assumption is generally applicable with a high degree of accuracy
whether the material behaves elastically or plastically, providing the
depth of the beam is small in relation to its span.
• In pure bending of a prismatic beam, the beam axis deforms into a
part of a circle of radius 𝜌,(rho) as shown in Fig. 1(b). For an element
defined by an infinitesimal angle 𝑑𝜃, the fiber length 𝑒𝑓 of the beam
axis is given as 𝑑𝑠 = 𝜌 𝑑𝜃
 𝑑𝜃 1
= =𝜅
𝑑𝑠 𝜌
• The reciprocal of 𝜌 defines the axis curvature 𝜅 (kappa). In pure
bending of prismatic beams, both 𝜌 and 𝜅 are constant.
• The difference between fiber lengths 𝑔ℎ and 𝑒𝑓 identified here as 𝑑𝑢
can be expressed as follows
𝑑 𝑢 = 𝜌 − 𝑦 𝑑𝜃 − 𝜌 𝑑𝜃 = −𝑦 𝑑𝜃
• By dividing above equation by 𝑑𝑠, the last term becomes 𝜅.
• Since the deflection and rotations of the beam axis are very small, the
cosines of angles involved in making the projections of 𝑑 𝑢 and 𝑑𝑠
onto the horizontal axis are very nearly unity. Therefore, it is possible
to replace 𝑑 𝑢 by 𝑑𝑢, the axial fiber deformation, and 𝑑𝑠 by 𝑑𝑥.
• The normal stress, 𝜀𝑥 = −𝜅 𝑦
 The Elastic Flexure Formula
• The normal longitudinal stress,
𝜎𝑥 = 𝐸 𝜀𝑥 = −𝐸 𝜅 𝑦
• Using nontrivial equations of equilibrium:
𝐹𝑥 = 0 or 𝜎𝑥 𝑑𝐴 = −𝐸 𝜅 𝑦 𝑑𝐴 = 0
• The summation of the infinitesimal forces must be carried out over the
entire cross-sectional area 𝐴 of the beam.
• By definition, this integral 𝑦 𝑑𝐴 = 𝑦𝐴, where 𝑦 is the distance from the
origin to the centroid of an area 𝐴. Therefore, the 𝑧 axis must pass through
the centroid of a section.
• Along the 𝑧 axis so chosen, both the normal strain 𝜀𝑥 and the normal stress
𝜎𝑥 are zero. This axis is referred to as the neutral axis of a beam.
• Linear variation in strain is schematically shown in Fig. 1(c). The
corresponding elastic stress distribution is shown in Fig. 1(d).
• Both the absolute maximum strain 𝜀𝑥 and the absolute maximum
stress 𝜎𝑚𝑎𝑥 occurs at the largest value of 𝑦.
• Alternative representations of the elastic bending stress distribution
in a beam are illustrated in Fig. 2. in three-dimensional problem, the
locus of a neutral axis along a length of a beam defines the neutral
surface, Fig. 3

Fig.2: Alternative representations of bending moment

 Fig. 3: Neutral surface

• The second equation of equilibrium: the sum of the externally-applied

and the internal resisting moments must be in equilibrium.

Fig. 4: Segment of
a beam in pure
flexure
• For the beam segment in Fig. 4(a), this yields

• A negative sign in front of the integral is necessary because the

compressive stresses 𝜎𝑥 develop a counterclockwise moment around
the 𝑧 axis.
• The tensile stresses below the normal axis, where 𝑦’s have a negative
sign, contribute to this moment in the same manner.
𝑀𝑧 = 𝐸𝜅 𝑦 2 𝑑𝐴

• The last integral, depending only on the geometrical properties of a

cross-sectional area, is called the rectangular moment of inertia or
second moment of the area 𝐴 and will be designated by 𝐼.
• It must be found with respect to the cross section’s neutral
(centroidal) axis. We can identify it with a subscript corresponding to
such an axis. For the case considered, this subscript is 𝑧, i.e.,
𝑀𝑧
𝐼𝑧 = 𝑦2 𝑑𝐴 and 𝜅 =
𝐸 𝐼𝑧
• This is the basic relation giving the curvature of an elastic beam
subjected to a specific moment. The elastic flexure formula for
beams:
𝑀𝑧
𝜎𝑥 = − 𝑦
𝐼𝑧
• The derivation of this formula was carried out with the coordinate
axis shown in Fig. 5(a). If the derivation for a member having a doubly
symmetric cross section were done with the coordinates shown in Fig.
𝑀𝑧
5(b), the longitudinal stress 𝜎𝑥 = + 𝑦
𝐼𝑧
• Here the sign reversal is necessary because a positive 𝑀𝑦 causes
tensile stresses for positive 𝑧’s.

Fig. 5: Definitions of positive moments

𝑀𝑐
𝜎𝑚𝑎𝑥 =
𝐼
• Due to Poisson's ratio, the compressed zone of a beam expands
laterally; the tensile zone contracts.
• The strains in the 𝑦 and 𝑧 direction are 𝜀𝑦 = 𝜀𝑧 = −𝜈𝜀𝑥 = −𝜈 𝜎𝑥 𝐸.
• Poisson’s effect, deforms the neutral axis into a curve of large radius;
and the neutral surface becomes curved in two opposite directions.

Procedure Summary and Extensions

• Three basic concepts:
1. Equilibrium conditions (statics) are used for determining the
internal resisting bending moment at a section.
2. Geometry of deformation (kinematics) is used by assuming that
plane sections through a beam remains plane after deformation.
The normal strains along a beam section vary linearly from the
neutral axis.
3. Properties of materials (constitutive relations) in the form of
Hooke’s law are assumed to apply to the longitudinal normal
strains. Poisson’s effect of transverse contraction and expansion is
neglected.
• Consider the beam having the cross-section shown in Fig. 6(a). This
beam is made up of two materials, 1 and 2,bounded together at their
interface.
• The elastic moduli for the two materials are 𝐸1 and 𝐸2 , assuming
𝐸2 > 𝐸1 .

Fig. 6: Beam of two elastic materials in bending where 𝐸2 > 𝐸1

• When a composite beam is bent, as for a beam of one material, the

strains vary linearly as shown in Fig. 6(b).
• The longitudinal stresses depend on the elastic moduli and are shown
in Fig. 6(c).
• At the interface between the two materials, whereas the strain for
both materials is the same, the stresses are different and depend on
the magnitudes of 𝐸1 and 𝐸2 .
• Let 𝐸𝑖 be such an elastic modulus for the ith material in a composite
cross section.
𝜎𝑥 = 𝐸𝑖 𝜀𝑥 = −𝐸𝑖 𝜅 𝑦
• From Fig. 6(a), 𝑦 = 𝑦𝑏 − 𝑦𝑏 . In this relation 𝑦𝑏 is arbitrarily measured
from the bottom of the section, and 𝑦𝑏 locates the neutral axis as
shown.
• Since for pure bending the force 𝐹𝑥 at a section in the 𝑥 direction
must vanish,
 𝐹𝑥 = 𝜎𝑥 𝑑𝐴 = −𝜅 𝐸𝑖 𝑦 𝑑𝐴 = 0
−𝜅 𝐸𝑖 𝑦𝑏 𝑑𝐴 + 𝜅𝑦𝑏 𝐸𝑖 𝑑𝐴 = 0 and
𝐸𝑖 𝑦𝑏 𝑑𝐴
𝑦𝑏 =
𝐸𝑖 𝑑𝐴
• The equation defines the modulus-weighted centroid and locates the
neutral axis.
• For this case plane sections through a beam remain precisely plane
after bending. However, in usual applications, per Saint-Venant’s
principle, it is generally assumed that the stresses, at a distance about
equal to the depth of a member away from the applied moment, are
essentially uniform.
• The local stresses at points of force application or change in cross
section are calculate using stress concentration factors.
• In Fig. 6(d), the tension 𝑇 is equal to the compression 𝐶, and the 𝑇- 𝐶
couple is equal to the moment 𝑀𝑧 .

Computation of the Moment of Inertia

• The rectangular moment of inertia (𝐼) must be computed around the
neutral axis, which passes through the centroid of the cross-sectional
area.
• For symmetric cross sections, the neutral axis is perpendicular to the
axis of symmetry. The moment of inertia around such an axis is either
a maximum or minimum, and for that reason, this axis is one of the
principal axis for an area.
• To find 𝐼 for an area composed of several simple shapes, the parallel
axis theorem (transfer formula) is necessary.
 Fig. 7: Area for deriving the parallel-axis theorem

• Consider that the area 𝐴 shown in Fig. 7 is a part of a complex area of

a cross section of a beam in flexure. The centroidal axis 𝑧𝑐 for this
area is at a distance 𝑑𝑧 from the centroidal 𝑧 axis for the whole cross-
sectional area.
• The moment of inertia 𝐼𝑧𝑐 of the area 𝐴 around its 𝑧𝑐 axis is
𝐼𝑍𝑐 = 𝑦𝑐2 𝑑𝐴
• The moment of inertia 𝐼𝑧 of the same area 𝐴 around the 𝑧 axis is
𝐼𝑧 = 𝑦𝑐 + 𝑑𝑧 2 𝑑𝐴 = 𝑦𝑐2 𝑑𝐴 + 2𝑑𝑧 𝑦𝑐 𝑑𝐴 + 𝑑𝑧2 𝑑𝐴
• Here the first integral is equal to 𝐼𝑍𝑐 , the second integral vanishes as
𝑦𝑐 passes through the centroid of 𝐴, and the last integral reduces to
𝐴𝑑𝑧2 . Hence,
𝐼𝑧 = 𝐼𝑍𝑐 + 𝐴𝑑𝑧2
• Parallel axis theorem: The moment of inertia of an area around any
axis is equal to the moment of inertia of the same area around a
parallel axis passing through the area’s centroid, plus the product of
the same area and the square of the distance between the two axis.
𝐼𝑧 (whole section) = 𝐼𝑍𝑐 + 𝐴𝑑𝑧2
 Applications of The Flexure Formula
• Both 𝐼 and 𝑐 are constants for a given section of a beam. Hence, 𝐼 𝑐
is a constant.
• Since this ratio is only a function of the cross-sectional dimensions of
a beam, it can be uniquely determined for any cross-sectional area.
This ratio is called the ‘elastic section modulus’ of a section and will
be designated by 𝑆.
𝑀𝑐 𝑀 𝑀
𝜎𝑚𝑎𝑥 = = =
𝐼 𝐼 𝑐 𝑆
• The distance 𝑐 used here is measured from the neutral axis to the
most remote fiber of the beam.
• The efficient sections for resisting elastic bending have as large an 𝑆
as possible for a given amount of material.
• To facilitate its use, section moduli for many manufactured cross
sections are tabulated in handbooks.
• Also, once the bending moment for a beam is determined and an
allowable stress is decided upon, flexure formula maybe solved for
the required section modulus. Using this information, we can select a
beam for design.
• The members resisting flexure maybe proportioned so as to have a
different maximum stress in tension than in compression. This is
significant for materials having different strengths in tension and in
compression.
• As the cast iron is strong in compression and weak in tension, the
proportions of a cast-iron member may be so set as to have a low
maximum tensile stress. the potential capacity of the material may
thus be better utilized.
 Stress Concentrations
• If notches, grooves, bolt holes, or an abrupt change in the cross-
sectional area of the beam occur, high local stresses arise.
• For the axial and torsion members, most of the information regarding
the actual stress distribution came from accurate photoelastic
experiments. Numerical methods employing finite elements are now
extensively used for the same purpose.
• The ratio 𝐾 of the actual maximum stress to the nominal maximum
stress in the minimum section, is defined as the stress-concentration
factor in bending, Fig. 8.
• In general,
𝑀𝑐
𝜎𝑚𝑎𝑥 𝑎𝑐𝑡𝑢𝑎𝑙 = 𝐾
𝐼
 Fig. 8: Meaning of stress-concentration factor in bending

• In this equation 𝑀𝑐 𝐼 is for the small width of the bar.

• Fig. 9 and 10 are plots of stress-concentration factors for two
representative cases. The factor 𝐾, depending on the proportions of
the member, maybe obtained from these diagrams.
• A study of these graphs indicates the desirability of generous fillets
and the elimination of sharp notches to reduce local stress
concentrations.
• Inelastic behavior of material tends to reduce these factors.
 Fig. 9: Stress-concentration factors in pure Fig. 10: Stress-concentration factors in bending
bending for flat bars with various fillets for grooved flat bars

• For ductile materials, where the applied forces are static, stress
concentrations are less important.
• If the cross-sectional area has reentrant angles, stress concentrations
become particularly significant, e.g. an 𝐼 beam. To minimize these,
commercially rolled or extruded shapes have a generous fillet.
 Elastic Strain Energy in Pure Bending
• The normal stress varies linearly from the neutral axis, as shown in
Fig. 11, and according to the stress 𝜎 = − 𝑀𝑦 𝐼.
• The volume of a typical infinitesimal beam element is 𝑑𝑥 𝑑𝐴, where
𝑑𝑥 is its length and 𝑑𝐴 is its cross sectional area.

• Fig. 11: A beam segment for deriving strain energy in bending

• Substituting these expressions, and integrating over the volume 𝑉 of

the beam, the elastic strain energy 𝑈 in a beam in pure bending can
be computed.
 2
𝜎𝑥2 1 𝑀𝑦
𝑈= 𝑑𝑉 = − 𝑑𝑥 𝑑𝐴
2𝐸 2𝐸 𝐼
• As the moment 𝑀 at a section of a beam is constant and the order of
performing the integration is arbitrarily,
𝐿 2
𝑀2 2
𝑀 𝑑𝑥
𝑈= 2
𝑑𝑥 𝑦 𝑑𝐴 =
2𝐸𝐼 0 2𝐸𝐼

• Above equation reduces the volume integral for the elastic energy of
prismatic beams in pure flexure to a single integral taken over the
length L of a beam.
• Alternatively, consider an elementary segment of a beam 𝑑𝑥 long, as
shown in Fig. 12.
 • After the application of the bending moments,
extensions of the same two planes, which remains
plane, intersects at 𝑂, and the angle induced between
these two planes is 𝑑𝜃 .
• Since the full value of the moment 𝑀 is attained
gradually, the average moment acting through an
angle 𝑑𝜃 is 12 𝑀.
• For small deflections, 𝑑𝑥 ≈ 𝜌 𝑑𝜃, where 𝜌 is the radius
of curvature of the elastic curve with 1 𝜌 = 𝑀 𝐸𝐼.
• From the principle of conservation of energy, the
internal strain energy of an element of a beam,
1 1 𝑑𝑥 𝑀2 𝑑𝑥
𝑑𝑈 = 𝑑𝑊𝑒 = 𝑀 𝑑𝜃 = 𝑀 =
2 2 𝜌 2𝐸𝐼
Fig. 12: Beam Segment
Beams of Composite Cross Section
• Wooden beam are sometimes reinforced by metal straps, plastics are
reinforced with fibers, and reinforced concrete is concrete with steel
reinforcing bars.

Fig. 13: Elastic beam of composite cross section in bending

• Consider an elastic beam of several materials bonded together with a
vertical axis of symmetry as shown in Fig. 13(a).
• For a homogeneous material, the longitudinal extensional strains 𝜀𝑥 are
assumed to vary linearly as shown in Fig. 13(b).
• The neutral axis for this section, passing through the modulus-weighted
centroid, is located by the distance 𝑦𝑏 .
• At the interfaces between two materials, depending on the relative
values of their 𝐸𝑖 ’s, a sharp discontinuity in stress magnitudes arises.
• The resisting bending moment
𝑀𝑧 = 𝜅 𝐸𝑖 𝑦 2 𝑑𝐴 = 𝜅 𝐸𝐼 ∗

𝑀𝑧 𝑀𝑧
𝜀𝑥 = − ∗ 𝑦 and 𝜎𝑥 = −𝐸𝑖 ∗ 𝑦
𝐸𝐼 𝐸𝐼
• Sometimes, it is useful to introduce the concept of an equivalent or
transformed cross-sectional area in one material.
• Arbitrary selecting a reference 𝐸𝑖 , and defining it as 𝐸𝑟𝑒𝑓 . For constant
curvature 𝜅,
𝐸𝑖
𝐸𝑖 𝑦 𝑑𝐴 = 𝐸𝑟𝑒𝑓 𝑦 𝑑𝐴 = 𝐸𝑟𝑒𝑓 𝑦 𝑛𝑖 𝑑𝐴 = 0
𝐸𝑟𝑒𝑓
• Where, 𝑛𝑖 𝑑𝐴 = 𝐸𝑖 𝐸𝑟𝑒𝑓 𝑑𝐴. Therefore a beam of composite cross-
section can be considered to have the mechanical properties of the
reference material. After transforming a cross section in this manner,
conventional elastic analysis is applicable.
• In transformed sections the stresses vary linearly from the neutral axis
in all materials.
• The actual stresses are obtained for the reference material, whereas
the stresses in other materials must be multiplied by 𝑛𝑖 .
 Curved Bars
• Consider a curved member such as shown in Figs. 14(a) and (b). The
outer fibers are at a distance of 𝑟0 from the center of curvature 𝑂.
The inner fibers are at a distance of 𝑟𝑖 . The distance from 𝑂 to the
centroidal axis is 𝑟.

Fig. 14: Curved bar in pure bending

• Assumption: sections perpendicular to the axis of the beam remain
plane after a bending moment 𝑀 is applied. This is diagrammatically
represented by the line 𝑒𝑓 in relation to an element of the beam
𝑎𝑏𝑐𝑑. The element is defined by the central angle 𝜙.
• The initial length of a beam fiber such as 𝑔ℎ depends upon the
resistance 𝑟 from the center of curvature.
• Although the total deformation of beam fibers 𝑑𝜙 follows a linear
law, strains do not.
• The elongation of generic fiber 𝑔ℎ is R − r 𝑑𝜙, where R is the
distance from 𝑂 to the neutral surface, and its initial length is rϕ.
• The strain 𝜀 of any arbitrary fiber is 𝑅 − 𝑟 𝑑𝜙 𝑟𝜙, and the normal
stress 𝜎 on an element 𝑑𝐴 of the cross-sectional area is
𝑅 − 𝑟 𝑑𝜙
𝜎 = 𝐸𝜀 = 𝐸
𝑟𝜙
• The location of the neutral axis follows from the condition that the
summation of the forces acting perpendicular to the section must be
equal to zero, i.e.,
𝐸 𝑅−𝑟 𝑑𝜙
𝐹𝑛 = 0 𝜎 𝑑𝐴 = 𝑑𝐴 = 0
𝑟𝜙
• However, since 𝐸, 𝑅, 𝜙, and 𝑑𝜙 are constant at any one section of a
stressed bar, they may be taken outside the integral sign and a
solution for 𝑅 obtained.
𝐸 𝑑𝜙 𝑑𝐴 𝐴
𝑅 − 𝑑𝐴 = 0 𝑅 =
𝜙 𝑟 𝑑𝐴 𝑟
• The summation of moments is made around the 𝑧 axis, which is
normal to the plane of Fig. 14(a).
• Remembering that 𝐸, 𝑅, 𝜙, and 𝑑𝜙 are constant at a section,
𝐸 𝑑Φ 𝑅−𝑟 2 𝜎𝑟 𝑅−𝑟 2
𝑀= 𝑑𝐴 = 𝑑𝐴
Φ 𝑟 𝑅−𝑟 𝑟
𝜎𝑟 2 𝑑𝐴
= 𝑅 − 𝑅 𝑑𝐴 − 𝑅 𝑑𝐴 + 𝑟 𝑑𝐴
𝑅−𝑟 𝑟

• Since R is a constant, the first two integrals vanish as may be seen

from one of the equations in previous slide. The third integral is A,
and the last integral is 𝑟𝐴 where 𝑟 is the radius of the centroidal axis.
• Normal stress acting on a curved beam at a distance 𝑟 from the
center of curvature is
𝑀 𝑅−𝑟
𝜎=
𝑟𝐴 𝑟 − 𝑅
• If positive 𝑦 measured toward the center of curvature from the
neutral axis, and 𝑟 − 𝑅 = 𝑒
• The flexure formula for straight beams,
𝑀𝑦
𝜎=
𝐴𝑒 𝑅 − 𝑦
• This equation indicate that the stress distribution in a curved bar
follows a hyperbolic pattern.
• A comparison of this result with the one that follows from the
formula for straight bars is shown in Fig. 14(c).
• In the curved bar, the neutral axis is pulled toward the center of the
curvature of the beam. This results from the higher stresses
developed below the neutral axis.
• This developed theory applies, only to elastic stress distribution and
only to beam in pure bending.
• There is 7% error in the maximum stress for 𝑟 ℎ = 5. where, ℎ is the
height of rectangular cross section.
• When 𝑟 ℎ = 1.5, a 28% error occurs.
• For greater ratios of 𝑟 ℎ, this error diminishes.
• As the curvature of beam increases, the stress on the concave side
rapidly increases over the one given by the usual flexure formula.
• Second, the evaluation of the integral for 𝑅 over the cross-sectional
area may become very complex. Also, calculations of 𝑅 must be very
accurate since difference between 𝑅 and numerically comparable
quantities are used in the stress formula.
• Yet another approach consists of working “in reverse”.
• Curved beams of various cross sections, curvatures, and applied
moments are analyzed for stress; then these quantities are divided by
a flexural stress that would exist for the same beam if it were straight.
These ratios are than tabulated.
• Stress in a curved beam,
𝑀𝑐
𝜎= 𝐾
𝐼
Where, the coefficient 𝐾 is obtained from a table or graph and 𝑀𝑐 𝐼
is computed as in the usual flexure formula.
• An expression for the distance from the center of curvature to the
neutral axis of a curved beam of circular cross-sectional area:
𝑟 + 𝑟2 − 𝑐2
𝑅=
2
where 𝑟 is the distance from the center of curvature to the centroid
and 𝑐 is the radius of the circular cross-sectional area.
Inelastic Bending of Beams
• Consider a beam having a cross-section such as shown in Fig. 15(a).

Fig. 15: Inelastic bending of beam

• Assuming plane sections remain plane after deformation, the
longitudinal normal strains vary linearly as in Fig. 15(b).
• For the several selected strains, 𝜀1 , 𝜀2 ,…, 𝜀5 in this diagram, the
corresponding stresses 𝜎1 , 𝜎2 , …, 𝜎5 are defined on the given stress-
strain diagram in Fig. 15(c).
• A plot of these stresses along a section establishes a possible stress
distribution in the beam along the curved line 𝐴𝐵 as illustrated in Fig.
15(d).
• Stresses acting on the corresponding area of the cross section, develop
a compressive force 𝐶 above the neutral axis, and a tensile force 𝑇
below it. For neutral axis, 𝑇 = 𝐶.
𝜎 𝑑𝐴 = 0
where, 𝜎 is the normal bending stress acting on a section.
• The resisting bending moment 𝑀𝑧 at the same section is known to be
𝐶 𝑎 + 𝑏 or 𝑇 𝑎 + 𝑏 , Fig. 15(d).
𝑀𝑧 = − 𝜎𝑦 𝑑𝐴
• If the beam cross section is symmetric around the horizontal axis and
material properties are the same in tension and compression, neutral
axis passes through the centroid of the section, and above equation
can be directly applied.
• The behavior of such a beam in bending is shown qualitatively in Fig.
16.
• A sequence of progressively increasing strains associated with plane
sections is shown in Fig. 16(b). These maximum strains define the
maximum stresses maximum stresses in the outer fiber of beam, Fig.
16(c).
• As can be seen from Figs. 16(a) and (c), the maximum attainable stress is
𝜎3 . The instantaneous stress distribution in the beam associated with 𝜎3 ,
for this brittle material, is given by the curved line 𝐴𝐵 in Fig. 16(c).

Fig. 16: Rectangular beam in bending exceeding the proportional limit of the material
• The nominal stress in the extreme fibers is computed by applying the
elastic flexure formula, using the experimentally determined ultimate
bending moment. This stress is called the ‘rupture modulus’ of the
material in bending. This stress is associated with the line 𝐶𝐷 in Fig.
16(c) and is larger than the stress actually attained.
• Consider a rectangular beam of elastic-plastic material, Fig. 17. a sharp
separation of the members into distinct elastic and plastic zones is
possible. If the strain in the extreme fiber is double that at the
beginning of yielding, only the middle-half of the beam remains
elastic, Fig. 17(a). In this case, the outer quarter of the beam yield.
• At higher strains, the elastic zone, or core, diminishes. Stress
distribution corresponding to this situation is shown in Fig. 17 (b) and
(c).
Fig. 17: Elastic-plastic beam at large level of straining
Part B- Unsymmetrical Bending and Bending with
Axial Loads
Bending about Both Principle Axes
• Consider the rectangular beam shown in Fig. 18. the applied moment
act in the plane 𝑎𝑏𝑐𝑑. By using the vector representation for 𝑀
shown in Fig. 18(b), this vector forms an angle 𝛼 with the 𝑧 axis and
can be resolved into the two components, 𝑀𝑦 and 𝑀𝑧 .
• Because of the symmetry, the product of inertia for this section is
zero, and the orthogonal axis shown are the principal axes for the
cross section. This also holds true for the centroidal axes of singly
symmetric areas.
• By assuming elastic behavior of material, a superposition of the stress
caused by 𝑀𝑦 and 𝑀𝑧 is the solution to the problem.
 Fig. 18: Unsymmetrical bending of a beam with doubly symmetric cross section

𝑀𝑧 𝑦 𝑀𝑦 𝑧
𝜎𝑥 = − +
𝐼𝑧 𝐼𝑦
• A graphical illustration of superposition is given in Fig. 19. neutral
axes 𝜎𝑥 = 0 forms at an angle 𝛽 with the 𝑧 axis.
 Fig. 19: Superposition of elastic bending stress

𝑀𝑧 𝑦 𝑀𝑦 𝑧 𝑦 𝑀𝑦 𝐼𝑧
− + = 0 or tan 𝛽 = =
𝐼𝑧 𝐼𝑦 𝑧 𝑀𝑧 𝐼𝑦
• Since, in general, 𝑀𝑦 = 𝑀 sin 𝛼 and 𝑀𝑧 = 𝑀 cos 𝛼,
𝐼𝑧
tan 𝛽 = tan 𝛼
𝐼𝑦
• Unless 𝐼𝑧 = 𝐼𝑦 , or 𝛼 is either 0° or 90°, the angles 𝛼 and 𝛽 are not
equal. In general, the neutral axis and the normal to a plane in which
the applied moment acts do not coincide.
• Consider a beam with the arbitrary cross section shown in Fig. 20.

Fig. 20: Pure bending around a principal axis

• Let such an elastic beam be bent about the principal 𝑧 axis and assume
that the stress distribution is given as 𝜎𝑥 = − 𝑀𝑧 𝑦 𝐼𝑧 .
• If this stress distribution causes no bending moment 𝑀𝑦 around the 𝑦
axis, this is correct solution of the problem.
𝑀𝑧 𝑀𝑧
𝑀𝑦 = − 𝑦𝑧 𝑑𝐴 = − 𝑦𝑧 𝑑𝐴 = 0
𝐼𝑧 𝐼
• The constants are placed in front of the second integral, which is equal
to zero because by definition a product of inertia for principal axis
vanishes.
• When unsymmetrical bending of a beam is caused by applied
transverse forces, The applied forces are first resolved into components
that act parallel to the principal axes of the cross-sectional area.
• Then the bending moment caused by these components around the
respective axes are computed for use in the flexure formula.
• To avoid torsional stresses, the applied transverse forces must act
through the shear center.
• For bilaterally symmetrical sections, e.g., a rectangle, a circle, an Ι beam,
etc., the shear center coincides with the geometric center (centroid) of the
cross section.
• For other sections, such as a channel, the shear center lies elsewhere, as at
𝑆 shown in Fig. 21, and it is at this point that the transverse force must be
applied to prevent occurrence of torsional stresses.
• For analysis of unsymmetrical bending, the applied forces must be
resolved at the shear center parallel to the principal axes of the cross
section.

Fig. 21: Lateral force through shear center 𝑆 causes no torsion

Elastic Bending with Axial Loads
• A solution for pure bending around both principal axes of a member
can be extended to include the effect of axial loads by employing
superposition.

Fig. 22
 𝑃 𝑀𝑧 𝑦 𝑀𝑦 𝑧
𝜎𝑥 = − +
𝐴 𝐼𝑧 𝐼𝑦
• Where, 𝑃 is taken positive for axial tensile forces, and bending takes
place around the two principal 𝑦 and 𝑧 axes.
• Consider the case shown in Fig. 22(a). By applying two equal but
opposite forces 𝑃 at centroid 𝐶, as shown in Fig. 22(b).
• A couple 𝑃𝑑 developed by the opposed forces 𝑃 a distance 𝑑 apart
causes unsymmetrical bending. The two components along the
principal axes are 𝑀𝑦 = 𝑃𝑧0 and 𝑀𝑧 = 𝑃𝑦0 . Since the sense of these
moments coincides with the positive directions of 𝑦 and 𝑧 axes, these
moments are positive.
• To complete a solution, the normal stress caused by axial force must be
superposed.
• The equation of a plane is given as,
𝐴𝑥 + 𝐵𝑦 + 𝐶𝑧 + 𝐷 = 0
here, setting 𝐴 = 1, 𝑥 = 𝜎𝑥 , 𝐵 = 𝑀𝑧 𝐼𝑧 , 𝐶 = − 𝑀𝑦 𝐼𝑦 , and 𝐷 = − 𝑃 𝐴
• Similarly, since 𝜀 = 𝜎 𝐸,
𝜀𝑥 = 𝑥 = − 𝑏𝑦 + 𝑐𝑧 + 𝑑
where, 𝑎 = 1, and 𝑏, 𝑐, and 𝑑 are constants.
• Because of the presence of axial strain due to 𝑃, the plane sections also
translate an amount 𝑃 𝐴𝐸.
• The longitudinal strain magnitudes in members subjected to bending
and axial forces can be represented by distances from a reference plane
to an inclined plane. These inclined planes intersect the reference plane
in line. This line of zero stress is analogous to the neutral axis occurring
in pure bending.
• However, when 𝑃 ≠ 0, this line does not pass through the centroid of a
section. For large axial force and small bending moments, the line of
zero stress or strain may lie outside a cross section. The normal stresses
or strains vary from it linearly.
• The bending moment in a member is caused by transverse force rather
than by an eccentrically applied force such as illustrated in Fig. 22.
• If the resultant of all vertical forces acts within the middle third of the
rectangular cross section, there is no tension in the material at this
section. The resultant of these forces may be made to intersect the
plane of the cross section, as shown in Fig. 23.
• At the point of intersection of this resultant with the section, it may be
resolved into horizontal and vertical components.
• If the vertical component of the resultant fulfils the conditions, no
tension will be developed at point 𝐵, as the horizontal component
causes only shear stresses.
 Fig. 23: Resultant causing no tension at 𝐵

• A more general “middle-third” rule may be stated thus: there will be no

tension at a section of a member of a rectangular cross section if the
resultant of the forces above these section intersects one of the axes of
symmetry of the section within the middle third.
Inelastic Bending with Axial Loads
• Plane sections normal through a beam taken normal to its axes
remain plane after a beam is bent. Plane sections perpendicular to a
beam axis move along it parallel to themselves when an inelastic
member is loaded axially.
• For small deformations, the normal strains corresponding to these
actions can be superposed.
• The superposition of strains for a planar member simultaneously
subjected to an axial force 𝑃 and a bending moment 𝑀 is shown
schematically in Fig. 24.
• Superposition of strains due to 𝑃 and 𝑀 moves a plane section axially
and rotates it as shown.
• If axial force 𝑃 causes strain larger than any strain of opposite sign that is
caused by 𝑀, the combined strains will not change their sign within a
section.
• By supplementing these basic kinematic assumptions with the stress-
strain relations and conditions of equilibrium, one can solve either elastic
or inelastic problems.

Fig. 24: Superposition of strains

Bending of Beams with Unsymmetrical
(Arbitrary) Cross Section
• Two basic requirements for equilibrium: (1) the total axial force on
any section of a beam must be zero, and (2) the external bending
moment at a section must be developed by the internal stresses
acting on the cross section. Hooke’s law is postulated for uniaxial
normal strain.
• Consider a beam having an arbitrary cross section such as shown in
Fig. 25.the orientation of the 𝑦 and 𝑧 orthogonal axes is chosen
arbitrarily. Let this beam subjected to a pure bending moment 𝑀
having the components 𝑀𝑦 and 𝑀𝑧 , respectively, around the 𝑦 and 𝑧
axes respectively, Fig. 25(a).
 Fig. 25: Bending of unsymmetrical cross section

• During bending, a plane section through a beam would rotate and

intersect the 𝑦𝑧 plane at an angle 𝛽 with the 𝑧 axis.
• A generic infinitesimal area 𝑑𝐴 in the positive quadrant of the 𝑦 and 𝑧
axes is located by the perpendicular distance 𝑟 from this line.
• The longitudinal normal stress,
𝜀𝑥 = −𝜅𝑟
 𝑟 = 𝑦 cos 𝛽 − 𝑧 sin 𝛽
• The longitudinal elastic stress 𝜎𝑥 acting on the cross section is,
𝜎𝑥 = −𝐸𝜅𝑦 cos 𝛽 + 𝐸𝜅𝑧 sin 𝛽
where, 𝜅 cos 𝛽 is the projected curvature 𝜅𝑦 in the 𝑥𝑦 plane, as may be
seen from the limiting case of setting 𝛽 equal to zero. similarly 𝜅 sin 𝛽 is
the projected curvature 𝜅𝑧 in the 𝑥𝑧 plane.
𝜎𝑥 = −𝐸𝜅𝑦 𝑦 + 𝐸𝜅𝑧 𝑧
• A mathematically defined positive curvature, causing an increase in the
slope of a bent beam with an increase in the distance from the origin,
gives rise to two different cases. In the 𝑥𝑦 plane, positive curvature and
positive bending moment have the same sense. The opposite is true in
the 𝑥𝑧 plane. Hence, the normal stress due to 𝜎𝑥 must be of opposite
sign, Fig. 26.
 Fig. 26: Relationships between positive moments and curvatures in 𝑥𝑦- and 𝑥𝑧-planes

• The sum of all forces in the 𝑥 direction must equal to zero, i.e.,
𝐹𝑥 = 0, can be written as
𝜎𝑥 𝑑𝐴 = −𝐸𝜅𝑦 𝑦 𝑑𝐴 + 𝐸𝜅𝑧 𝑧 𝑑𝐴 = 0
• The arbitrary orthogonal axes in Fig. 25 are shown passing through the
centroid 𝐶 of the cross section.
• By imposing the conditions of moment equilibrium at a section, two
moment component equations can be written requiring that the
externally applied moment around either axis is balanced by the internal
system of stresses.
𝑀𝑧 = −𝜎𝑥 𝑦 𝑑𝐴 = 𝐸𝜅𝑦 𝑦 2 𝑑𝐴 − 𝐸𝜅𝑧 𝑦𝑧 𝑑𝐴

and 𝑀𝑦 = +𝜎𝑥 𝑧 𝑑𝐴 = −𝐸𝜅𝑦 𝑦𝑧 𝑑𝐴 + 𝐸𝜅𝑧 𝑧 2 𝑑𝐴

• These integrals define the moments and product of inertia for a cross
sectional area as 𝐼𝑧 , 𝐼𝑦 , and 𝐼𝑦𝑧 .
𝑀𝑧 = 𝐸𝐼𝑧 𝜅𝑦 − 𝐸𝐼𝑦𝑧 𝜅𝑧

and 𝑀𝑦 = −𝐸𝐼𝑦𝑧 𝜅𝑦 + 𝐸𝐼𝑦 𝜅𝑧

• Solving these two equations simultaneously gives,
 𝑀𝑧 𝐼𝑦 +𝑀𝑦 𝐼𝑦𝑧
𝐸𝜅𝑦 = + 2
𝐼𝑦 𝐼𝑧 −𝐼𝑦𝑧
𝑀𝑦 𝐼𝑧 +𝑀𝑧 𝐼𝑦𝑧
and 𝐸𝜅𝑧 = − 2
𝐼𝑦 𝐼𝑧 −𝐼𝑦𝑧
• The expression for the elastic bending stress 𝜎𝑥 for any beam cross
section with arbitrarily directed orthogonal coordinate axes is
𝑀𝑧 𝐼𝑦 +𝑀𝑦 𝐼𝑦𝑧 𝑀𝑦 𝐼𝑧 +𝑀𝑧 𝐼𝑦𝑧
𝜎𝑥 = − 2 𝑦+ 2 𝑧
𝐼𝑦 𝐼𝑧 −𝐼𝑦𝑧 𝐼𝑦 𝐼𝑧 −𝐼𝑦𝑧

• If the principal axes for a cross section are used, where 𝐼𝑦𝑧 = 0, this
equation simplifies to basic form.
• The angle 𝛽 for locating the neutral axis in the arbitrary coordinate system
is obtained,
𝑦 𝑀𝑦 𝐼𝑧 +𝑀𝑧 𝐼𝑦𝑧
tan 𝛽 = =
𝑧 𝑀𝑧 𝐼𝑦 +𝑀𝑦 𝐼𝑦𝑧
 Part C- Area Moments of Inertia
Area Moments and Products of Inertia
• Concept of Moments of inertia, or second moments of area is
generalized for two orthogonal axes for any cross-sectional shape.
• With the 𝑦𝑧 coordinates chosen as shown in Fig.27, the moments and
product of inertia of an area are given as,
𝐼𝑧 = 𝑦 2 𝑑𝐴 𝐼𝑦 = 𝑧 2 𝑑𝐴 and 𝐼𝑦𝑧 = 𝑦𝑧 𝑑𝐴
• These axes are chosen to pass through the centroid 𝐶 of the area. The
product of inertia vanishes either for doubly or singly symmetric
areas, Fig. 28.
 Fig. 27: Rotation of Coordinate axes Fig. 28: (a) Doubly and (b) singly symmetric cross sections
• In Fig. 28 (b), due to symmetry, for each 𝑦 +𝑧 𝑑𝐴, there is a 𝑦 −𝑧 𝑑𝐴,
and their sum vanishes.
• In calculating moments of inertia for symmetric cross sections having
complex areas, it is advantageous to subdivide such areas into simple
parts for which the moments of inertia are available as formulas. Then by
applying the parallel-axis theorem to each part and adding, the moment
of inertia for the whole section is obtained.
• For the transfer of a moment of inertia for an area from the 𝑧𝑐 to the 𝑧 or
𝑦𝑐 to the 𝑦 axis remains applicable.
𝐼𝑧 = 𝐼𝑧𝑐 + 𝐴𝑑𝑧2
and
𝐼𝑦 = 𝐼𝑦𝑐 + 𝐴𝑑𝑦2
• Where, 𝐼𝑧𝑐 and 𝐼𝑦𝑐 are, moments of inertia around the 𝑧𝑐 and 𝑦𝑐 axes, 𝐴 is
the area considered, and 𝑑𝑧 and 𝑑𝑦 are, the distances from 𝐶 to the axes 𝑧
and 𝑦.
• Following the same procedure as before for 𝐼𝑧 and 𝐼𝑦 , the transfer-of-axis
formula for the product of inertia, becomes
𝐼𝑦𝑧 = 𝑦𝑐 + 𝑑𝑧 𝑧𝑐 + 𝑑𝑦 𝑑𝐴 = 𝐼𝑦𝑐 𝑧𝑐 + 𝐴𝑑𝑦 𝑑𝑧
• Where, 𝐼𝑦𝑐𝑧𝑐 is the product of inertia of the area 𝐴 around the centroidal
𝑦𝑐 and 𝑧𝑐 axes.
 Principal Axes of Inertia
• As can be seen from Fig. 27, the axes are rotated through an angle 𝜃,
forming a new set of 𝑦 ′ 𝑧′ coordinates. In order to transform these
quantities from one set of coordinates to another,
𝑦 ′ = 𝐶𝑃 + 𝑃𝑆 = 𝑦 cos 𝜃 + 𝑧 sin 𝜃
𝑧 ′ = 𝑁𝑅 − 𝑅𝑆 = 𝑧 cos 𝜃 − 𝑦 sin 𝜃
• Then, based on the definitions for moments and product of inertia,
𝐼𝑧′ = 𝑦′ 2 𝑑𝐴 = 𝑦 cos 𝜃 + 𝑧 sin 𝜃 2 𝑑𝐴
= cos2 𝜃 𝑦 2 𝑑𝐴 + sin2 𝜃 𝑧 2 𝑑𝐴 + 2 sin 𝜃 cos 𝜃 𝑦𝑧 𝑑𝜃
= 𝐼𝑧 cos2 𝜃 + 𝐼𝑦 sin2 𝜃 + 2𝐼𝑦𝑧 sin 𝜃 cos 𝜃
1+cos 2𝜃 1−cos 2𝜃
= 𝐼𝑧 + 𝐼𝑦 + 𝐼𝑦𝑧 sin 2𝜃
2 2
 𝐼𝑧 +𝐼𝑦 𝐼𝑧 −𝐼𝑦
𝐼𝑧′ = + cos 2𝜃 + 𝐼𝑦𝑧 sin 2𝜃
2 2
𝐼𝑧 +𝐼𝑦 𝐼𝑧 −𝐼𝑦
similarly, 𝐼𝑦′ = − cos 2𝜃 − 𝐼𝑦𝑧 sin 2𝜃
2 2
𝐼𝑧 −𝐼𝑦
and 𝐼𝑦′𝑧′ = − sin 2𝜃 + 𝐼𝑦𝑧 cos 2𝜃
2
• These equations relate the moments and the product of inertia of an
area (second moments) in the new 𝑦 ′ 𝑧′ coordinates to the initial ones in
the 𝑦𝑧 coordinates through the angle θ.
also 𝐼𝑦′ + 𝐼𝑧′ = 𝐼𝑦 + 𝐼𝑧
• For a maximum or minimum value of 𝐼𝑧′ or 𝐼𝑦′
𝑑𝐼𝑧′
= − 𝐼𝑧 − 𝐼𝑦 sin 2𝜃 + 2𝐼𝑦𝑧 cos 2𝜃 = 0
𝑑𝜃
 2𝐼𝑦𝑧
• Hence, tan 2𝜃1 =
𝐼𝑧 −𝐼𝑦
• This equation gives two roots within 360° that are 180° apart. Since this
is for a double angle 2𝜃1 , the roots for 𝜃1 are 90° apart. These roots
locates two conjugate axis for maximum and minimum moment of
inertia. These are principal axes of inertia, for which the product of
inertia is zero.
• Expressions for principal moments of inertia:
𝐼𝑧 +𝐼𝑦 𝐼𝑧 −𝐼𝑦 2 2
𝐼𝑚𝑎𝑥 = 𝐼1 or 𝐼2 = ± + 𝐼𝑦𝑧
𝑚𝑖𝑛 2 2

where by definition, 𝐼1 = 𝐼𝑚𝑎𝑥 , and 𝐼2 = 𝐼𝑚𝑖𝑛 .