lOMoARcPSD|27577562

column analysis

Mechanics of Solids (Bangladesh University of Engineering and Technology)

Studocu is not sponsored or endorsed by any college or university

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com)
 lOMoARcPSD|27577562

5/29/2016

What is Column?
Any axial members that support compressive axial loads such as the
weight of a building, are called columns – and not all structural
members behave the same. If a compressive axial force is applied to a
long, thin wooden strip, then it bends significantly as shown in fig. 1a. If
the columns of a building were to bend the way, the building itself would
collapse. And when a column buckles, the collapse is usually sudden and
catastrophic.

Fig. 1(a)
Fig. 1(b)

Geometry, materials, boundary conditions, and imperfections all

affects the stability of columns.

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 1

 lOMoARcPSD|27577562

5/29/2016

Load-carrying structures may fail in a variety of ways, depending upon

- the type of structure,
- the conditions of support,
- the kinds of loads, and
- the materials used.

Types of Column
Columns may be divided into three general types:
- Short Columns,
- Intermediate Columns, and
- Long Columns.
The sharp line of demarcation between types of columns is not well
defined, but generally accepted measure is based on the following criteria:
(a) Depending on the failure mode:
A compressive member fails due to the combination of two failure
mode – buckling and crushing.
If a compression member is relatively slender, it may deflect laterally and
fail by bending (Fig.2a) rather than failing by direct compression of the
material or crushing (Fig. 2c). When lateral bending occurs, we say that

the column has buckled. Under an increasing axial load, the lateral
deflections will increase too, and eventually the column will collapse
completely. When a structure (subjected usually to compression)
undergoes visibly large displacements transverse to the load then it is
said to buckle. Buckling is an instability of equilibrium.

Fig.2a: buckling failure Fig.2b

“Long” columns fail structurally (buckling failure)
“Short” columns fail materially (yielding failure) Fig.2c: Crushing
Long Columns: fail by buckling, or excessive lateral deflection
Short Columns (or short compression block): fail by crushing
Intermediate Columns: by a combination of crushing and buckling

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 2

 lOMoARcPSD|27577562

5/29/2016

(b) Depending on the Slenderness Ratio:

The Slenderness Ratio is the (effective) length of the column divided
by its radius of gyration.
The radius of gyration is the distance from an axis which, if the entire
cross sectional area of the object (beam) were located at that distance, it
would result in the same moment of inertia that the object (beam)
possesses. Or, it may be expressed as:
Radius of Gyration: rxx = (Ixx/A)1/2 (radius of gyration about xx-axis)
So, Slenderness Ratio = Le / r.
The subscript “Le” indicates ‘effective length’ i.e. the length of the column,
between points of inflection (that is, points of zero moment) in its
deflection curve, assuming that the curve is extended (if necessary) until
points of inflection are reached (it is not use the actual length).
The effective length is given by: Le = KL
where K is called an effective length constant or effective length
factor. The values for K depends on how the column is supported and L is
the actual length of the column.

A generally accepted relationship between the slenderness ratio and the

type of Column is as follows:
Short Column: 0 < Le / r < 60
Intermediate Column: 60 < Le / r <120
Long Column: 120 < Le / r < 300

(c) Depending on the Failure Stress:

A compressive member fails due to the combination of two stresses
– flexural stress and direct compressive stress.
If the member is short (and the eccentricity is small), the lateral
deflection is small and flexural stress is insignificant compared with the
direct compressive stress. Thus a short column carries principally
Direct compressive stress.
long member however, is quite flexible, hence a relatively low value of
load may cause a large flexural stress accompanied by a negligible direct
compressive stress. Thus, a long column is subjected primarily to
Flexural stress.

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 3

 lOMoARcPSD|27577562

5/29/2016

As the length of the column increases, the importance of the direct

compressive stress decreases and that of the flexural stress increases.
Unfortunately, in the intermediate column range it is not possible to
determine the rate of change in these stresses, or the proportion of each
stress that constitutes the resultant stress. So, it is assumed that the
intermediate column is subjected to a combination of flexural and
direct compressive stress.
Buckling and Stability of Structures
Whenever a member is designed, it is necessary that it satisfy specific
strength, deflection, and stability requirements. In the previous chapters
we had two primary concerns in discussing the analysis and design of
various structures:
 the strength of the structure, i.e. its ability to support a specified load
without experiencing excessive stresses;
 the ability of the structure to support a specified load without
undergoing unacceptable deformations.
i.e. the members strength and deflection assuming that the member
was always in static equilibrium.

Now we shall be concerned with stability of the structure,

- with its ability to support a given load without experiencing a
sudden change in its configuration.
The concept of stability of equilibrium can be demonstrated very easily
by considering the equilibrium of a ball on three different surfaces, as
illustrated in Fig. In all three situations the solid-colored ball is in an
equilibrium position, that is, it satisfies ∑Fx = 0, ∑Fy = 0, and ∑M = 0 .

Fig: Stability of equilibrium.

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 4

 lOMoARcPSD|27577562

5/29/2016

In Fig. a, the ball is said to be in stable equilibrium because, if it is

slightly displaced to one side and then released, it will move back
toward the equilibrium position at the bottom of the “valley.’’

If the ball is on a perfectly flat, level surface, as in Fig. b, it is said to

be in a neutral-equilibrium configuration. If slightly displaced to
either side, it has no tendency to move either away from or toward the
original position, since it is in equilibrium in the displaced position as
well as the original position.

Finally, the ball on the top of the “hill’’ in Fig.c is also in an equilibrium
configuration, because it also satisfies, ∑Fx = 0, ∑Fy = 0, and ∑M = 0.
In this case, however, the ball is in an unstable-equilibrium
configuration; if slightly displaced to either side, the ball will tend to
move farther from the equilibrium position at the top of the “hill.’’

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 5

 lOMoARcPSD|27577562

5/29/2016

To investigate the stability of real columns, consider a hypothetical

structure consists of two rigid bars each of length L/2. They are joined at
A by a pin connection and held in a vertical position by a spring having
stiffness k. [explanation Gere]
Assumptions
 The column is initially perfectly straight, and it is made of linearly
elastic material. Elasticity of the buckling model is concentrated in
the spring (real model can bend throughout its length) as shown in fig.
3a.
 The column is free to rotate, at its ends, about frictionless pins; that
is, it is restrained like a simply supported beam. Each pin passes
through the centroid of the cross section.
 The column is symmetric about the xy plane, and any lateral
deflection of the column takes place in the xy plane.
 The column is loaded by an axial compressive force P has its line of
action along the longitudinal axis. Consequently, the spring is
initially unstressed and the bars are in direct compression.

Fig. 3c
Fig. 3a Fig. 3b
(Hibbeler)

Now suppose that the structure is disturbed by some external force

that causes point A to move a small distance Δ laterally (Fig. 3b).

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 6

 lOMoARcPSD|27577562

5/29/2016

The rigid bars rotate through small angles θ and a moment

develops in the spring. The direction of this moment is such that it
tends to return the structure to its original straight position, and
therefore it is called a restoring moment.
At the same time, however, the tendency of the axial compressive
force is to increase the lateral displacement. Thus, these two actions
have opposite effects—
- the restoring force tends to decrease the displacement and
- the axial force tends to increase it.
Now consider what happens when the disturbing force is removed.
If the axial force P is relatively small, the action of the restoring
force will predominate over the action of the axial force and the
structure will return to its initial straight position.
The structure is said to be stable.
If the axial force P is large, the lateral displacement of point A will
increase and the bars will rotate through larger and larger angles until
the structure collapses. Under these conditions,
The structure is unstable and fails by lateral buckling.

Critical Load
The transition between the stable and unstable conditions occurs at
a special value of the axial force known as the critical load, Pcr.
The critical load of the buckling model can be determined by
considering the structure in the disturbed position (Fig. 3b) and
investigating its equilibrium.
First, we consider the entire structure as a free body and sum moments
about support A in the x-direction. Consider upper bar as a free body
(Fig. 3c) subjected to axial forces P and force F in the spring.
Restoring Force, F is equal to the stiffness k times the displacement ∆,
F = k∆
The applied load P develops two horizontal components,
Px = P tanθ which tends to push the pin (and the bars)
further out of equilibrium.
Since θ is small, the lateral displacement of point A is
Δ ≈ θ(L/2) and tanθ ≈ θ

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 7

 lOMoARcPSD|27577562

5/29/2016

Thus the restoring spring force becomes F = k∆= kθ(L/2) and

the disturbing force is 2Px = 2P tanθ = 2Pθ
Applying equilibrium and solving for P: 2Px = F
Pcr = kL/4
What is Critical Load?
The loading repressents a case of the mechanism being in neutral
equilibrium. Since Pcr is independent of (small) displacement θ of the
bars, any slight disturbance given to the mechanism will not cause it to
move further out of equilibrium, nor will it be restored to its original
position. Instead, the bars will remain in the deflected position.
The critical load is the only load for which the structure will be in
equilibrium in the disturbed position. At this value of the load, the
restoring effect of the moment in the spring just matches the
buckling effect of the axial load.
Therefore, the critical load represents the boundary between the
stable and unstable conditions.

If the axial load is less than Pcr, i.e. (0 < P < Pcr) the effect of the
force in the spring predominates and the structure returns to the
vertical position after a slight disturbance; Thus, the structure is in
equilibrium only when it is perfectly straight (θ = 0).
If the axial load is larger than Pcr, i.e. (P > Pcr) the effect of the
axial force predominates and the structure buckles:
If P < Pcr , the structure is stable
If P > Pcr , the structure is unstable
The stability of the structure is increased either by increasing its
stiffness or by decreasing its length.
At the critical load (P = Pcr), the structure is in equilibrium even
when point A is displaced laterally by a small amount. In other words,
the structure is in equilibrium for any small angle θ, including θ = 0.
However, the structure is neither stable nor unstable—it is at the
boundary between stability and instability. This condition is referred
to as neutral equilibrium.
Pcr = kL/4 Neutral Equilibrium

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 8

 lOMoARcPSD|27577562

5/29/2016

The three equilibrium conditions

for the idealized structure are shown in
the graph of axial load P versus angle
of rotation θ. The two heavy lines,
one vertical and one horizontal,
represent the equilibrium conditions.
The point where the equilibrium
diagram branches, is called a
bifurcation point. At this transition
point the applied load is equal to the
critical load.
Fig. Equilibrium diagram for
The horizontal line for neutral buckling of an idealized structure
equilibrium extends to the left and
right of the vertical axis because the angle θ may be clockwise or
counterclockwise. The line extends only a short distance, however,
because the analysis is based upon the assumption that θ is a small
angle.

Stable equilibrium

Unstable equilibrium
Pcr = kL/4 Neutral Equilibrium
A critical load, therefore, can be interpreted as the maximum axial
load to which a column can be subjected and still remain straight,
although in such an unstable condition that a slight sideways thrust
will cause it to bow out.

Ideal Column
The column to be considered is an ideal column, meaning that -
is perfectly straight before loading and have no imperfections.
is made of homogeneous material with constant cross-section.
is subjected to axial compressive load (or upon which the load is
applied through the centroid of the cross-section)
the material behaves in a linear-elastic manner and that the column
buckles or bends in a single plane.

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 9

 lOMoARcPSD|27577562

5/29/2016

Real/Actual Column
In reality, the conditions of column straightness and load application
are never accomplished. Actual column always have imperfections of
material and fabrication, as well as unavoidable accidental eccentricities
of load which produce the effect shown, greatly exaggerated.
The initial crookedness of the column,
together with the placement of the load,
Unavoidable
causes an indeterminate eccentricity e or accidental
with respect to the centroid of a typical eccentricity
section m-n. The loading on this section is Actual central line due
similar to that on an eccentrically loaded to initial crookedness
(greatly exaggerated)
short strut and the resultant stress is due to
a combination of a direct compressive e = eccentricity of P at
a typical section m-n
stress and a flexure stress.
However, the analysis to be
performed on an “ideal column” is Mathematically
straight center line
similar to that used to analyze initially
crooked columns or those having an
eccentric load application.

Ideal column with pin supports

An ideal column is straight and theoretically the axial load P could be
increased until failure occurs either by fracture or yielding of the material.
When the critical load Pcr is reached, the column will be on the verge of
becoming unstable, so that a small lateral force F (fig. 4b), will cause the
column to remain in the deflected position when F is removed (fig. 4c).
Any slight reduction of axial load P from Pcr will allow the column to
straighten out (fig. 4a), and any slight increase in P, beyond Pcr , will
cause further increases in lateral deflection.

Whether or not a column will remain

stable or become unstable when
subjected to an axial load will depend
on its ability to restore itself, which is
based on its resistance to bending.

Fig. 4a Fig. 4b Fig. 4c

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 10

 lOMoARcPSD|27577562

5/29/2016

Since a column can be considered as a beam placed in a vertical

position and subjected to an axial load, so in order to determine the
critical load and buckled shape of column, the relationship between the
internal moment in the column and its deflected curved shape can be
applied i.e.
Moment-curvature equation

Fig. 4e

Fig. 4d
B&J

In the fig. 4d, x denotes the distance from end A of the column to a
given point Q of its elastic curve, and by y the deflection of that
point. It follows that the x-axis will be vertical and directed
downward, and the y-axis horizontal and directed to the right.
Considering the equilibrium of the free body AQ (fig. 4e), it is found
that the bending moment at Q is -
Equating moments at the cut end;

M(x) = -Py and so

This equation simplifies to:

This is a linear, homogeneous, second order differential equation

with constant coefficient P/EI .

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 11

 lOMoARcPSD|27577562

5/29/2016

2 P
Consider,  
EI
So the equation can be written in the form,
d2y P
  2 y  0; where  2 
dx 2 EI

This linear, homogeneous, second order differential equation is

similar as that of the differential equation for simple harmonic motion
except that the independent variable is now the distance x instead of
time t.
The general solution of this equation is:
y = A sin αx + B cos αx ------ (1)
where A and B are found using boundary conditions.
Boundary Conditions are:
(i) At x = 0, y = 0, and
(ii) At x = L, y = 0,
From the first boundary condition: at x = 0, y = 0, we get B = 0
From the second boundary condition: at x = L, y = 0, we get
A sin  L  0
means, either A  0 or , sin  L  0

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 12

 lOMoARcPSD|27577562

5/29/2016

If A = 0, then equation (1) reduces to y = 0. This is a trival solution

which requires that the column is always straight, even though the
load may cause the column to become unstable and no bending
moment exists. It gives the behavior of an ideal column that is in
equilibrium (either stable or unstable) in the straight position (no
deflection) under the action of the compressive load P.
For any buckling to happen, the second condition has to be true
and that means, so the only logical solution is
sin(L)=0
and the only way that can happen is if L= n

------ (2)

The critical load is then

This formula gives the values of P that satisfy the buckling equation
and provide solutions (other than the trivial solution) to the
differential equation. For the values of P given by Eq. (2) is
theoretically possible for the column to have a bent shape. For all
other values of P, the column is in equilibrium only if it remains
straight. Therefore, the values of P given by Eq. (2) are the critical
loads for this column.

Effect of n on loads

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 13

 lOMoARcPSD|27577562

5/29/2016

The values of “n” defines the buckling mode shapes.

P1 P1
n=1 First mode of buckling P1 
 2 EI
L2

P2 P2
n=2
4 2 EI
Second mode of buckling P2 
L2
P3 P3
n=3
2
9 EI
Third mode of buckling P3 
L2

Fig. 4f: First three modes of buckling loads

Critical Buckling Load

However, since P1< P2< P3, the column buckles at P1 and never gets to
P2 or P3 unless bracing is placed at the points where y = 0 to prevent
buckling at lower loads.

The smallest value of P is obtained when n = 1 and is the lowest

critical load (or Euler Buckling Load) for the column. The critical
load for a pin ended column is therefore:
-------- (3)

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 14

 lOMoARcPSD|27577562

5/29/2016

The type of buckling described in this section is called Euler buckling.

Substituting the value of P into equation (1),we can write

Buckling Mode

This is the equation of elastic curve after the column has buckled.
The function that represents the shape of the deflected column is called a
mode shape, or buckling mode.
The value of maximum deflection, ym = A which determines the
direction (sign) and amplitude of the deflection, is indeterminate or
arbitrary, but it must be small. This is due to the fact that the governing
differential equation is a linearized approximation (slope very small) of
the actual governing differential equation for the elastic curve.

In case of a column with a circular cross-section or square cross-

section, the moment of inertia I of the cross-section is the same about
any centroidal axis, and the column is as likely to buckle in one plane as
another, except for the restraints that can be imposed by the end
connections. For other shapes of cross-section, the critical load
should be computed by making I = Imin in equ. (3); if buckling occurs,
it will take place in a plane perpendicular to the corresponding
principal axis of inertia.
Since Pcr is proportional to I, the column will buckle in the
direction corresponding to the minimum value of I as shown in the
following fig.

Fig. Column cross section showing

the direction of buckling

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 15

 lOMoARcPSD|27577562

5/29/2016

Critical Column Stress

A column can either fail due to the material yielding, or because the
column buckles, it is of interest to the engineer to determine when this
point of transition occurs.
 2 EI
Consider the Euler buckling equation: PE 
L2
PE 2 E
Dividing the buckling equation by A, gives: E  
A L / r 2
Where σE is the compressive stress in the column and must not
exceed the proportional limit stress σP of the material, i.e. σE < σP , and
PE/A is the average stress in the column when carrying its critical load.
L/r is called the slenderness ratio, it is a measure of the column's
flexibility.
The value of stress corresponding to the critical load is called
critical stress, σcr
-------- (4)

Where, I = Ar2
A = Cross-sectional area and r I
r = least radius of gyration A
(the minimum value of radius of gyration must be used)
Equ. (4) shows that the critical stress is proportional to the modulus of
elasticity of the material, and inversely proportional to the square of
the slenderness ratio of the column. The plot of σcr versus L/r is shown
in the following fig. for structural steel.
For structural steel, assume E =
200 Gpa and σY = 250 Mpa. No
factor of safety has been used in
plotting σcr . If the value obtained
for σcr > σY , this value is of no
interest since the column will
yield in compression and ceases
to be elastic before it has a
chance to buckle. Euler formula
is not valid in this case.

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 16

 lOMoARcPSD|27577562

5/29/2016

Limitations of Euler Formula

A column always tends to buckle in its most limber direction. For
this reason, and since flexural resistance varies with moment of inertia,
the values of I in the column formulas is always the least moment of
inertia of the cross-section. Any tendency to buckle, therefore, occurs
about the least axis of inertia of the cross-section.
Euler formula also shows that the critical load that causes buckling
depends not on the strength of the material but only on its
dimensions, and material property i.e. modulus of elasticity. For this
reason, dimensionally two identical slender struts, one of high-strength
steel and the other of ordinary structural steel will buckle under the
same critical load because, although their strength are different, they
have the same modulus of elasticity. One way to increase the elastic
buckling load of a member would be to use a member that is made of
material with a higher E value.
Good design also requires that a section have as large a moment
of inertia as possible. This can be done, without increasing the cross-
sectional area, by using thin-wall tubular members as shown in the

following figs. Hence, for a given area, the material should be

distributed as far as possible from the centroid and in such a way that
a moment of inertia about the principal axes are equal or as
nearly equal as possible. However, if the column wall is too thin,
local buckling can occur.

Fig. Column cross-sections.

In order for Euler formula to be applicable, the buckling stress

must not exceed the proportional limit.
Finally, the Euler formula determines the critical loads, not
working loads. It is therefore, necessary to divide the right side of
each formula by a suitable factor of safety --- usually 1.7 to 2.5,
depending on the material --- in order to obtain practical allowable
stress.

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 17

 lOMoARcPSD|27577562

5/29/2016

The limiting slenderness ratio

Conventionally, we defines long columns as those for which Euler
formula applies. The limiting slenderness ratio that fixes the lower
limit for Euler formula is easily found by substituting in equ. (4) the
known values of the proportional limit and the modulus of elasticity of
the specified material. This limiting ratio varies with different
materials and even with different grades of the same material.
Example: Steel has a proportional limit of 200 Mpa and for which
E = 200 Gpa, the limiting slenderness ratio is

Below this value of L/r, as shown in the above fig. by the dashed
portion of Euler’s curve, the Euler unit load exceeds the proportional
limit. Hence for L/r < 100, Euler's formula is not valid, and the
proportional limit is taken as the critical stress.
The curve also shows that the critical or allowable stress on a
column decreases rapidly as the slenderness ratio increases. If the

slenderness ratio is very large, say L/r > 100, the stress at buckling will
be very small. Therefore, the strength of the material is underutilized.
The design should be modified by, for example, adding lateral bracing
or changing the boundary conditions . Hence, it is good design to
keep the slenderness ratio as small as possible.

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 18

 lOMoARcPSD|27577562

5/29/2016

Fig: Graphs of Euler’s

formula for structural steel
and for an aluminum alloy.

Problem
A 2m long pin-ended column of square cross section is to be made of
wood. Assuming E =12.5 GPa, σallow=12MPa for compression parallel to
the grain, and using a factor of safety of 2.5 in computing Euler’s critical
load for buckling, determine the size of the cross section if the column is
to safely support (a) P = 100kN load and (b) P = 200kN load.

Solution
(a) For the 100-kN Load
Buckling criterion
Using given Factor of Safety
The required critical load is

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 19

 lOMoARcPSD|27577562

5/29/2016

Based on Euler’s formula, we have

For a square cross-section with side s

Stress criterion

[ Considering 99.3 mm ≈ 100 mm]

Since σ < σallow , a 100 x 100-mm cross-section is acceptable.

(a) For the 200-kN Load

x 2 = 15.588 x 10-6 m4

The value of normal stress in the column is:

Since σ > σallow , the dimension obtained is not acceptable, and the
cross-section must be selected on the basis of its resistance to
compression.

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 20

 lOMoARcPSD|27577562

5/29/2016

Columns With Other Types Of Support (Or, End) Conditions

Euler load was derived for a column that is pin supported or free to rotate
at its end. However, in practice, column may be supported in some other
way such as fixed ends, free ends, and elastic supports. The critical loads
for columns with various kinds of support conditions can be determined
from the differential equation of the deflection curve as used when
analyzing a pinned-end column.

The procedure is as follows:

First, with the column assumed to be in the buckled state, obtain an
expression for the bending moment in the column.
Second, we set up the differential equation of the deflection curve,
using the bending-moment equation.
Third, solve the equation and obtain its general solution, which
contains two constants of integration plus any other unknown
quantities.
Fourth, apply boundary conditions pertaining to the deflection and
the slope and obtain a set of simultaneous equations.
Finally, we solve those equations to obtain the critical load and the
deflected shape of the buckled column.

The critical load for other columns can be expressed in terms of the
critical buckling load for a pin- ended column.

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 21

 lOMoARcPSD|27577562

5/29/2016

Column Fixed at the Base and Free at the Top

The first case, consider is an ideal column that is fixed at the base,
free at the top, and subjected to an axial load P (Fig. 5a). The deflected
shape of the buckled column is shown in Fig. 5b. From this figure we see
that the bending moment at distance x from the base is-
Where δ is the deflection at the free
------- (1)
end of the column.
So, the differential equation of the
deflection curve then becomes

------- (2)
[ Consider ν instead of y,
deflection ]
Fig. 5a Fig. 5b The cut section of the
upper part of the column

Using the notation k2 = P/EI in eq. 2, we can rearrange the eq. into the
form

------- (3)

which is a nonhomogeneous (because of non-zero term on the right side)

linear differential equation of second order with constant coefficients.
The general solution of Eq. (3) consists of two parts:
(1) the homogeneous (complementary) solution, which is the solution
of the homogeneous equation obtained by replacing the right-hand
side with zero, and
(2) the particular solution, which is the solution of Eq. (3) that produces
the term on the right-hand side.
------- (4)

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 22

 lOMoARcPSD|27577562

5/29/2016

The homogeneous solution (also called the complementary solution)

is the same as equ. of pin ended column y = A sin αx + B cos αx

where C1 and C2 are constants of integration. Note that when vH is

substituted into the left-hand side of the differential equation (Eq. 3),
it produces zero.
The particular solution of the differential equation is

When vP is substituted into the left-hand side of the differential

equation, it produces the right-hand side, that is, it produces the term
k2δ. Consequently, the general solution of the equation, equal to the
sum of vH and vP, is
------- (4)
This equation contains three unknown quantities (C1 , C2 , and δ),
and therefore three boundary conditions are needed to complete
the solution.

At the base of the column, the deflection and slope are each equal to
Zero i.e.,
At x = 0, ν = 0 , so C2 = - δ and
At x = 0, dv/dx = 0 and so C1 = 0

The deflection curve is therefore:

Or, ------- (5)

Again at x = L, ν = δ, i.e. v(L) = δ

Using this condition we get
Or, ------- (6)

From this equation we conclude that either δ = 0 or cos kL = 0.

If δ = 0, there is no deflection of the bar and we have the trivial
solution—the column remains straight and buckling does not occur.
In that case, Eq. (6) will be satisfied for any value of the quantity kL, that
is, for any value of the load P. This conclusion is represented by the
vertical line in the load-deflection diagram.

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 23

 lOMoARcPSD|27577562

5/29/2016

The other possibility for solving Eq. (6) is

Or,

which is the buckling equation. In this case, Eq. (6) is satisfied

regardless of the value of the deflection δ. Thus, as already observed, δ
is undefined and may have any small value.
The equation cos kL = 0 is satisfied when

Using the expression k2 = P/EI, we obtain the following formula for

the critical loads:
------- (7)

Also, the buckled mode shapes are obtained from Eq. (5):

Fig: Ideal column fixed at the base and free at the top: (a) initially straight
column, (b) buckled shape for n = 1, (c) buckled shape for n = 3, and
(d) buckled shape for n = 5.

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 24

 lOMoARcPSD|27577562

5/29/2016

The lowest critical load is obtained by substituting n =1 in Eq. (7):

The corresponding buckled shape is

So, a column fixed supported at its base, free at its top, will
support only one-fourth the critical load that can be applied to a
column pin-supported at both ends.

Effective Lengths of Columns

The critical loads for columns with various support conditions can be
related to the critical load of a pinned-end column through the concept
of an effective length (Le).
The effective length of a column is the distance between points of
inflection (that is, points of zero moment) in its deflection curve,
assuming that the curve is extended (if necessary) until points of inflection
are reached.

The effective length Le for any column is the length of the

equivalent pinned-end column, that is, it is the length of a pinned-end
column having a deflection curve that exactly matches all or part of
the deflection curve of the original column.

Example: Consider the deflected

shape of a column fixed at the base
and free at the top (Fig. a). This
column buckles in a curve that is one-
quarter of a complete sine wave. If we
extend the deflection curve (Fig. b), it
becomes one-half of a complete sine
wave, which is the deflection curve for
a pinned-end column.
Thus, for a fixed-free column,
effective length is Le = 2L

Or,

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 25

 lOMoARcPSD|27577562

5/29/2016

We can write a general formula for critical loads as follows:

The effective length is often expressed in terms of an effective

length factor K:
Le =KL
where L is the actual length of the column.
Thus, the critical load is

Therefore:  L pin  pin

0.7 L fixed  pin

Le  
0.5 L fixed  fixed
 2 L fixed  free

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 26

 lOMoARcPSD|27577562

5/29/2016

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 27

 lOMoARcPSD|27577562

5/29/2016

Column with Both Ends Fixed Against Rotation

Fig: Buckling of a column with

both ends fixed against rotation

 2 EI
Pcr 
L2e

Le = L/2

Fig: FBD of the column

 2 EI  2 EI
Pcr  
L / 2 2 L2 4

Fig: Comparison with pin-ended column

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 28

 lOMoARcPSD|27577562

5/29/2016

Column Fixed at the Base and Pinned at the Top

Reactive moment,
The bending moment in the buckled column, at distance x from the base

 2 EI
Pcr 
L2e

So, the differential equation of the deflection

curve then becomes

------- (1)

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 29

 lOMoARcPSD|27577562

5/29/2016

Consider,

Then the differential equation of the deflection curve then becomes

------- (1)

This equation is a linear, non-homogeneous differential equation of

the second order with constant coefficients. A particular solution is
easily seen to be

The general solution is ------- (2)

At the base of the column, the deflection and slope are each equal to
Zero i.e.,
At x = 0, y = 0 , so B = 0 and
At x = L, dy/dx = 0

At x = L, y = 0 ------- (3)

Finally, computing

At x = L, dy/dx = 0 ------- (4)

Dividing (3) by (4) ------- (5)

Solving this equation by trial and error, we find that the smallest
value of pL which satisfies (5) is

So, we get from

The effective length of the column is obtained by

The effective length of a column with one fixed end and one pin-
connected end is Le = 0.699L ≈ 0.7L

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 30

 lOMoARcPSD|27577562

5/29/2016

Problem
An aluminum column of length L and rectangular cross-section has a
fixed end at B and supports a centric load at A. Two smooth and rounded
fixed plates restrain end A from moving in one of the vertical planes of
symmetry but allow it to move in the other plane.
a) Determine the ratio a/b of the two sides
of the cross-section corresponding to
the most efficient design against
buckling.
b) Design the most efficient cross-section
for the column.

Knowing that,
L = 20 in., E = 10.1 x 106 psi
P = 5 kips and factor of safety FS = 2.5
are required.

Solution
The most efficient design occurs when the resistance to buckling is
equal in both planes of symmetry i.e. the critical stresses corresponding
to the two possible modes of buckling are equal. This occurs when the
slenderness ratios are equal.
Buckling in xy Plane Buckling in xz Plane

The effective length of the The effective length of the

column with respect to buckling column with respect to buckling
in this plane Le = 0.7L (one in this plane Le = 2L (one
fixed-one pinned end) fixed-one free end)
1 3
2 I y 12 ab b2 b
ry    ry 
A ab 12 12
1 ba3
Le, y 2L
I a2 a 
rz2  z  12  rz  ry b / 12
A ab 12 12
Le, z 0.7 L

rz a 12

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 31

 lOMoARcPSD|27577562

5/29/2016

(a) Most efficient design: (b) Most efficient design:

The most efficient design is that
for which the critical stresses Le 2L 220 in  138.6
corresponding to the two possible   
ry b 12 b 12 b
modes of buckling are equal.
Pcr  FS P  2.55 kips   12.5 kips
Pcr 12500 lbs
 cr  
This will be the case if the two A 0.35b b
values Obtained for the effective
slenderness ratio are equal.  cr 
 2E


 2 10.1  106 psi 
Le r 2 138.6 b 2
Le, z Le, y
rz

ry 
12500 lbs  2 10.1  106 psi


0.35b b 138.6 b 2
0.7 L 2L
 b  1.620 in.
a 12 b / 12
a a  0.35b  0.567 in.
a 0.7
  0.35
b 2 b

Long column can be treated by Euler’s formula provided that the

slenderness ratio is larger than the value at which the average stress
reaches the proportional limit. For hinged steel columns, this limit is L/r
≈ 100 at 200 Mpa. Euler formula is not valid for smaller slenderness
ratio.

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 32

 lOMoARcPSD|27577562

5/29/2016

The figure shows that steel

have a yield point of 280
MPa and proportional limit
of 200 MPa.

A short column is one whose length does not exceed 10 times the
least lateral dimension, sets the upper limit of the slenderness ratio at
about 30 for rectangular section. For practical purposes, the limiting
stress on a short column found to be the stress at yield point.

Columns With Eccentric Axial Loads

In reality, the conditions of column straightness and load application are
never accomplished. Actual column always have imperfections of
material and fabrication, as well as unavoidable accidental
eccentricities of load which produce the effect shown, greatly
exaggerated. So, in reality, columns never suddenly buckle; instead they
begin to bend, although ever so slightly, immediately upon application of
the load.
Consider that a column is compressed by load P applied with a small
eccentricity e measured from the axis of the column (Fig. 7a). This
eccentric axial load is statically equivalent to a centric load P and a
couple of moment M0 = Pe (Fig. 7b). This moment exists from the
instant the load is first applied, and It is clear that no matter how small
the load P and the eccentricity e, the couple M0 will cause some
bending of the column and therefore the column begins to deflect at
the onset of loading. The deflection then becomes steadily larger as the
load increases. As the eccentric load is increased, both the couple M0
and axial force P increase, and both cause the column to bend further.

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 33

 lOMoARcPSD|27577562

5/29/2016

(Fig. 7a) (Fig. 7a)

Fig: Column with eccentric axial loads.

Assumptions
The column is initially perfectly straight,
The material is linearly elastic, and
The xy plane is a plane of symmetry for the cross-sectional area.
Ends A and B are supported so that they are free to rotate (pin
supported.)
The slope and deflections are small.

From the FBD of any arbitrary section,

the bending moment in the column at
point Q is (Fig. 7c)

M = M0 + P(-y) = Pe - Py

where y is the deflection of the column

(positive when in the positive direction of
the y axis). Note that the deflections of
the column are negative when the
eccentricity of the load is positive.
(Fig. 7c)

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 34

 lOMoARcPSD|27577562

5/29/2016

The differential equation of the deflection curve is

This is a nonhomogeneous (because of non-zero term on the right side)

linear differential equation of second order with constant coefficients.
Consider,
The general solution of this equation is (consisting of a complementary
and particular solution)

in which C1 and C2 are constants of integration in the

homogeneous solution and e is the particular solution.
The boundary conditions are
(i) At x = 0, y = 0, and This condition yield C2 = -e
(ii) At x = L, y = 0,

Therefore, the equation of the deflection curve is

------ (1)

For a column with known loads P and known eccentricity e, this

equation can be used to calculate the deflection at any point along the
x axis.
The behavior of a column with an eccentric load is quite different
from that of a centrally loaded column. The above eccentric deflection
equation shows that each value of the eccentric load P produces a
definite value of the deflection, just as each value of the load on a
beam produces a definite deflection. In contrast, the deflection
equations for centrally loaded columns give the buckled mode shape
(when P = Pcr) but with the amplitude undefined.
Because the column shown in Fig. 7 has pinned ends, its critical load
(when centrally loaded) is

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 35

 lOMoARcPSD|27577562

5/29/2016

Maximum Deflection
The maximum deflection δ produced by the eccentric loads occurs at
the midpoint of the column (Fig. 7d) and is obtained by setting x
equal to L/2 in Eq. (1):

------ (2)

This equation can be written in a slightly

different form by replacing the quantity k with
its equivalent in terms of the critical load

Fig. 7d

the maximum deflection becomes

------ (3)
Special cases
1. The deflection δ is zero when the eccentricity e is zero and P is not
equal to Pcr,
2. The deflection is zero when the axial load P is zero,
3. The deflection becomes infinitely large as P approaches Pcr.
These characteristics are shown in the load-deflection diagram.
To plot the load-deflection diagram, select a particular value e1 of the
eccentricity and then calculate δ for various values of the load P. The
resulting curve is labeled e = e1 in Fig. 7e. Note immediately that the
deflection δ increases as P increases, but the relationship is nonlinear.
Therefore, the principle of superposition cannot be used for calculating
deflections due to more than one load, even though the material of the
column is linearly elastic. As an example, the deflection due to an axial
load 2P is not equal to twice the deflection caused by an axial load P.

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 36

 lOMoARcPSD|27577562

5/29/2016

Additional curves, such as the curve

labeled e = e2, are plotted in a similar
manner. Since the deflection δ is linear
with e (Eq. 2), the curve for e = e2 has the
same shape as the curve for e = e1 but the
abscissas are larger by the ratio e2/e1.
As the load P approaches the critical
load, the deflection δ increases without
limit and the horizontal line corresponding
to P = Pcr becomes an asymptote for the
curves. In the limit, as e approaches zero, Fig 7e: Load-deflection
the curves on the diagram approach two diagram for a column with
straight lines, one vertical and one eccentric axial loads
horizontal (compare with Fig. 7f). Thus, as
expected, an ideal column with a
centrally applied load (e = 0) is the
limiting case of a column with an
eccentric load (e > 0).

Although the curves plotted in Fig. 7e

are mathematically correct, but the
differential equation is valid only for
small deflections. Therefore, when the
deflections become large, the curves are no
longer physically valid and must be
modified to take into account the presence
of large deflections and (if the proportional
limit of the material is exceeded) inelastic
Fig 7f: Load-deflection diagram for
an ideal, linearly elastic column.
bending effects.

The reason for the nonlinear relationship between loads and deflections
even when the deflections are small and Hooke’s law holds

The axial loads P are equivalent to centrally applied loads P plus couples
Pe acting at the ends of the column. The couples Pe, if acting alone,
would produce bending deflections of the column in the same manner as
for a beam. In a beam, the presence of the deflections does not change the

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 37

 lOMoARcPSD|27577562

5/29/2016

action of the loads, and the bending moments are the same whether the
deflections exist or not. However, when an axial load is applied to the
member, the existence of deflections increases the bending moments
(the increases are equal to the product of the axial load and the
deflections). When the bending moments increase, the deflections are
further increased—hence the moments increase even more, and so on.
Thus, the bending moments in a column depend upon the
deflections, which in turn depend upon the bending moments. This
type of behavior results in a nonlinear relationship between the
axial loads and the deflections.
In general, a straight structural member subjected to both bending
loads and axial compressive loads is called a beam-column. In the case
of a column with eccentric loads, the bending loads are the moments
M0 = Pe and the axial loads are the forces P.
Maximum Bending Moment
The maximum bending moment in an eccentrically loaded column
occurs at the midpoint where the deflection is a maximum

From equ. (2) aand (3):

----- (4)

When P is small, the maximum moment is

equal to Pe, which means that the effect of the
deflections is negligible. As P increases, the
bending moment grows nonlinearly and
theoretically becomes infinitely large as P
approaches the critical load. But the
equations are valid only when the deflections
are small, and they cannot be used when the
axial load approaches the critical load.
Fig 7g: Maximum bending Nevertheless, the preceding equations and
moment in a column with accompanying graphs indicate the general
eccentric axial loads
behavior of beam-columns.
Other End Conditions
If a column is fixed at the base and free at the top, Equs. (3) and (4)
can be used by replacing the actual length L with the equivalent length 2L.

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 38

 lOMoARcPSD|27577562

5/29/2016

However, the equations do not apply to a column that is fixed at

the base and pinned at the top. The use of an equivalent length equal
to 0.699Lgives erroneous results; instead, we must return to the
differential equation and derive a new set of equations.
In the case of a column with both ends fixed against rotation, the
concept of an eccentric axial load acting at the end of the column
has no meaning. Any moment applied at the end of the column is
resisted directly by the supports and produces no bending of the
column itself.
The Secant Formula For Columns
The maximum stresses in a column with eccentric axial loads occur
at the cross section where the deflection and bending moment have
their largest values, that is, at the midpoint.
The stresses due to the force P are equal to P/A, where A is the
cross-sectional area of the column, and the stresses due to the
bending moment Mmax are obtained from the flexure formula.

Thus, the maximum compressive stress, which occurs on the

concave side of the column, is

Where I is the moment of inertia in the plane of bending and

c is the distance from the centroidal axis to the extreme
point on the concave side of the column.
Note that in this equation compressive stress is considered to be positive,
since these are the important stresses in a column.

Since for pin-ended column,

and
Where r is the radius of gyration in the plane of bending and so

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 39

 lOMoARcPSD|27577562

5/29/2016

The maximum compressive stress:

This is the Secant formula for an

eccentrically loaded column with
pinned ends.
The secant formula gives the maximum
Fig. 7g: Column with
compressive stress in the column as a function of Eccentric axial loads
1. the average compressive stress P/A,
2. the modulus of elasticity E, and
3. Two nondimensional ratios
— the slenderness ratio L/r and
— the eccentricity ratio:

The eccentricity ratio is a measure of the eccentricity of the load as

compared to the dimensions of the cross section. Its numerical value
depends upon the position of the load, but typical values are in the
range from 0 to 3, with the most common values being less than 1.
A graph of the secant formula is shown in Fig. 7h. Curves are
plotted for several values of the eccentricity ratio ec/r2. These curves
are valid only when the maximum stress is less than the
proportional limit of the material, because the secant formula was
derived using Hooke’s law.
Special case
When eccentricity of the load disappears (e = 0), then it becomes
an ideal column with a centrally applied load. Under these conditions
the maximum load is the critical load,

and the corresponding maximum

stress is the critical stress
It is the Euler’s curve.

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 40

 lOMoARcPSD|27577562

5/29/2016

The proportional limit of

the material is the same as the
selected maximum stress, (in
the graph it is, 36 ksi). The
horizontal line on the graph
shows the limit.
The horizontal line and
Euler’s curve represent the
Fig 7h: Graph of the secant
limits of the secant-formula
Formula for a steel column.
curves as the eccentricity e
approaches zero.

The graph of the secant formula shows that the load-carrying capacity of a
column decreases significantly as the slenderness ratio L/r increases,
especially in the intermediate region of L/r values. Thus, long slender columns
are much less stable than short, stocky columns. For small value of Le/r, the
secant is almost equal to 1 and P/A can be assumed equal to

On the other hand, for large values of Le/r, the curves corresponding to the
various values of the ratio ec/r2 get very close to the Euler’s curve and thus the
effect of eccentricity of the loading on the value of P/A become negligible. The
Secant formula is chiefly useful for intermediate values of Le/r.
The graph also shows that the load-carrying capacity decreases with
increasing eccentricity e; furthermore, this effect is relatively greater for short
columns than for long ones.
Secant formula can also be used for a column that is fixed at the base and
free at the top. All that is required is to replace the length L in the secant
formula with the equivalent length 2L.
The secant formula is not valid for the other end conditions.

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 41

 lOMoARcPSD|27577562

5/29/2016

Problem:
The uniform column AB consists of an 8-ft section of structural tubing
having the cross-section shown.
a) Using Euler’s formula and a factor of safety
of two, determine the allowable centric load
for the column and the corresponding
normal stress.
b) Assuming that the allowable load, found in
part a, is applied at a point 0.75 in. from the
geometric axis of the column, determine the
horizontal deflection of the top of the
column and the maximum normal stress in
the column.

E  29  106 psi.

Effective Length
Since the column has one end fixed and
one end free, its effective length is

Critical load
Using the Euler’s formula

(a) Allowable Centric load and Stress

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 42

 lOMoARcPSD|27577562

5/29/2016

(b) Eccentric load:

Horizontal deflection at point A or end deflection

The maximum normal stress in the column

Intermediate Column: Empirical Formulas

Various methods have been proposed for bridging the gap between
the short column range and the long column range. However, none
of them has been accepted universally for intermediate columns, partly
because of their departure from the stress-strain relationship when the
stresses exceed the proportional limit, and partly because of their
indeterminate mixture of direct and flexural stresses.

Tangent-Modulus Theory
The Euler formula is extended to intermediate columns stresses above
the proportional limit by replacing the constant modulus E by a tangent
modulus Et (also called effective modulus), namely
tangent-modulus load The corresponding
(or critical load): critical stress is

The tangent modulus Et is obtained by using the slope of the tangent to

the stress-strain diagram at the point corresponding to the average stress
in the column. This yields a curve that connects the curve in fig. 8
representing the short and long column formulas.

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 43

 lOMoARcPSD|27577562

5/29/2016

The compressive stress-strain diagram

for the material of the column is
shown in Fig. The proportional limit
of the material is indicated as σp1 , and
the actual stress σA in the column
(equal to P/A) is represented by point
A (which is above the proportional
limit). If the load is increased, so that
a small increase in stress occurs, the
relationship between the increment of
stress and the corresponding
Fig 8: Compression stress-strain increment of strain is given by the
diagram for the material of the column slope of the stress-strain diagram at
Note that the tangent modulus point A. This slope, equal to the slope
decreases as the stress increases of the tangent line at A, is called the
beyond the proportional limit. When tangent modulus and is denoted by Et
the stress is below the proportional ; thus,
limit, the tangent modulus is the same
as the ordinary elastic modulus E.

Although this method is empirical because it violets the stress-strain

proportionality assumed in the derivation of Euler’s formula, actual
tests show close agreement with the theoretical curve.

Fig: Critical stress versus slenderness ratio for aluminum columns.

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 44

 lOMoARcPSD|27577562

5/29/2016

Other methods are empirical. In general, empirical equations

are either linear or non-linear depending on the best agreement that
can be found with data from tests of real columns. Modern column
studies which include the effects of residual stresses, will undoubtedly
lead to a more complete understanding of the behaviour of
intermediate column.
American Institute of Steel Construction (AISC) defines the
limit between the intermediate and long columns to be the value of the
slenderness ratio Cc given by

Where, the modulus of elasticity, E is 200 Gpa or 29 x 106 psi for

most grades of steel and σyp is the yield stress for the particular
grade of steel being used.

For columns of effective length Le and minimum radius of gyration

r, AISC specifies that for Le/r > Cc , the working stress, σw is given by

Note that this is Euler’s formula with a factor of safety 23/12 = 1.92.
For columns Le/r < Cc , AISC specifies the parabolic formula

Where the factor of safety, FS is given by

Observe that the factor of safety is 1.92 at Le/r = Cc and it becomes

smaller for larger values of the slenderness ratio.

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 45

 lOMoARcPSD|27577562

5/29/2016

Fig: the variation of σw with

Le/r for several grades of
steel in SI unit.

J.B. Johnson Formula

One of the semi-empirical formulas is the J.B. Johnson Formula in the
intermediate length range. We will not derive this formula, but make
several comments. The J.B. Johnson formula is the equation of a
parabola with the following characteristics. For a graph of stress versus
slenderness ratio, the parabola has its vertex at the value of the yield
stress on the y-axis. Additionally, the parabola is tangent to the Euler
curve at a value of the slenderness ratio, such that the
corresponding stress is one-half of the yield stress.
In the diagram below, we have a steel member with a yield stress of
40,000 psi. Notice the parabolic curve beginning at the yield stress and
arriving tangent to the Euler curve at 1/2 the yield stress.
We first note that at the point where the Johnson formula and Euler's
formula are tangent, we can relate the stress to Euler's formula as
follows (where C represents the slenderness ratio when the stress is 1/2
the yield stress):

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 46

 lOMoARcPSD|27577562

5/29/2016

From this we find an expression for

C (critical slenderness ratio) of:

For our particular case, where we have a steel member with a yield
stress of 40,000 psi, and a Young's modulus of 30 x 106 psi., we find
C = sqrt(2 * 3.142 * 30 x 106 / 40,000 psi) = 122. If our actual beam
has a slenderness ratio greater than the critical slenderness ratio
we may use Euler’s formula. If on the other hand our actual
slenderness ratio is small than the critical slenderness ratio, we
may use the J.B. Johnson Formula.
Example: As an example let us now take a 20 foot long W12 x 58
steel column (made of same steel as above), and calculate the critical
stress using the J.B. Johnson formula. (Beam information and Johnson
formula shown below.)

- - - Flange Flange Web Cross Section Info. Cross Section Info.

Designat
ion Area Depth Width thick thick x-x axis x-x axis x-x axis y-y axis y-y axis y-y axis
- A-in2 d - in wf - in tf - in tw - in I - in4 S -in3 r - in I - in4 S -in3 r - in

W 12x58 17.10 12.19 10.014 0.641 0.359 476.0 78.1 5.28 107.00 21.40 2.51

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 47

 lOMoARcPSD|27577562

5/29/2016

J.B. Johnson's formula:

For our beam the slenderness ratio = (20 ft * 12 in/ft)/2.51in = 95.6

(where 2.51 in. is the smallest radius of gyration, about y-y axis).
Inserting values we find:
Critical Stress = [ 1 - (95.62/2* 1222)]*40,000 psi. = 27,720 psi.
This is the critical stress that would produce buckling. Note we did
not have a safety factor in this problem. As a result we really would
not want to load the column to near the critical stress, but at a lower
'allowable' stress.
The Critical Load will equal the product of the critical stress and
the area, or Pcr = 27,720 psi. * 17.10 in2 = 474,012 lb.

The Aluminium Association. Inc., lists column specification for

each of the various types of aluminium alloy. In their specifications, the
allowable stress is a constant for short columns, a straight line relation
approximating the tangent-modulus formula is used for long columns.
For example the specifications for 2014-T6 aluminium alloy are

The column length L in the specifications is defines as the “length of

the compression member between points of lateral supports, or twice
the length of a cantilever column (except where analysis shows that a
shorter length can be used)”.

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 48

 lOMoARcPSD|27577562

5/29/2016

For wooden columns, the National Lumber Manufacturers

Associations recommends Euler’s formula in the following form:

Here, load duration and moisture content adjustments must also be made.
For rectangular column with least lateral dimension d, r = √(b/12) and
the above equation reduces to

All the empirical equations reduce the safe working load as

the slenderness ratio increases, although in varying proportions.
Because of variations in design codes, the same column may,
therefore, support any one of the several safe legal loads.

Problem

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 49

 lOMoARcPSD|27577562

5/29/2016

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 50

 lOMoARcPSD|27577562

5/29/2016

For the detail of tangent modulus

- go through the book Mechanics of Materials - James M. Gere

What is the difference between Buckling and Bending?

Buckling and bending are similar in that they both involve bending
moments. In bending these moments are substantially independent of
the resulting deflections, whereas in buckling the moments and
deflections are mutually inter-dependent - so moments, deflections
and stresses are not proportional to loads.

Downloaded by IMRAN HOSSAIN (hossainimran.buet@gmail.com) 51