Structural Mechanics (CE- 312)

Buckling and Stability of Columns

Dr. Nauman KHURRAM

Assistant Professor
Department of Civil Engineering

UNIVERSITY OF ENGINEERING
& TECHNOLOGY LAHORE
 DESIGN CONSIDERATION OF STRUCTURE

1: Strength: The ability of structure to support a specified

load without experiencing excessive load.
2: Deformation: The ability of structure to support a
specified load without undergoing appreciable deformation.
3: Stability: The ability of structure or structural member to
support a given load without experiencing a sudden change in
its configuration (Buckling).
We define instability instead of stability
 Change in geometry of a structure or structural component
under compression , resulting in loss of ability to resist
loading is defined as instability.
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 2
Structure is in unstable equilibrium when small perturbations
(disturbance) produce large movements and the structure
never returns to its original equilibrium position.
Structure is in neutral equilibrium when we cant decide
whether it is in stable or unstable equilibrium. Small
perturbation cause large movements but the structure can be
brought back to its original equilibrium position with no
work.
Thus, stability talks about the equilibrium state of the
structure.

Stable Equilibrium Unstable Equilibrium Neutral Equilibrium

By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 3

The definition of stability had nothing to do with a change in
the geometry of the structure under compression.
Change in geometry of structure under compression that
results in its ability to resist loads called instability.
Not true :this is called buckling.
Buckling is a phenomenon that can occur for structures under
compressive loads.
Stability of equilibrium:
As the loads acting on the structure are increased, when does
the equilibrium state become unstable?
The equilibrium state becomes unstable due to:
 Large deformations of the structure
 Inelasticity of the structural materials

By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 4

COLUMN
A column is a line element (long slender bar) subjected to
axial compression. The term is frequently used to describe a
vertical member.
Structural members (i.e., columns) are generally stable when
subjected to tensile loading and fail when the stress in the
cross section exceeds the ultimate strength of material.
In case of elements (i.e., column) subjected to the
compressive loading, secondary bending effect e.g.,
imperfections within material and/or fabrication process,
inaccurate positioning of loads or asymmetry of cross
section can induce premature failure either in part of cross
section or of the whole element. In such case failure mode is
normally the Buckling.
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 5
Buckling is categorized into the following
1. Overall buckling
2. Local buckling
3. Lateral Torsional buckling
The design of the most of the compressive members is
governed by over-all buckling capacity. i.e., the maximum
compressive load which can be carried before the failure
occurs due to the excessive deflection in the plane of greatest
slenderness ratio.
Typical overall buckling occur in columns of frame structure
and in compression members of trusses

By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 6

SLENDERNESS RATIO (Le /rmin)
It is the ratio of the effective length of column (Le) to the
minimum radius of gyration (rmin) of cross sectional area.
If the columns is free to rotate at each end then buckling
takes place about that axes for which the radius of gyration is
minimum.
TYPES OF THE COLUMNS
The compression elements (Columns) are sub-divided into the
following three categories.
1. Short Column
The column which has a relatively low slenderness ratio is
called the short column (e.g., length of not greater than the 10
time to the least cross sectional dimension).
Failure occur when stress over the cross section reaches the
yield or crushing value of the material.
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 7
Such element fail by crushing of material induced by
predominantly axial compressive stress (flexure stresses are
not dominant).
2. Slender Column
The column which has a relatively high slenderness ratio is
called the slender or long column (e.g., length is greater than
the 30 time to the least cross sectional dimension).
Such element fail due to excessive lateral deflection (i.e.,
buckling) at a value of stress considerably less than the yield
or crushing value.
In slender column flexure stress are dominant and
compressive stress are not too important.
3. Intermediate Column
The failure of columns is neither short nor slender and occur
due the combination buckling and yielding/crushing.
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 8
For Intermediate column Length is in between 10 to 30 time
to the least cross sectional dimension.
Ideal Column
An ideal column has the following properties.
1. Its is prismatic (having the constant cross section through
out the length).
2. Material is homogeneous.
3. Loading is perfectly axial.
4. Pin ended condition (simply supported) are frictionless.
Real Column
1. Imperfection are present (i.e., structural and geometric)
2. Its not perfectly prismatic
3. Centroid may not lie on line joining the centroid of the end
section.
4. Load is not acting along the centroidal line.
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 9
 e t ep
Stress in Eccentric Column P

Actual
center line

Theoretical
center line

e = Total eccentricity e  et  e p
et = Theoretical eccentricity
ep = Loading eccentricity P P.e
Z = section Modulus  
A Z
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 10
CRITICAL LOAD OF COLUMNS
The critical load of as slender bar (columns) subjected to axial
compression is that value of the axial load that is just sufficient
to keep the bar a slightly deflected configuration.
Case-I: P < Pcr P < Pcr P = Pcr P > Pcr
Stable Equilibrium and No Buckling
Case-II: P = Pcr
Equilibrium State and Slight deflection
Case-III: P > Pcr
Unstable State and Buckling d

P < Pcr P = Pcr P > Pcr

By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 11
EULER FORMULA FOR PIN ENDED COLUMN

In 1759 a Swiss mathematician Leonhard Euler developed a

theoretical analysis of premature failure due to buckling.
Let suppose a pin ended
column AB of length L is
subjected to a slight
bending. Since column can
be considered a beam
placed in vertical direction
and subjected to axial load,
thus deformation at any
point of column can be
represented by equation of
elastic curve.
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 12
 d2y Here in figure, bending moment at
EI 2  M (1) point Q having co-ordinate (x , y)
dx can be represent as given in Eqn.
M   P.y (2) (2). The negative sign indicate the
negative bending moment.

d2y P
(1)  EI 2   P. y Let k 2
( 4)
dx EI
d 2 y P. y d2y
2
 0 (3) ( 4)  2
 k 2
y  0 (5)
dx EI dx

Eqn. (5) represent a second order Homogeneous Differential

Equation for simple harmonic motion and general solution
of the equation is given as Eqn. (6)
y  C sin kx  D cos kx (6)
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 13
 Coefficient C & D can be determined by applying the
boundary condition.
At End A: x=0&y=0 At End B: x=L&y=0
(6)  0  C sin( k 0)  D cos(k 0) (6)  0  C sin kL  0 cos kL
D0 0  C sin kL (7 )
In Eqn. (6) either A = 0 or sinkL=0. if A = 0 it will be zero
everywhere along the column and we will have a trivial
solution (member will be straight for any loading) the only

sin kL  0 (8) ( 4)  k
P
EI
To satisfy the Eqn. (8) P
(9)  .L  n 
kL  n (radian) (9) EI
n 2 2 EI
n = 1, 2, 3, …. P (10)
L2
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 14
n values of 1, 2, 3, represent the buckling
shape (eigenvalue) corresponding to 1st,
2nd and 3rd buckling mode shape,
respectively.
The smallest (critical) value load, Pcr
occurs when n = 1, which corresponding
to first (least) buckling mode.

 2 EI
Pcr  2
(11)
L

The Eqn. (11) is called the Euler formula and deflection

corresponding to this load is
Pcr
( 6)  y  C sin kx  C sin x (12)
EI
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 15
Substituting the value of Pcr from Eqn. (11)
 2 EI  .x
(12)  y  A sin 2
x  A sin (13)
EIL L
Eqn. (13) represents the equation of elastic curve after the
column has been buckled. From the equation (13) deflection
will be maximum when
x
If sin 1 (13)  ym  A
L
Above solution is indeterminate this is due to the fact that
differential Eqn. (2) used is the linearized approximation of
actual differential equation.
If P < Pcr the condition sin( 𝜋𝑥/𝐿) = 0 cannot be satisfied then
we must have C = 0 as only in this case configuration of
column will be straight, which is stable condition.
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 16
INFLUENCE OF END CONDITION
Effective Length (Le)
It is the length of the column corresponding to the half sigh
wave or length between the point of contra-flexure.
The Euler critical load for fundamental buckling mode
depends upon the effective length.
Effective Length Factor (K)
It is the ratio between the effective length and original length
Le
K
L
 Le  KL

The Factor K depends upon the end/boundary Condition of

the column
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 17
Effect of K-factor on Critical Buckling Load

By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 18

Critical Stress (σcr )
It is the stress corresponding to the Euler Critical Load and can
be calculated as following.
Pcr  2 EI  2E
 cr   2  2
(14)  I  Ar 2
A Le A ( Le / r )

Critical Stress (σcr ) for Slender Column

The critical stress for slender columns may be fixed by
dividing proportional or yield stress by factor of safety and
corresponding limiting slenderness ratio can be determined by
using the Eqn. (14).
 pl
Let  cr   2 200 103
F .O.S . (14)  200 
( Le / r ) 2
250
 cr   200 MPa  Le / r  100
1.25
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 19
Alternatively
For slender columns, Length > 30(least X-sectional dimension)
Assuming a rectangular cross-section of bxh.
hb 3 / 12 Le
rmin 
I min
 
b Let  100
A bh rmin
2 3
Le 30b  2 200 103
  103  cr  2
 200 MPa
rmin b / 2 3 (100)

Critical Stress (σcr ) for Short Column

For Short columns critical stress is taken equal to the crushing or
yield stress and slenderness ratio may be fixed by considering
the, Length = 10 (least X-sectional dimension)
Le 10b Le
  34.6 Let  30
rmin b / 2 3 rmin

By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 20

EXAMPLE PROBLEM
A steel bar of rectangular cross section of 40x60 mm2 and
pinned at the both ends is subjected to axial compression. If
proportional limit of material is 230 MPa and E =200 GPa.
a) Determine the maximum length for which the Euler
Equation may be used to calculate the buckling load.
b) For the same column determine the Euler Buckling load if
length of the column is equal to 2m.
Data
Area = bxh = 40x60 mm2
σPl = 230 MPa , E = 200 GPa , L=?
b) Pcr = ? If L = 2 m

By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 21

Example 10.01 (Bear & Johnston 6th Ed.)
A 2.0 m long pin-ended column of square cross section is to be
made of wood. Assuming E =13 GPa, σall = 12 MPa, 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) A 100 kN load
b) A 200 kN load
Data
σall = 12 MPa , E = 13 GPa
F.O.S. = 2.5
L = 2.0 m
Size of square column, b = ?

By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 22

σcr FOR INTERMEDIATE COLUMNS
Tangent Modulus Theorem (Inelastic Buckling)
By this method a modified version of Euler equation is adopted
to determine the stress-slenderness relationship in which the
value of the modulus of elasticity at any given level.
Consider a column manufactured from the a material, whose
stress-strain curve is shown in the figure below.
The slope of the tangent to the
stress-strain curve at any stress
value σ (σ is greater than σPl and is
within the inelastic range) is equal
to the value of Tangent Modulus of
Elasticity, Et.
Et is different from the E which is
the value at Elastic limit.
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 23
The value of Et can be used is Euler equation to calculate
the modified slenderness corresponding to any successive
value of σ.
The curve for to intermediate column can be plotted by
obtaining the slenderness value corresponding the any
successive stress value (σ = σcr) ranging between than σPl
and σult or crushing value .
 2 Et Le  2 Et
 cr    (15)
( Le / r ) 2
r  cr
Although, the nonlinearity of the stress-strain diagram
beyond the proportional limit is considered in Eqn. (15), its
theoretical basis is somewhat weak. Therefore, this equation
should be viewed as an empirical formula. However, the
results obtained from Equation are in satisfactory agreement
with experimental results.
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 24
 Compression Yielding Inelastic Buckling  2 Et
(Tangent Modulus Theorem)
 cr 
( Le / r ) 2
σy
Insignificant zone of
Euler Equation

Euler’s Curve  2E
σcr (Elastic Buckling)
 cr 
( Le / r ) 2

30 100 KL / r
Short Intermediate Long Columns
Columns Columns (KL / r)max
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 25
Rankin-Gordon Formula
Euler formula is only suitable for the slender columns with
small imperfections. In practice, most of the intermediate
columns fail due to the combined effect of compression and
flexure and experimentally obtained results are much less than
the Euler prediction.
Gordon suggested an empirical formula based on the
experimental results to predict the load of intermediate
columns, which was further modified by Rankin.
According to Rankin intermediate columns/members fail due
to buckling and compression to more or less degree and load
carrying capacity of such member can be calculated as
following.
1 1 1 Pc Pe Pc
   PR   (16)
PR Pc Pe Pc  Pe 1  Pc / Pe

By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 26

In Eqn. (16) For pin endded column
PR = Rankin – Gordon buckling load  2 EI
Pe 
Pe = Euler buckling Load Le
2

Pc = Ultimate compressive load Pc   c A or  y A

 y .A  y .A  y .A
(16)  PR   
 y .A  y . ALe 2
 y  Le  2
1 2 1 2 1  2 . 
 EI Le 2
 E ( Ar ) 2
 E r 
 y .A
PR  2
(17)
 Le 
1  a 
 r  a = Rankin constant, which
y depends upon the boundary
 a 2
 E condition and material properties
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 27
Graphical Presentation of Rankin Formula

Rankin constant for various Materials

Material σy (MPa) Rankin constant, a
Mild Steel 325 1 / 7500
Wrought Iron 250 1 / 9000
Cast Iron 560 1 / 1600
Timber 35 1 / 3000
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 28
Example Problem
A cast Iron column of 200 mm external diameter is 20 mm
thick and 4.5 m long. Assuming the both end rigidly fixed,
calculate the safe load using Rankin Formula if Rankin
constant, a = 1 / 1600, σy = 550 MPa F.O.S. = 4.0.

Data
σy = 550 MPa , F.O.S. = 4.0
Do = 200.0 mm K = 0.5 ( both Ends fixed)
t = 20 mm a= 1 / 1600
Psafe = PR / FOS
 y .A
PR  2
 Le 
1  a 
 r 
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 29
Practice Problem
For the given data determine the length of the Column for
which Euler formula cease to apply.

Data
σy = 325 MPa , E = 200 GPa
Do = 38.0 mm L = 2.25 m
Di = 33.0 mm K = 1.0 ( both Ends pinned)
a= 1 / 7500

By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 30

AISC SPECIFICATIONS FOR STEEL COLUMNS
American Institute of Steel Construction (AISC) specifies two
method for the computation of the compressive strength of the
columns. Both design specification bound the maximum
slenderness ratio equal to 200.
1. Allowable stress design (ASD)
2. Load and Resistance Factor Design (LRFD)
1.0 Allowable stress design (ASD)
It is the old method and according to this method columns
made of structural steel can be designed on the basis of
formulas proposed by the Structural Stability Research
Council (SSRC). Factors of safety are applied to these
formulas.
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 31
It consider only intermediate (short) and long column and
there is no straight portion between the stress~slenderness
ratio curve. A specific slenderness ratio value Rc is used to
differentiate between the slender and intermediate (or short)
column.
Experimental studies showed that compressive residual
stresses can exist in rolled-formed steel sections their
magnitude may be as much as one-half the yield stress.
Consequently, if the stress in the Euler formula is greater σy
/2 then equation is not valid. Thus, limiting slenderness
ratio Rc for the long columns can be determined by putting
the σcr = σy /2 in Euler Equation.
y  2E Le  2E
  Rc   (18)
2 ( Le / r ) 2
r y

By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 32

Slender Column
Le
If 200   Rc Its long column
r
In long column allowable stress can be calculated through
the Euler equation divided by the Factor of safety.

 cr   2E  1 12  2 E
 all    
2 
 . (19)
FOS  ( Le / r )  FOS 23 ( Le / r ) 2

Short Column
Le
If  Rc Its short / Intermedioate column
r
The short column are designed on the base of an empirical
formula which is parabolic in form and maximum stress by
this formula is given as following.
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 33
  Le / r 2 
 max  1  
2  y
(20)
 2 Rc 

5 3  Le / r  1 Le / r 
3
FOS      (21)
3 8  Rc  8 Rc 3

 max
 all  (22)
FOS
FOS becomes 5/3 or 1.67 when
Le /r = 0 and increases to 1.92 or
23/12 at slenderness value equal
to Rc.
All the above equation may be
used both in SI and FPS
System.
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 34
Example Problem
Determine the allowable axial load (Pallow) for a W310x129
wide-flange steel column with both end pinned, for the
following lengths.
(a) L = 6 m (b) L = 9 m
Assume E = 200 GPa and σy = 340 MPa
Data
For W310x129 Section
A= 16,500 mm2 σy = 340 MPa
rz = ry = 78.0 mm E = 200 GPa
K = 1.0 ( both Ends pinned)

By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 35

ECCENTRICALLY LOADED COLUMN
(SECANT FORMULA)
In practice it is difficult to apply the end thrust (axial load)
along the longitudinal centroidal axes of columns. In such case
we have to consider the effect of eccentrically applied load “P”
on a prismatic column of flexural stiffness EI.

M Q  M A  P. y
= M Q  P (e  y )

By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 36

 2 Suppose axial load is acting at an
d y
EI 2  M (1) eccentricity “e” from the weaker
dx axes (y-axis) the equation of elastic
M Q   P (e  y ) (23) curve and moment at any arbitrary
point Q can be given in Eqn. (23).

d2y P
(1)  EI 2   P(e  y ) Let k2  ( 4)
dx EI
d 2 y P. y P.e d2y
2
   ( 24) ( 24)  2
 k 2
y   k 2
e (25)
dx EI EI dx

The complete solution of Eqn. (25) is given as following

y  C sin kx  D cos kx  e (26)

General solution Particular solution
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 37
 Coefficient C & D can be determined by applying the
boundary condition.
At End A: x=0&y=0
(26)  0  C sin(k 0)  D cos(k 0)  e  De
At End B: x=L&y=0
(26)  0  C sin kL  e cos kL  e kL
 sin kL  2 sin cos
kL
C sin kL  e(1  cos kL) 2 2
2 kL
 kL kL   2 kL   (1  cos kL)  2 sin
C  2 sin cos   e 2 sin  2
 2 2   2 
kL
C  e tan Substituting the value C & D in Eqn. (26)
2
 kL 
y  e  tan sin kx  cos kx  1 (27)
 2 
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 38
 The Eqn. (27) represents the equation of deflection (y) at any
point (x) along the columns. The value of maximum deflection
(ymax) can be calculated by setting x = L /2.

 kL kL kL 
(27)  ymax  e  tan sin  cos  1
 2 2 2 
 sin 2 kL / 2  cos 2 kL / 2 
ymax  e   1 The Eqn. (29) shows that (ymax)
 cos kL / 2  becomes infinite when P = P .
cr
 kL 
ymax  e sec  1 (28)    2 EI L  
 2  ymax  e sec   1
  EIL 2  
2

P
(4)  k    
EI ymax  e sec  1
 2 
  P L 
ymax  e sec   1
 (29) 
  EI 2    sec  
2
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 39
In actual cases deflection does not become infinite even the
load exceed the elastic limits also P should not be reached to
the Pcr (Euler critical load)
 2 EI Pcr L2
Pcr   EI 
L2
2
Replacing the value of EI in
Eqn. (29)
  P 2 L  
ymax  e sec   1
  Pcr L 2  
2

Source of eccentricity in
  P  column
ymax  e sec  1 (30)
 2 Pcr 

Note: In above equation secant angle is in radians

By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 40
MAXIMUM STRESS IN ECCENTRIC COLUMN

The maximum stress σmax occurs in the section

of the column where the bending moment is
maximum, i.e., in the transverse section
through the midpoint C, and can be obtained
by adding the normal stresses due to the axial
force and the bending couple exerted on that
section
 M C  P( ymax  e)
P M max  c
 max   (31)  kL 
A I  ymax  e sec  1
P P( ymax  e)  c  2 
 max     P 
A Ar 2 OR ymax  e sec  1
 2 Pcr 
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 41
 P P( ymax  e)  c
 max  
A Ar 2 e = eccentricity of loading
P   kL   c  c = distance from the N.A.
 max  1  e sec  1  e 2  to extreme fibres
A   2  r 
r = radius of gyration
P   kL  ec 
 max  1  sec  1  1 2  A = cross-sectional area of
A  2 r  column
P  ec kL 
 max  1  sec (32)
A  r 2 2 

kL P L  P
Replacing the value of kL /2 as following  
2 EI 2 2 Pcr

P  ec  P 
(32)   max  1  2 sec  (33)
A r 2 Pcr 

By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 42

The Eqn. (33) can be used for any end condition as long as
the appropriate (K) value is used to calculate Pcr.
Since σmax does not vary linearly with load P, the principal
of superposition is not applicable to determine the stress due
to the simultaneously application of applied loads.
For the same reason any factor of safety should be used with
load not the stress.
P  ec P L
(32)   max  1  2 sec 
A r EI 2   I  Ar 2
P  ec P Le 
 max  1  2 sec  ec
A r EAr 2 2  2
 Eccentricity ratio
r
P  max
 (34)
A  ec  1 P Le  P Le
1  2 sec   Euler angle
 2
EAr 2
 r  2 EA r 
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 43
The formula given in Eqn. (33) is referred to as the secant
formula; it defines the force per unit area (P/A), that causes a
specified maximum stress (σmax) in a column of given
effective slenderness ratio (Le /r), for a given value of the
eccentricity ratio (ec/r2).
If the material properties, the dimensions of the column, and
the eccentricity e are known then we have two variables in
the secant formula: P and σmax. If P is also given, σmax can be
computed from the formula without difficulty.
 On the other hand, if σmax is specified, the determination of
P is considerably more complicated because Eqn. (33),
being nonlinear in P, must be solved by trial-and-error.
 The secant formula is chiefly useful for intermediate
values of Le /r. However, to use it effectively, we should
know the value of the eccentricity e of the loading
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 44
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 45
 Due to imperfections in manufacturing or specific
application of the load, a column will never suddenly
buckle; instead, it begins to bend.
 The load applied to a column is related to its deflection
in a nonlinear manner, and so the principle of
superposition does not apply.
 As the slenderness ratio increases, eccentrically loaded
columns tend to fail at or near the Euler buckling load.

By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 46

Exercise: Plot the load-displacement curves of a rectangular column
for the given data with eccentricity ranging from 5-25 mm.
Data
L= 2.5 m K = 1.0, A = 30x60 mm2, e = 5 – 25 mm

Solution Imin= 60x303/12= 135,000 mm4 , rmin = 8.66 mm

 2  200 103 135 103  kL    P 

Pcr  42.64kN ymax  e sec  1  e sec  1
(1 2500) 2  2   2 Pcr 

P (kN) Sec(kL/2) y (mm)

0 e= 5 e= 10 e= 15 e= 20 e= 25
10
20
30
40
42.64

By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 47

Problem 10.34 (Mech. of Materials, Bear & Johnston, 6th Ed)
The axial load P is applied at a point located on the x-axis at a
distance e from the geometric axis of the rolled-steel column BC.
When P = 350 kN, the horizontal deflection of the top of the
column is 5 mm. Using E =200 GPa. determine
(a) the eccentricity e of the load,
(b) the maximum stress in the column.
Data
For W250x58 Section
A= 7420 mm2 Sy = 185x103 mm3
Ix = 8700x104 mm4, rx = 108 mm
Iy = 1870x104 mm4, ry = 50.3 mm
E = 200 GPa, d = 250 mm, bf =203 mm
K = 2.0 (cantilever case)
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 48
Sample Problem 10.3
(Mech. of Materials, A Pytel, 2nd Ed)
A W14x61 section is used as a simply supported column of 25 ft
long. When the 150-kip load is applied with the 4-in. eccentricity
shown, Determine
(1) the maximum compressive stress in the column;
(2) the factor of safety against yielding; and
(3) The maximum lateral deflection.
Assume that the column does not buckle about the y-axis.
Use E = 29x106 psi and σy = 36x103 psi..

For W250x58 Section

A= 17.9 in2 Iz = 640 in4 ,
Sz = 92.1 mm e = 4 in ,
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 49
Problem 13.53
(Mech. of Materials by RC Hibbler, 8th Ed)
The W200x22, A-36-steel column is fixed at
its base. Its top is constrained to rotate about
the y–y axis and free to move along the y–y
axis. Also, the column is braced along the x–x
axis at its mid-height. Determine the
allowable eccentric force P that can be
applied without causing the column either to
buckle or yield. Use against buckling F.O.S. =
2.0 and F.O.S. = 1.5 against yielding.
For W250x58 Section
A= 28600 mm2 Ix = 20x106 mm4, rx = 83.6 mm
e = 100 mm , Iy = 1.42x106 mm4, ry = 22.3 mm
d = 200 mm, bf =102 mm , E = 200 GPa
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 50
INITIALLY CURVED COLUMN
(PERRY - ROBERTSON FORMULA)

In practice a column cannot be made perfectly straight and

Pcr is never reached. Consideration of small deviation from
the straight configuration makes the analysis more realistic.
According to Perry-Robertson Formula, all practical
imperfections (e.g. properties of the real columns) could be
represented by a hypothetical initial curvature (a0) of
column.
Let consider a columns AB of length L has an initial
imperfection y0 prior to the application of the and y is the
additional deformation due to the applied load P. the equation
of the elastic curve for any arbitrary point Q can be represented
as following.
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 51
 d2y P
EI 2  M (1)
dx A
x=0,y=0 yo
M Q   P ( y0  y ) (35)
y
d2y ym
(1)  EI 2   P( y0  y ) Q
dx
L P
d 2 y P. y P. y0
2
  (36)
dx EI EI
ao MQ= - P(y0+y)
P
Let k  2
( 4)
EI
x=L,y=0
d2y B
(36)  2
 k 2
y   k 2
y0 (37) P
dx

y0 = initial deviation of the column and x

y0  a0 sin (38)
is represented by the sinusoidal curve L
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 52
 d2y x
(37)   k 2
y   k 2
a 0 sin (38)
dx 2 L
The complete solution of Eqn. (38) is given as following
k 2 a0 x
y  C sin kx  D cos kx  sin (39)
 2
 L
 2  k 2 
L 
Applying the boundary condition
At End A: x=0&y=0 At End B: x=L&y=0
k 2 a0 L
(39)  0  C sin( k 0)  D cos(k 0)  0 0  C sin kL  0 cos kL  2 sin
  L
 D0  2  k 2 
L 
In Eqn. (40) either C or sinkL is zero 0  C sin kL (40)
Assuming k any non-zero value (as deflection will always be
due to some applied load P) we must have C=0
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 53
Substituting the values of C and D in Eqn. (39)
k 2 a0 x a0 x
(39)  y 2 sin  sin P
  L   2
 L  k2 
 2  k 2   2 2  1 EI
L  k L 
x x  2 EI
y 2
a0
sin 
a0
sin (41)  Pcr 
  EI  L  cr  P L L2
 2  1   1
 LP   P 

For pin ended column the deflection is maximum (ym) at

center when x = L/2
a0  ( L / 2) a0
(41)  ym  sin  (42)
 Pcr  L  Pcr 
  1   1
 P   P 

In Eqn. (41) & (42) y and ym are the additional deflection due
to the applied P as compared to the initial deflection a0.
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 54
Using Eqn. (42)
if P  0.2 Pcr  ym  0.25a0

if P  0.5 Pcr  ym  a0
if P  0.75 Pcr  ym  3a0
if P  0.9 Pcr  ym  9a0

if P  Pcr  ym   Load-deflection Curve of initially

curved column
The relationship of P and ym as shown in the figure depicts
that the initially deformed columns fails before reaching the
Pcr (Euler critical load) and ym increases rapidly with the
increase of load P.
At any definite displacement before the failure the Eqn. (42)
be written as following.
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 55
  Pcr 
(42)  ym   1  a0
 P 
P P y a
ym cr  ym  a0  cr  m  0
P P Pcr Pcr
Relating with the equation of
straight Line
y  mx  c
 Pcr   1   a0 
    m  
   y  (43)
 P   Pcr   Pcr  South-Well Plot

The values of ym /P and ym are plotted from a column test

then these variables can be related by a straight line.
While plotting initial values may be discarded (40% to 80%
data may be plotted).
This plot is called the South-well plot and it is used to
determine the initial deflection of a column, experimentally.
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 56
Total deflection at any distance x is given as
 
x a0 x x  1 
yt  y  y0  a0 sin  sin  a0 sin   1
L  Pcr  L L  Pcr  1 
  1  
 P   P 
x  P  x  P 
yt  a0 sin   1  a0 sin  cr 
L  Pcr  P  L  Pcr  P 
x  Pcr  A x   cr 
yt  a0 sin    a0 sin   (44)
L  Pcr  P  A L   cr   

Displacement will be maximum at x = L /2

 ( L / 2)   cr    cr 
 yt max  a0 sin    a0   (45)
L   cr      cr   

By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 57

 MAXIMUM STRESS IN DEFLECTED COLUMN

The maximum stress σmax occurs in the section of the column

where the bending moment or displacement is maximum.

P M max  c    cr 
 max   (46)  max   1     (47)
A I    cr   
P P( yt ) max  c
 max   
a0 c
  2  Initial deflection ratio
A Ar 2 r
  cr     averge applied stress
Pa0    c
P   cr      cr  Euler critical stress
 max  
A Ar 2
If applied load P is given the
P  a0 c   cr 
 max  1  2   maximum stress can be
A r   cr   
determined by using the Eqn. (47)
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 58
If σmax are specified then to determine the safe applied load
the Eqn. (47) is to transformed in term of applied stress σ.

   cr        cr 
(47)   max   1        cr 
   cr      cr   
 max . cr      cr   2   cr
 2    max   cr   cr    max cr   2    max  (1   cr )    max cr  0


1
 max  (1   ) cr   1  max  (1   ) cr 2   max cr (48)
2 4

 We need not to consider positive square root since we are only

interested in smaller values of square roots in the Eqn. (48).
 This equation represents the average value of stress in the cross-
section at which the maximum stress would be attained at mid-
height of the column for any given value of η.
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 59
To determine the average applied stress (σ) at which yield
occurs then σmax is replaces by the σy.


1
2
 
 y  (1   ) cr 
1
4
 
 y  (1   ) cr 2   y cr (48)

Experimental evidence obtained by Perry and Robertson

indicated that for a mild steel the hypothetical initial curvature
of the column could be represented as following.
L
  0.003 (49)     cr   y
r

It is that value of slenderness ratio when the yield stress is

first attained in one of the extreme fibres.
2
1 L  1 L 
   y  (1  0.003 ) cr    y  (1  0.003 ) cr    y cr (50)
2 r  4 r 
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 60
By Dr. Nauman KHURRAM Department of Civil Engineering, UET Lahore 61