Download as pdf or txt
Download as pdf or txt
You are on page 1of 20

MECHANICS OF SOLIDS

Mechanical III/I

CURVED BEAM

Session 29: Circumferential Stress in


Curved Beams
2024
Learning Objectives
• Understand the difference between a straight beam and a curved
beam.

• Understand the assumptions of curved beam formula.

• Derive the curved beam formula.

• Understand the section properties required for a curved beam


problem.
The theory of pure beam bending is limited to straight beams or to
beams that are mildly curved relative to their depth.
However, if the ratio of the radius of curvature to depth for a beam
is less than 5, the flexure formula 𝜎 = 𝑀𝑦/𝐼 is generally
inadequate for describing the flexural stresses in the beam.

For beams that are curved in such a manner, the theory of bending
must also include consideration of the curvature.
There are two important differences with respect to straight-beam
bending result.

First, the flexural stress distribution in a curved beam is nonlinear.


Based on this result the neutral axis will not coincide with the
centroidal axis of the cross section when the beam is subjected to
pure bending.
Second, a curved beam carries radial stresses as a consequence of
the internal bending moment.

These radial stresses have important design implications for thin-


wall cross sections and for materials (such as wood and
unidirectional composites) with relatively low tensile strength in
the radial direction.

Timoshenko and Goodier (1970) presented a solution based on the


theory of elasticity for the linear elastic behavior of curved beams
of rectangular cross sections for the loading shown in Figure
6.1(a). They obtained relations for the radial stress 𝜎𝑟𝑟 , the
circumferential stress 𝜎𝜃𝜃 , and the shear stress 𝜎𝑟𝜃 [Figure
6.1(b)].
However, most curved
beams do not have
rectangular cross sections.
Therefore, in this chapter,
we present an approximate
curved beam solution that is
generally applicable to all
symmetrical cross sections.

This solution is based on two simplifying assumptions: (a) plane


sections before loading remain plane after loading and (b) the
radial stress 𝜎𝑟𝑟 and shear stress 𝜎𝑟𝜃 are sufficiently small so that
the state of stress is essentially one dimensional. The resulting
formula for the circumferential stress 𝜎𝜃𝜃 is the curved beam
formula.
6.1 Circumferential Stress
in Curved Beams
Consider the curved beam
shown in Figure 6.2(a).
The cross section of the
beam has a plane of
symmetry and the polar
coordinates 𝑟, 𝜃 lie in the
plane of symmetry, with
origin at 0, the center of
curvature of the beam.
We assume that the applied
loads lie in the plane of
symmetry. A positive
moment is defined as one
that causes the radius of
curvature at each section of
the beam to increase in
magnitude.

A free-body diagram of an
element 𝐹𝐵𝐶𝐻 of the beam
is shown in Figure 6.2(b).
The normal traction 𝑁, at
the centroid of the cross
section, the shear 𝑉 , and
moment 𝑀, acting on face
𝐹𝐻 are shown in their
positive directions.

These forces must be


balanced by the resultants
due to the normal stress 𝜎𝜃𝜃
and shear stress 𝜎𝑟𝜃 that act
on face 𝐵𝐶.
The effect of the shear
stress 𝜎𝑟𝜃 on the
computation of 𝜎𝜃𝜃 is
usually small, except for
curved beams with very thin
webs.
However, ordinarily,
practical curved beams are
not designed with thin webs
because of the possibility of
failure by excessive radial
stresses. Therefore,
neglecting the effect of 𝜎𝑟𝜃
on the computation of 𝜎𝜃𝜃
is reasonable.
Let the 𝑧 axis be normal to
face 𝐵𝐶 [Figure 6.2(b)]. By
equilibrium of forces in the
𝑧 direction and of moments
about the centroidal 𝑥 axis,
we find

෍ 𝐹𝑧 = න𝜎𝜃𝜃 𝑑𝐴 − 𝑁 = 0

∴ 𝑁 = න𝜎𝜃𝜃 𝑑𝐴 ….. (6.1)

෍ 𝑀𝑥 = න 𝜎𝜃𝜃 𝑅 − 𝑟 𝑑𝐴 − 𝑀𝑥 = 0

∴ 𝑀𝑥 = න 𝜎𝜃𝜃 𝑅 − 𝑟 𝑑𝐴 ….. (6.2)


The integrals of Equations (6.1) and (6.2) cannot be evaluated until
𝜎𝜃𝜃 is expressed in terms of 𝑟. The functional relationship between
𝜎𝜃𝜃 and 𝑟 is obtained from the assumed geometry of deformation
and stress-strain relations for the material.
The elongation 𝑑𝑒𝜃𝜃 of a typical element in the 𝜃 direction is
equal to the distance between faces 𝐹𝐻 and 𝐹 ∗ 𝐻 ∗ and varies
linearly with the distance 𝑅𝑛 − 𝑟 . However, the corresponding
strain 𝜖𝜃𝜃 is a nonlinear function of 𝑟, since the element length
𝑟𝑑𝜃 also varies with 𝑟. This fact distinguishes a curved beam from
a straight beam.

Thus, by Figure 6.2(b), we obtain for the strain


𝑑𝑒𝜃𝜃 𝑅𝑛 − 𝑟 ∆ 𝑑𝜃 𝑅𝑛
𝜖𝜃𝜃 = = = − 1 𝜔 ….. (6.3)
𝑟𝑑𝜃 𝑟𝑑𝜃 𝑟
where 𝜔 = ∆ 𝑑𝜃 \ 𝑑𝜃 ….. (6.4)
It is assumed that the transverse normal stress 𝜎𝑥𝑥 is sufficiently
small so that it may be neglected. Hence, the curved beam is
considered to be a problem in plane stress. Although radial stress
𝜎𝑟𝑟 may, in certain cases, be of importance, here we neglect its
effect on 𝜖𝜃𝜃 .

Then, by Hooke's law, we find


𝑅𝑛 − 𝑟 𝐸𝜔𝑅𝑛
𝜎𝜃𝜃 = 𝐸𝜖𝜃𝜃 = 𝐸𝜔 = − 𝐸𝜔 ….. (6.5)
𝑟 𝑟
Substituting Equation (6.5) into Equations (6.1) and (6.2), we
obtain
𝑑𝐴
𝑁 = 𝑅𝑛 𝐸𝜔 න − 𝐸𝜔 න 𝑑𝐴 = 𝑅𝑛 𝐸𝜔𝐴𝑚 − 𝐸𝜔𝐴 ….. (6.6)
𝑟
and 𝑑𝐴
𝑀𝑥 = 𝑅𝑛 𝑅𝐸𝜔 න − 𝑅 + 𝑅𝑛 𝐸𝜔 න 𝑑𝐴 + 𝐸𝜔 න 𝑟𝑑𝐴
𝑟
𝑀𝑥 = 𝑅𝑛 𝑅𝐸𝜔𝐴𝑚 − 𝑅 + 𝑅𝑛 𝐸𝜔𝐴 + 𝐸𝜔𝑅𝐴 = 𝑅𝑛 𝐸𝜔 𝑅𝐴𝑚 − 𝐴
….. (6.7)
where 𝐴 is the cross-sectional area of the curved beam and 𝐴𝑚 has
the dimensions of length and is defined by the relation
𝑑𝐴
𝐴𝑚 = න ….. (6.8)
𝑟

Equation (6.7) can be rewritten in the form


𝑀𝑥
𝑅𝑛 𝐸𝜔 = ….. (6.9)
𝑅𝐴𝑚 − 𝐴

Then substitution into Equation (6.6) gives


𝑀𝑥 𝐴𝑚 𝑁
𝐸𝜔 = − ….. (6.10)
𝐴 𝑅𝐴𝑚 − 𝐴 𝐴
The circumferential stress distribution for the curved beam is
obtained by substituting Equations (6.9) and (6.10) into Equation
(6.5) to obtain the curved beam formula
𝑁 𝑀𝑥 𝐴 − 𝑟𝐴𝑚
𝜎𝜃𝜃 = + ….. (6.11)
𝐴 𝐴𝑟 𝑅𝐴𝑚 − 𝐴

The normal stress distribution given


by Equation (6.11) is hyperbolic in
form; that is, it varies as 1/𝑟. For
the case of a curved beam with
rectangular cross section ሺ𝑅/ℎ =
The curved beam formula, Equation (6.11), requires that 𝐴𝑚 ,
defined by Equation (6.8), be calculated for cross sections of
various shapes. Explicit formulas for 𝐴, 𝐴𝑚 , and 𝑅 for several
curved beam cross-sectional areas are listed in Table.
Often, the cross section of a curved beam is composed of two or
more of the fundamental areas listed in Table. The values of 𝐴,
𝐴𝑚 , and 𝑅 for the composite area are given by summation. Thus,
for composite cross sections,
𝑛

𝐴 = ෍ 𝐴𝑖 ….. (6.12)
𝑖=1
𝑛

𝐴𝑚 = ෍ 𝐴𝑚𝑖 ….. (6.13)


𝑖=1

σ𝑛𝑖=1 𝑅𝑖 𝐴𝑖
𝑅= 𝑛 ….. (6.14)
σ𝑖=1 𝐴𝑖

where 𝑛 is the number of fundamental areas that form the


composite area.
Location of Neutral Axis of Cross Section
The neutral axis of bending of the cross section is defined by the
condition 𝜎𝜃𝜃 = 0. The neutral axis is located at distance 𝑅𝑛 from
the center of curvature. The distance 𝑅𝑛 is obtained from Equation
(6.11) with the condition that 𝜎𝜃𝜃 = 0 on the neutral surface 𝑟 =
𝑅𝑛 . Thus, Equation (6.11) yields
𝐴𝑀𝑥 ….. (6.15)
𝑅𝑛 =
𝐴𝑚 𝑀𝑥 + 𝑁 𝐴 − 𝑅𝐴𝑚

For pure bending, 𝑁 = 0, and then Equation (6.18) yields


𝐴 ….. (6.16)
𝑅𝑛 =
𝐴𝑚
Review Questions
1. What are the differences between a straight beam and a curved
beam.?
2. List the assumptions of a curved beam formula.

3. Write down the curved beam formula.

4. How section properties 𝐴, 𝐴𝑚 , and 𝑅 are determined for a


simple section and a composite section.

You might also like