Steel Structures

SE-505

Design of Locally Unstable

Compression Members
Instructor: Dr. Nauman Khurram
LOCALLY UNSTABLE
MEMBERS IN COMPRESSION
INTRODUCTION
The design of locally unstable sections, at least with
respect to overall buckling (local instability does not
occur before the chances of overall buckling), is
discussed earlier.
Sometimes, thin / slender elements are used in the
compression members, which may carry substantial
loads even after this local instability.
➢ In fact, the thin plate cold-formed sections are
always made such that the individual elements are
not locally stable.
ELASTIC STABILITY OF PLATES
➢ The buckling of a plate section, having a size
of b × t, depends on the equivalent
slenderness ratio.
➢ This equivalent slenderness ratio is equal to
the width / thickness ratio denoted by λ, equal
to b / t.
DIFFERENTIAL EQUATION FOR BENDING OF
HOMOGENEOUS PLATES

The general forces acting a 3-d element are shown in

Figure 11.1. Following nomenclature is used for
various force effects:
Mx = Bending moment per unit length on x-face.
My = Bending moment per unit length on y-face.
Mxy = Twisting moment per unit length on x-face.
Myx = Twisting moment per unit length on y-face.
Qx = Shearing force per unit length in z-direction
acting on x-face.
Qy = Shearing force per unit length in z-direction
acting on y-face.
q = Intensity of continuously distributed load in
z-direction.
Nx = Normal force per unit length on x-face.
Ny = Normal force per unit length on y-face.
w = Deformation in z-direction load “q”.
 w
= Slop in x − direction
x
w
= Slop in y − direction
y
2w
= Curvature in x − direction , proportional to moment M x
x 2

2w
= Curvature in y − direction , proportional to moment M y
y 2

 2 w Change of x − direction slope measured in y − direction or vice versa showing

=
xy torsional shear curvature proportional to torsional moment M xy and M yx
 3 w First derivative of x − direction curvature w.r.t. x − axis (indicating rate
=
x 3
of change moment in x − direction) proportional to shear force Q x
 3 w First derivative of y − direction curvature w.r.t. y − axis (indicating rate
=
y 3
of change moment in y − direction) proportional to shear force Q y
 4 w Second derivative of x − direction curvature w.r.t. x − axis (indicating rate of
=
x 4
change shear force in x − direction) proportional to load change in x − axis
 4 w Second derivative of y − direction curvature w.r.t. y − axis (indicating rate of
=
y 4
change shear force in y − direction) proportional to load change in y − axis
4w Second derivative of y − direction curvature w.r.t. x − axis (indicating rate of
=
x 2y 2 change shear force in y − direction)proportional to load change in y − axis
 E.t 3
D =
(
12 1 −  2 )
D = Flexural rigidity of the plate
t = Thickmess of plate

The differential equation for bending of a plate

element may be written by adding the load resistance
by flexure and shear in the two directions (the related
derivatives along with the constant of proportionality
equal to the flexural rigidity of the plate, D) and
equating it to the applied load.
The D-term may be taken on the right hand of
the equation.
4w 4w 4w q
+2 2 2 + 4 =
x 4
x y y D

After solving this equation for the deflection function

w(x), analytically for some simple cases or numerically,
the corresponding load effects may be calculated by
using the following expressions:
 2w 2w 
Mx = − D. 2 +  2 
 x y 
 2w 2w 
My = − D. 2 +  2 
 y x 
 2w
M xy = M xx = − D(1 −  )
xy
M xy M x   2w 2w 
Qx = + = − D  2 + 2 
y x x  x y 
M y M xy   2w 2w 
Qy = + = − D  2 + 2 
y x y  x y 

BUCKLING OF UNIFORMLY COMPRESSED PLATE

The buckling of a uniformly compressed plate may be

studied by considering a thin plate, just at the stage of
buckling, with free to rotate edges and subjected to
compressive force uniformly distributed at the edges.
 b Nx = F crt

Nx Transverse Component of Nx

X
Lateral Deflection, w

∂w
∂x
∂N
dx x
Nx + dx
Y ∂x
Considering a thin plate element of size a x b,
subjected to a critical buckling stress on the edges
denoted by Fcr, the applied axial force per unit length
on the edges will become Fcr t = Nx in our general
nomenclature for the plate element.
Now considering a differential element of size dx x dy,
a component of the force Nx after buckling acts as the
transverse load q on the element.
The magnitude of this load may be estimated by
considering the equilibrium of the element in the z-
direction after buckling as follows:

w  N x   w  2 w 
N x dy −  Nx + dx .dy. + 2 dx  = qdxdy
x  x   x x 
➢ Where Nxdy is the total inclined force, ∂w /∂x is the
slope or tangent of the angle and Nx dy.∂w /∂x is
the z-direction component of this load.
➢ Remember that for small angles in radians, the
angle itself, its sine and tangent are almost equal.
➢ The second term in the equation is the same
expression developed for the right end of the
element.
➢ After opening the brackets, the following is
obtained:
w w 2w N x w N x  2 w 2
N x dy − Nx dy − N x 2 dxdy − dxdy − dx dy = qdxdy
x x x x x x x 2

  2 w N x w N x  2 w 
−  N x 2 + + dx 2 dxdy = qdxdy
 x x x x x 
Neglecting the product of infinitesimal terms, the
expression simplifies to:
2w
q = − Nx. 2
x
Using his load, the D.E. of plate bending may provide
all the required results. The form of this equation will
become as under:
4w 4w 4w N x 2w
+2 2 2 + 4 = −
x 4
x y y D x 2

This is a partial differential equation involving two

variables. For its solution, a deflection function to be
fitted may be assumed to be a product of a x-function
F1(x) and a y-function F2 (y).
 w = F1 ( x) F2 ( y )
Further assuming that the buckling will yield a
sinusoidal variation along the x-axis, following function
may be tried:
 mx 
F1 ( x) = sin  
 a 
where a is the length of the plate and m is an integer
number. This function satisfies the boundary conditions
as shown below:

X = 0  F1 ( x ) = 0  w = 0 (BC #1)
2w  mx  
m 2 2
= F2 ( y ) − sin  2
x 2
 a  a
 2w
x=0  =0 (BC #2)
x 2

(Moment at edge is equal to zero)

x = a  F1 ( x ) = m  w=0 (BC #3)

 w
2
x=0  =0 (As above) (BC #4)
x 2

Substituting the trial function into the governing

differential equation, the following result is obtained:

 m mx  m  mx d 2 mx d 4

2

  sin F2 ( y ) − 2  sin F2 ( y ) + sin F2 ( y )
 a   a 
2 4
a a dy a dy
N x  m  mx
2

=   F2 ( y )sin
D  a  a
 4
 m  d
2 2   m  N x  m  
4 2

F2 ( y ) − −2 F2 ( y ) +    F2 ( y ) = 0
d
  − 
 a  dy  a  D  a  
4 2
dy

This is an ordinary fourth order homogeneous

differential equation in terms of only one variable, that
is, y. The solution of this equation is:

F2 ( y ) = C1 sinh y + C2 sinh y + C3 sinh y + C4 sinh y

 m  N x  m 
2 2

Where  =   +  
 a  D  a 

 m  N x  m 
2 2

and  = −  +  
 a  D  a 
The compete solution for plate deflection
function is:

  m .x 
w = sin   (C1 sinh y + C2 cosh y + C3 sinh y + C4 cosh y )
  a 

➢ This function must satisfy all the boundary

conditions.
➢ However, sine and sinh functions are not
symmetrical about x = 0 line.
➢ For identical support conditions along the two edges
parallel to the direction of loading (y =+- b / 2), the
odd function coefficients must be zero.
C1 = C2 = 0
 m .x 
w = (C2 sinh y + C4 sinh y )sin  
 a 
w  m .x 
= (C2 cosh y + C4  cosh y )sin  
y  a 
2w  m .x 
y 2
= ( 2 2
)
C2 cosh y + C4  cosh y sin  
 a 
The boundary conditions are that the edges,
y = +- b / 2, are simply supported.

2w
x = 0 and =0
x 2

Which give the following results

b b
C2 cosh  + C4 cosh  = 0
2 2
b b
C2 cosh  − C4  cosh 
2 2
= 0
2 2
 b b 
 cosh  2 cosh 
2  C2  = 0
 b b    
 2 cosh  −  2 cosh   C4  0 
 2 2
For a non-trivial solution, C2 and C4 must be
non-zero and the determinant of the coefficients
matrix must be zero:
 b  2 b  2 b  b
−  cosh   cosh   −   cosh   cosh   = 0
 2  2  2  2

( 2 2
) b
 +  cosh  cosh 
2
b
2
= 0

The condition that α 2= −β represents a trivial solution

giving N = 0 and cosh(αb /2) can not be zero (it is
always greater than or equal to 1.0).
The only way in which the above equation may be
satisfied for a real problem is the following:
b b  3 5
cosh  = 0 or  = , , , etc.
2 2 2 2 2
The first mode of buckling along the width (quarter
wave in b / 2 distance) is usually the most critical, which
is represented by the least values out of the above
solutions. b 
 =
2 2

 m  N x  m  
2 2
b
−  +   =
2  a  D  a  2

 m  N x  m  2
2 2

−  +   =
 a  D  a  b2
2 2
N x  m 
2
 2
 m  
  =  2 +  
D  a   b  a  
2 2
Nx D   m  
2
=  2 +  
t t  b  a  
 D  1 a
2
2
b
or Fcr = + m
b 2t  m b a 

The term in the brackets is called the plate- buckling

coefficient (k ), m is the number of half sine waves in
the buckled shape along the x-axis (or along the length)
and D is the plate rigidity defined earlier.
2
1 a b Et 3
k= + m  and D =
 m b a  12 1 −( 2
)
The expression for Fcr becomes :
 2E
=
12(1 −  )(b / t )
Fcr 2 2
k

The buckling coefficient depends on the type of stress

(uniform compression or otherwise), edge conditions
(value of m will be different), and the aspect ratio a / b.
For a larger value of a / b and if m also becomes
larger, the k-curve becomes flatter and approaches a
constant value of 4.0. For example for a / b of 15, the
value of k becomes:
2 2
1 a b  15 m 
k =  +m  =  + 
m b a   m 15 
m = 1  k = 227
m=4  k = 16.1
m=8  k = 5 .8
m = 12  k = 4.02

It is to be noted that the less value of k gives less

buckling strength and is more critical. Hence for
simply supported ends, a value of 4.0 is taken.
For the other end conditions, the critical values
are listed below:
Kmin value for two edges simply supported = 4.0
Kmin value for two edges fixed = 6.97
Kmin value for one edge fixed other simply supported = 5.42
Kmin value for one edge fixed and other free = 1.277
Kmin value for one edge free and other simply supported = 0.425

The form of equation for Fcr may be modified as

under in order to study the factors affecting the
buckling strength:

Fcr  2 Ek 1
= =
Fy .12(1 −  2 )(b / t ) c 2
2
Fy
The square root of the ratio of yield strength to the
elastic critical buckling strength may be denoted by
the parameter λc , while the b / t ratio may be denoted
by the parameter λ.
The Poisson’s ratio for steel may be taken equal to 0.3
to simplify the above equation to the following:

Fy
c = 1.052 .
Ek

A graph between the parameter λ (b / t ratio) and

the ratio Fcr/Fy is shown in Figure. This graph
shows distinct phases of strength under the action
of edge compression as described below:
 0.724

0.476

λc=0.5 λc=1.0 λc=1.45

a) For very low values of λ, the strength becomes

higher than Fy due to strain hardening, without
any buckling.
➢ In such cases, post-buckling strength is absent
but the entire plate reaches strain hardening after
undergoing all the yielding.
➢ Hence, Fcr/ Fy may become greater than unity.
➢ The value λo is an equivalent elastic value of λc at
which chances of inelastic buckling just start
corresponding to the yielding stress.
➢ Some typical values of λo are as under:
λo = 0.455 for long hinged flanges.
λo = 0.461 for fixed flanges.
λo = 0.588 for hinged webs.
λo = 0.579 for fixed webs.
An average value of 0.5 may be considered for the
general discussion.
b) The value of λc is equal to one at the point
where no elastic buckling occurs up to Fy stress.
c) Inelastic buckling occurs for values of λc less
than approximately1.45.
d) The point B indicates a situation where the
elastic buckling formula gives strength equal to Fy.
e) Elastic buckling according to the derived formula
when the value of λc is greater than or equal to
1.45.
f) Post buckling strength with stress redistribution
and large deformations results after λc equal to
1.45.
Point A
Fcr/ Fy = 1.0 according to the inelastic buckling formula
c  0.5 for elastic formula
Ek
 for elastic c of 0.5  0.475
Fy
Point B
Fcr/ Fy = 1.0 on the elastic curve
c on the elastic curve = 1.0
Ek
  0.951 using the elastic formula
Fy

Corresponding value of Fcr / Fy calculated using the

inelastic straight line = 0.724
Point C
Fcr/ Fy = 0.476 according to the elastic buckling formula
c  1.45
Ek
  1.378 for elastic formula
Fy
Fy
Slop of straight line between A and C = 0.58
Ek
The values of the factor λ for the three points are listed
below for different critical values of k-factor:
For k = 0.425 For Overhanging Flanges

Ek E
  0.475 = 0.31
Fy Fy
 Ek E
  0.951 = 0.62
Fy Fy
Ek E
  1.378 = 0.898
Fy Fy

AISC Table B4.1:

Flexure in flanges of rolled I-shaped sections:

E
p  0.38 = 10.748 for A36 Steel
Fy

Uniform compression in flanges of rolled I-shaped

sections:
E
p  0.56 = 15.84 for A36 Steel
Fy
For k = 4.0 For Stiffened Web
Ek E
  0.475 = 0.95
Fy Fy
Ek E
  0.951 = 1.902
Fy Fy
Ek E
  1.378 = 2.756
Fy Fy

AISC Table B4.1:

Flexure in web of doubly symmetric I-shaped sections:
E
p  3.76 = 106.35 for A36 Steel
Fy

(Half of the web is in compression and that

compression also varies along the member.)
Uniform compression in webs of doubly symmetric
I-shaped sections:
E
r  1.49 = 42.14 for A36 Steel
Fy

➢ Any plate with no residual stresses develops uniform

axial stresses up to the stage when the portions away
from the side supports buckles in out-of-plane direction
(Figure).
➢ After buckling, the stresses become non-uniform and
post-buckling strength is available near the relatively
stable ends.
➢ The post-buckling strength becomes larger as the width-
to-thickness ratio is increased.
➢ The AISC values are in general excessively
conservative because of the presence of residual
stresses and imperfections
➢ For design, the local buckling of a column
component must be prevented if it occurs before
achieving full strength of the column based on its
overall slenderness ratio KL / r.
➢ This means that the acceptable b/t ratios vary depending
on the overall slenderness ratio of the column.

Fcr component  Fcr overall column

Design of Compression Members With Some Parts
Locally Unstable
AISC 2005 – E7
➢ Plate elements in compression, either “stiffened” or
“unstiffened” have post buckling strength.
➢ Stiffened elements have large post buckling strength while
unstiffened elements have little.
Post Buckling Strength
Stiffened Elements
Plate elements under axial compression, showing
actual stress distribution and an equivalent system

Nominal Strength of Stiffened Plate

t
Pn = t  f ( y )dy = t  be  f max
0

Pn = Aeff  f max

b = Effective width over which the maximum stress

may be considered uniform and still gives almost
correct answer.
Unstiffened Elements
Plate elements under axial compression, showing
actual stress distribution and an equivalent system

Nominal Strength of Stiffened Plate

Unstiffened elements have less post-

Pn = t  b  f avg buckling strength. They may be
Pn = Ag  f avg idealized as not buckled but subjected
to a reduced equivalent stress.
Nominal strength of the member having both
stiffened and unstiffened elements (W-section)

f avg Aeff
Pn = f avg  Aeff = . f max  . Ag
f max Ag
Pn = Qs .Qa  f max Ag = Q  f max Ag
Pn = Fcr Ag

A compression member consisting of both stiffened

and unstiffened elements may be treated as
unstiffened for establishing the stress favg, while
effective area is determined after deducting the
ineffective area out of the stiffened elements.
Q = Full reduction factor for slender compression elements,
1.0 for members with compact and non-compact
elements
Qs = Reduction factor for slender unstiffened elements, 1.0
if no slender unstiffened element is present
Qa = Reduction factor for slender stiffened elements , 1.0 if
no slender stiffened element is present
Critical/ Ultimate Compressive Strength, ɸcFcr
KL E
For  4.71 or Fe  0.44QFy
r QFy
  
 QFy 

 2E
= 0.658 e   QFy  =
F
Fcr Fe




(KL / r )2
 KL E
For  4.71 or Fe  0.44QFy
r QFy
Fcr = 0.877  Fe

Reduction Factor Qs for Unstiffened Element

For Columns
Fcr , plate Fcr , plate
Qs = 
Fcr ,column Fy

For Compression Flanges of Beams

Fcr , plate Fcr , plate

Qs = 
Fcr , beam flange Fy
a) For single angles and double angles with
separators
b E
For  0.45  Qs = 1.0
t Fy
E b E
For 0.45   0.91
Fy t Fy

 b  Fy
 Qs = 1.34 − 0.76  
t E
b
12.8   25.8 For A36 Steel
t
b E 0.53E
For  0.91  Qs = 2
t Fy  
b
Fy  
t
b) For flanges, angles and plates projecting
from hot rolled beams and columns
b E
For  0.56  Qs = 1.0
t Fy
E b E
For 0.56   1.03
Fy t Fy

 b  Fy
 Qs = 1.415 − 0.74  
t E
b
15.8   29.1 For A36 Steel
t
b E 0.69 E
For  1.03  Qs = 2
t Fy  
b
Fy  
t
c) For flanges, angles and plates projecting
from projecting from built-up columns or
other compression members
b Ekc
For  0.64  Qs = 1.0
t Fy
Ekc b Ekc
For 0.64   1.17
Fy t Fy

 b  Fy
 Qs = 1.415 − 0.65  
 t  Ekc
b E 0.90 Ekc
For  1.17  Qs = 2
t Fy b
Fy  
t
4
kc = 0.35  kc  0.76
h / tw
d) For Stem of Tees
d E
For  0.75  Qs = 1.0
t Fy
E d E
For 0.75   1.03
Fy t Fy

 d  Fy
 Qs = 1.908 − 1.22  
t E
d E
For  1.03
t Fy
0.69 E
 Qs = 2
d 
Fy  
t
d = the full nominal depth of tee
Reduction Factor Qa for stiffened Element
Aeff
Qa =
Ag
a) For flanges of square and rectangular
sections of uniform thickness
b E
For  1.40
t f
E  0.38 E 
 be = 1.92t 1 − 
f  b/t f 
otherwise be = b

f = Computed elastic compressive stress in the

stiffened element
= Pn /Aeff (may conservatively be taken equal to Fy.
b) For other uniformly compressed
elements

b E
For  1.49
t f
E  0.34 E 
 be = 1.92t 1 − 
f  b/t f 
otherwise be = b

f is taken as Fcr With Fcr calculated with Q= 1.0.

c) For axially loaded circular sections

E D E
For 0.11   0.45
Fy t Fy
0.038 E 2
 Q = Qa = +
Fy (D / t ) 3
D = Outside diamter, mm
t = Wall thickness, mm
Example: Design a double equal leg angle compression
member of width 416 mm connected by 10 mm gusset
plate and stay plates. Steel with F = 420 MPa is to be
used. P = 1700 kN, KL = 6m.

Solution:
Assume Slendeness Ratio, R = 90
KL E
  4.71 or Fe  0.44 Fy
r Fy
 2E  2  2 105
Fe = = = 243.69 MPa
(KL / r )2
(90)2

   
 Fy 
  420 
 
c Fcr = 0.658 e   Fy = 0.9  0.658 
  420 = 183.74 MPa
F 243.69

   
 
 Pu 17000 1000
Areq = = = 9252 mm 2 For 2 Ls
c Fcr 183.74
9252
Areq = = 4626 mm 2 For Single Angle
2
Trial Section - 1: 2L 203 x 203 x 12.7
A = 5000 mm2 For single angle
rmin = rx , of single angle = 63.5 mm
KL 6000
R = =  95
rmin 63.5
 2E  2  2 105
Fe = = = 218.72 MPa
(KL / r )2
(95)2

  Fy 
     420 
 
c Fcr = 0.658   Fy = 0.9  0.658   420 = 169.21 MPa
Fe   218.72 

   
 
Based on assumption that member is locally stable
c Pn = c Fcr  Areq = 169.21  (2  5000 ) / 1000
c Pn = 1692.1 kN
c Pn  Pu = 1700 kN Revise the section
Trial Section - 2: 2Ls 203 x 203 x 14.3
A = 5600 mm2 For single angle
rmin = rx , of single angle = 63.5 mm
KL 6000
R = =  95
rmin 63.5
 2E  2  2 105
Fe = = = 218.72 MPa
(KL / r )2
(95)2

  Fy 
     420 
 
c Fcr = 0.658   Fy = 0.9  0.658   420 = 169.21 MPa
Fe   218.72 

   
 
Based on assumption that member is locally stable
c Pn = c Fcr  Areq = 169.21  (2  5600) / 1000
c Pn = 1895.1 kN  Pu = 1700 kN (OK )
Check Local Stability
b 203 E
= = = 14.2  r = 0.45 = 9.8
t 14.3 Fy
  r = 9.8 Section is locally unstable
E
r = 0.91 = 19.9
Fy
E b E
For 0.45   0.91
Fy t Fy

 b  Fy
 Qs = 1.34 − 0.76  
t E
Q = Qs = 1.34 − 0.76(14.2 )
420
2 105
Q = 0.846 (15.4 % Reduction )
  
 QFy 

c Fcr = 0.658 e   QFy
F

 
 
 
 0.846420 

c Fcr = 0.9  0.658  218.72 
 0.846  420
 
c Fcr = 161.96 MPa
c Pn = c Fcr  Areq = 161.96  (2  5600) / 1000
c Pn = 1814.1 kN  Pu = 1700 kN (OK )
Example: Calculate the factored axial load capacity of
the shown 300 mm x 300 mm non-standard structural
tube having a thickness (t ) of 5mm and an effective
length of KL = 5.5 m. Use Fy = 345 MPa.

Solution:
A = 300 2 − 290 2 = 5900 mm 2
300 4 290 4
Ix = Ix = −  8560 10 4 mm 4
12 12
I 8560 10 4
rx = ry = =  120 mm
A 5900
Straight width, b = 300 − 2  (2  5) = 280
b 280
 = = = 56
t 5
 E 2 105
r = 1.40 = 1.40 = 33.7
Fy 345
 = 56  r = 33.7 Stiffened, locally unstable element

The section does not have unstiffened elements

Q = Qs  Qa For stiffened elements Qs = 1.0

Q = Qa
KL 5500
Overall Stabilty R = = = 45.83
rmin 120
E  0.38 E 
 be = 1.92t 1 − 
f  b/t f 
Assuming f = Fy
2 105  0.38 2 105 
be = 1.92  5 1 −  = 193.4 mm
345  56 345 
Ineffective width = 280 − 193.4 = 86.6 mm
Effective Area = 5900 − 86.6  4  5 = 4168 mm 2
Aeff 4168
Qa = = = 0.70
Ag 5900
 2E  2  2 105
Fe = = = 939.79 MPa
(KL / r ) 2
(45.83) 2

  
 QFy 

c Fcr = 0.658 e   QFy
F

 
 
 
 0.7345 

c Fcr = 0.9  0.658  939.79 
 0.7  345 = 195.19 MPa
 
c Pn = c Fcr  Areq = 195.19  5900 / 1000
c Pn = 1151.6 kN
If we ignore local buckling

  
 Fy 

c Fcr = 0.658 e   Fy
F

 
 
  345 
 
c Fcr = 0.9  0.658  939.79 
345 = 266.27 MPa
 
c Pn = c Fcr  Areq = 266.27  5900 / 1000
c Pn = 1571 kN
Example: Determine the compression capacity of the
given built-up I-section for an effective length of KL =
2.5m. Fy = 345 MPa. Ignore the residual stresses.

Solution:
A = 2 10  250 + 5  280
A = 6400 mm 2
10  250 4
Iy = 2
12
Iy = 2605 10 4 mm 4
I 2605 10 4
ry = = = 63.8 mm
A 6400
Locally Stability Check for Flange
bf 250
f = = = 12.5
2t f 2 10

kc =
4
(0.35 to 0.76)
h / tw
4
kc = = 0.534
280 / 5
Ekc 2 105  0.534
r = 0.64 = 0.64 = 11.26
Fy 345
 f = 12.5  r = 11.26 Flange is locally unstable
For Built − up Section
Ekc Ekc
0.64 = 11.26 and 1.17 = 20.6
Fy Fy
11.26    26.6
  b  Fy
 Qs = 1.415 − 0.65  
 t  Ekc

= 1.415 − 0.65  (12.5)

345
Qs = 0.953
2 10  0.534
5

Locally Stability Check for Web

h 280
w = = = 56
tw 5
E 2 105
r = 1.49 = 1.49 = 35.9
Fy 345
w = 56  r = 35.9 Web is locally unstable

Assume Q =1
KL 2500
Overall Stabilty R = = = 39.2
rmin 63.8
  2E  2  2 105
Fe = = = 1284.6 MPa
(KL / r )2
(39.2)2

  
 Fy 

f = Fcr = 0.658 e   Fy
F

 
 
  345 
 
c Fcr = 0.9  0.658  1284.6 
  345 = 308.3 MPa
 
E  0.34 E 
 be = 1.92t 1 − 
f  b/t f 

2 105  0.34 
be = 1.92  5 1− 308.3  = 206.7 mm
308.3  56 
Ineffective width = 280 − 206.7 = 73.3 mm
Aeff = 6400 − 73.3  5 = 6033 mm 2
 Aeff 6033
Qa = = = 0.943
Ag 6400
Q = Qa  Qs = 0.943  0.953
Q = 0.898
  
 QFy 

c Fcr = 0.658 e   QFy
F

 
 
 
 0.898345 

c Fcr = 0.9  0.658  1284.6 
 0.898  345 = 252.06 MPa
 
c Pn = c Fcr  Areq = 252.06  6400 / 1000
c Pn = 1613 kN
Concluded