Chapter 5

Design of
Compression Members and
Axially-Loaded Columns
 Lecture Outline
• Introduction
• Types of Compression Members
• Euler Buckling Load for Columns
• Effect of Initial Crookedness on Column
Buckling
• Inelastic Buckling of Columns
• Design Compressive Strength of Columns
 Lecture Outline
• Calculations Using LRFD Design Tables
• Limiting Slenderness Ratio
• Effective Length of Columns in Framed
Structures
• Design of Double-Angle Compression
Members
 Introduction
• Steel trusses have members that are subjected only to axial
compressive force. Such members are called “compression
members”.

• Axially-loaded columns in typical steel buildings

may be designed as pure compression members.

• In cases where a steel member is

subjected to combined axial force
and bending moment, such a member
is called “beam-column element”,
and its design requires a knowledge of column design
(which is covered in this chapter), and beam design
(which will be covered in the next chapter).
 Introduction (Cont’d)
The major differences between tension and compression
members are:
1) The gross area (Ag) is always used in the analysis of
compression members (not net area An).

2) Compression members are susceptible to buckling.

3) The failure stress (by bucking) is, in most cases,

less than the yield stress of steel used.
The buckling stress depends on the
member length and cross-section size.
 Types of Compression Members
There are many types of compression members
according to their application, e.g. hot-rolled sections
(W, S, M, T and L sections), built-up sections, and
round or rectangular tubes.

Rolled Shapes Built-up Sections Tubes

Euler Buckling Load for Columns
Assumptions:

• The column is perfectly straight,

prismatic, made of elastic homogenous
material.
• The column is pin-ended (hinged at both
end).
• The column remains elastic until buckling
occurs (elastic buckling).
Euler Buckling Load for Columns (Cont’d.)

At any distance “z”; M = P x

where:
M = the bending moment
about axis y-y, and can be
substituted for according to
the moment-curvature
relation to give:
2
d x
EI 2 = − M = − P x
dz
Euler Buckling Load for Columns (Cont’d.)

where: I = the moment of inertia about

the bending axis y-y
2
d x P
2
+ x = 0 ……………. (1)
dz EI
P
let = c (some constant) ……… (2)
2

EI
2
d x 2
eq. (1) 2
+c x = 0 ……………. (3)
dz
Euler Buckling Load for Columns (Cont’d.)

which is a 2nd order, linear, homogeneous

differential equation with constant
coefficients; and its solution is:

x = A sin(cz) + B cos(cz) …..….(4)

where: A and B are constants to be

obtained from the following boundary
conditions:
Euler Buckling Load for Columns (Cont’d.)

i. at z = 0, x = 0:

eq. (4) 0 = 0 + B => B = 0

eq. (4) x = A sin (cz) …….….(5)

ii. at z = L, x = 0:

eq.(5) 0 = A sin (cL)

which implies that either A = 0 (which

gives x = 0 for any “z” ➔ rejected)
Euler Buckling Load for Columns (Cont’d.)

OR cL = n, where n = 1,2,3,…etc. (n =

0 is rejected since it implies P = 0)

n …….….(6)
c=
L
and the eq. of the buckled shape (eq.(5))
becomes:

 n z  …….….(7)
x = A sin  
 L 
Euler Buckling Load for Columns (Cont’d.)

from eq.(2) & (6):

P n 2
n  EI
2 2 2

= 2 = P = 2
…….….(8)
EI L L
and eq.(8) is called “Euler Formula”; and
the smallest load at which buckling may
occur is obtained by putting n = 1; and is
called the critical buckling load:
 EI
2

eq. (8) Pe = 2
…….….(9)
L
Euler Buckling Load for Columns (Cont’d.)

and the buckled shape eq. is given by

eq. (7) with n = 1

  z  …….….(10)
eq. (7) x = A sin  
 L 

which is the equation of a half-sine wave

Euler Buckling Load for Columns (Cont’d.)

Comments:
1. The solution given by Eq.(10) is valid for
any value of “A” indicating that the
buckled shape is a half sine wave with an
indeterminate amplitude A > 0. This is
unstable behavior, implying that column
buckling is a failure mode.
2. If the first buckling mode is prevented,
by providing lateral support at mid-
height, then Pcr will occur at n = 2
Euler Buckling Load for Columns (Cont’d.)

4 EI  EI
2 2
eq. (8) Pe = =
( L 2)
2 2
L

 EI
2
or Pe = …….….(11)
(K L )
2

where: K = 1/2 in this case.

and the buckled shape equation will be

given by eq.(7) with n = 2
Euler Buckling Load for Columns (Cont’d.)

 2 z  …….….(12)
eq. (7) x = A sin  
 L 
which is the equation
of a complete sine
wave.
Euler Buckling Load for Columns (Cont’d.)

In eq.(11) above, (KL) is called the

effective buckling length of the column
which is the chord length of that part of
the column that takes the shape of a
half sine wave at buckling (or the chord
length between the points of inflection,
M = 0 points).
  2EI
Euler Buckling Load for Columns (Cont’d.) Pe =
(K L )
2

Comments:
 EAg
2

Let (I=r2Ag) in eq.(11) to get: Pe =

(K L r )
2

then divide both sides by “Ag” to get:

Pe Euler buckling stress

Fe = =
Ag or Elastic critical stress

 E
2
…….(13)
Fe =
(K L r )
2
(AISC - LRFD Eq. (E3-4), p. 16.1-33)
Euler Buckling Load for Columns (Cont’d.)

Eq.(13) above suggests that

buckling occurs about the
axis that has the largest
slenderness ratio (KL/r). For
symmetric shapes, you must
use the larger of (KL)x/rx
and (KL)y/ry.

The following table (from LRFD Manual, p.

16.1-240) shows the values of the
effective length factor “K” for different
boundary conditions.
Effective Buckling Length Factor
TABLE C-C2.2 (p. 16.1-240)
 Effect of Initial Crookedness
on Column Buckling
In reality, perfectly straight members
cannot be manufactured.
• For any steel member to be acceptable,
we must have: L
e
1000
• For rolled steel sections, the average
value of “e” is: L
e=
1500
Effect of Initial Crookedness on Column
Buckling (Cont’d.)

• Based on experimental and

theoretical studies, it was
found that Pcr decreases
when initial crookedness
exists according to the
shown figure, therefore;
the LRFD specifications re-
writes Eq. (11) for
e = L/1500 as follows:

Pcr = 0.877Pe
Effect of Initial Crookedness on Column Buckling (Cont’d.)

and also Eq. (13) becomes:

AISC - LRFD Eq. (E3-3),

Fcr = 0.877Fe p. 16.1-33

 2E
Notes: Fe =
(K L r )
2

1. LRFD Eq. (E3-3), above, is used only for

(elastic bucking):

KL E
 4.71 (or Fe  0.44Fy )
r Fy
Effect of Initial Crookedness on Column Buckling (Cont’d.)

2. LRFD design requirement for columns is

fc Pn > Pu , where: fc = 0.9; and

Pn = Fcr . Ag AISC - LRFD Eq. (E3-1), p. 16.1-33

thus;
fc Pn = 0.9Ag (0.877Fe ) = 0.79Ag Fe
 2E
Fe =
(K L r )
2
 Inelastic Buckling of Columns
• Due to the presence of
residual compressive stresses
“frc” at some parts of the
column x-sec. (e.g. the
flange tips and web mid-
height in W-shapes), these
parts could yield locally
before buckling of the
column occurs. This buckling
is then called inelastic
buckling.
  2E
Inelastic Buckling of Columns (Cont’d.) Fe =
(K L r )
2

• According to LRFD specifications:

when e = L/1500, and

KL E
 4.71 (or Fe  0.44Fy )
r Fy
the critical inelastic buckling stress is given
by:

 Fy
 AISC - LRFD Eq. (E3-2),
Fcr = 0.658 Fe
 Fy p. 16.1-33
 
Inelastic Buckling of Columns (Cont’d.)

LRFD design requirement for columns is:fcPn>Pu

where: fc = 0.9; and

Pn = Fcr . Ag AISC - LRFD Eq. (E3-1), p. 16.1-33

thus;

 
Fy

fc Pn = 0.90
0.85Agg 0.658 Fe
 Fy
 
 Summary of the Design
Compressive Strength of Columns

LRFD design requirement for columns states:

fc Pn > Pu where fc = 0.9; and

Pn = Fcr . Ag AISC - LRFD Eq. (E3-1), p. 16.1-33

hence: fc Pn = 0.9 Fcr . Ag

In Summary (Cont’d.)

Elastic Buckling

a) when KL
 4.71
E
(or Fe  0.44Fy )
r Fy

AISC - LRFD Eq. (E3-3),

Fcr = 0.877Fe p. 16.1-33

hence: fc Pn = 0.9 Fcr . Ag

where:

 E2
(AISC - LRFD Eq. (E3-4),
Fe =
(K L r ) p. 16.1-33)
2
In Summary (Cont’d.)
 2E
Fe =
(K L r )
2

Inelastic Buckling

b) when KL
 4.71
E
(or Fe  0.44Fy )
r Fy

 Fy
 AISC - LRFD Eq. (E3-2),
Fcr = 0.658 Fe
 Fy p. 16.1-33
 

hence: fc Pn = 0.9 Fcr . Ag

Buckling of Columns (Cont’d.)

• The following figure summarizes the two cases

of elastic and inelastic buckling:
 Limiting Slenderness Ratio
For members designed on the basis of
compression, the slenderness ratio KL/r
preferably should not exceed 200 (AISC -
LRFD-E2, p. 16.1-33).

KL/r should be calculated for all principle axes:

 KL 
   200
 r max
Buckling of Columns (Cont’d.)

• However, Fcr given by the above 2 equations is

correct only if local buckling of the thin
rectangular elements in the column x-section
does not occur before column buckling takes
place.
 Local Buckling of Columns

Local buckling can be avoided

only if the width-to-
thickness ratio (b/t) of
each element in the cross
section of the column does
not exceed a specific value
called lr as given in Table
B4.1 of the LRFD, pages
16.1-16 to 16.1-18 (provided
after a few slides).
 Local Buckling of Columns

b b b
t t t

b b

t b

(b/t) for unstiffened* compression elements

(*= Elements having one free edge parallel to
the direction of load)
 Local Buckling of Columns
b

t
t b b t b

(b/t) for stiffened* compression elements

(*= Elements having lateral supports along both
edges parallel to the direction of load)
Notes:
If (b/t) of some elements in the column x-section exceeds
the limits given in Table B4.1, then the column design
strength must be reduced according to the equations given
in the LRFD Manual, Section E. However, this case is
beyond the scope of this course.
 Calculations Using LRFD
Design Tables
The LRFD design Tables are provided for steel
type A572 Grade 50 ; Fy=50 ksi. For other types of
steel, manual calculations are a must.

These tables are based on the assumption that

buckling occurs about the y-axis:
[i.e. (KL/r)y > (KL/r)x] and the user should use the
table with (KL)y.
This is not the case for many columns and
therefore a modified (KL)y-eq. should be used.
 Calculations Using LRFD Design Tables
(Cont’d)
To compute the design strength using the Table, one
should follow the procedure provided below:

• If
(KL )x  (KL )y , then:
rx ry

 KL 
➢   governs the design (as assumed by LRFD Manual)
 r y

➢ User should enter the table with (KL)y

➢ Get fc Pn based on (KL)y

 Calculations Using LRFD Design Tables
(Cont’d)

• Else if
(KL )x  (KL )y , then:
rx ry

 KL 
➢   governs the design
 r x

➢  KL  r
User should enter the table with (KL)eq.=   y
 r x

➢ Get fc Pn based on (KL)eq.

 Calculations Using
LRFD Design Tables
(Cont’d)

LRFD Manual, starting from

Page 4-10
 Example 1

An A572 Grade 50 axially loaded column

with a cross-section of W14x90, has
effective length defined by KL = 20’ for
both principle axes.
Using LRFD, show that local buckling
does not occur and then compute the
column design strength; both manually
and using the LRFD tables.
 Solution
For W14x90:
Ag = 26.5 in2 rx = 6.14 in. ry= 3.70 in.

Check local buckling:

bf
= 10.2  0.56 E / Fy = 13.48 (Table B4.1 p. 16.1-16)
2t f
No F.L.B
h
= 25.9  1.49 E / Fy = 35.88 (Table B4.1 p. 16.1-16)
tw
No W.L.B
Kdes = K design
Kdet = K detailing
Check overall (global) buckling:

(KL/r)x = KLx / rx = 20.0 (12) / 6.14 = 39.09

(KL/r)y = KLy / ry = 20.0 (12) / 3.70 = 64.86

(KL/r)max = (KL/r)y = 64.86 < 200 (ok)

Note:
Since KLx = KLy = 20’ and rx > ry then the
maximum (KL/r) is due to buckling about the y-
axis (i.e. for r = ry).
Using Table C-E3.1 (p.16.1-260, AISC–LRFD):

(113 and 22 are limiting values of KL/r and Fe)

KL E 
= 64.86  4.71 = 113 
r Fy 

 2E 
(or = 67.97  0.44Fy = 22) 
(K L r )
2

Indicating a case of inelastic buckling
Therefore:
 Fy
 AISC – LRFD
Fcr = 0.658 Fe
 Fy Eq. (E3-2),
  p. 16.1-33

 2E
Fe = = 67.97 ksi
( KL r )
2

 50

Fcr = 0.658 67.97
 50 = 36.75 ksi
 
hence: fc Pn = 0.9 Fcr . Ag
fc Pn = 0.9 (36.75) (26.5) = 876 kips
Using the LRFD Tables:

Check local buckling:

No Need for it

Check overall (global) buckling:

(KL/r)x = KLx / rx = 20.0 (12) / 6.14 = 39.09

(KL/r)y = KLy / ry = 20.0 (12) / 3.70 = 64.86

(KL/r)max = (KL/r)y = 64.86 < 200 (ok)

From the LRFD Column Design Tables,
Table 4-1, p. 4-13:

For KLy = 20 ft :

fc Pn = 876 kips
 Example 2

An A572 Grade 50 axially loaded column

with a cross-section of W14x90, has
effective lengths defined by (KL)x = 20’
and (KL)y = 10’.
Using LRFD, show that local buckling
does not occur and then compute the
column design strength.
 Solution
For W14x90:
Ag = 26.5 in2 rx = 6.14 in. ry= 3.70 in.

Check local buckling:

bf
= 10.2  0.56 E / Fy = 13.48 (Table B4.1 p. 16.1-16)
2t f
No F.L.B
h
= 25.9  1.49 E / Fy = 35.88 (Table B4.1 p. 16.1-16)
tw
No W.L.B
Kdes = K design
Kdet = K detailing
Explicit Calculations for Design Strength:
To find maximum (KL/r) the slenderness ratios
about the two principle axes are calculated and
the maximum is selected as follows:

 KL  ( K L )x 20(12)
KLx = 20’,   = = = 39.09
 r x rx 6.14
 KL  ( K L ) y 10(12)
KLy = 10’,   = = = 32.43
 r y ry 3.70
 (KL/r)max = 39.09 < 200 (ok)
 Using (KL/r)max = 39.09

KL E AISC – LRFD
= 39.09  4.71 = 113 Table C-E3.1
r Fy (113 and 22
are limiting
 E
2

values of
(or = 187.14  0.44Fy = 22) KL/r and F ),e

( 39.09 )
2 p. 16.1-260

Therefore, we have inelastic buckling:

 Fy
 AISC – LRFD, Eq. (E3-2),
Fcr =  0.658  Fy p. 16.1-33
Fe

 
 2E
Fe = = 187.14 ksi
( 39.09 )
2

 50

Fcr = 0.658187.14  50 = 44.71 ksi
 
hence:
fc Pn =0.9 Fcr.Ag =0.9(44.71)(26.5)=1066 kips
Using the LRFD Tables:

Check local buckling:

No Need for it

Check overall (global) buckling:

(KL/r)x = KLx / rx = 20.0 (12) / 6.14 = 39.09

(KL/r)y = KLy / ry = 10.0 (12) / 3.70 = 32.43

(KL/r)max = (KL/r)x = 39.09 < 200 (ok)

We can get the same value for (fc Pn) from the
LRFD Column Design Tables, Table 4-1, p. 4-13:

 KL  39.09x3.7
(KL ) table = (KL ) eq. =  ry = = 12.05'
 r x 12

(KL)table (ft) 12 12.05 13

fc Pn (kips) 1070 ??? 1050

 1070 − 1050 
fc Pn = 1070 +  (12.05 − 12.0) = 1069 kips
 12 − 13 
 Example 3
An A572 Grade 50 column is loaded with
an ultimate axial compression Pu = 1000
kips. Knowing that (KL)x=20’ and
(KL)y=10’, find the lightest W10-section
that can withstand this load satisfactorily.
Solution
Assuming that (KL)y = 10’ controls the design,
using the LRFD design tables, and selecting the
section that gives fcPn > 1000 kips:
we select W10x100 section.
LRFD Tables page 4-19, fcPn = 1140 kips

Check overall (global) buckling:

(KL/r)x = KLx / rx = 20.0 (12) / 4.6 = 52.17

(KL/r)y = KLy / ry = 10.0 (12) / 2.65 = 45.28

(KL/r)max = (KL/r)x = 52.17 < 200 (ok)

To be able to use the LRFD Column Design
Tables:
 KL  52.17 x 2.65
(KL ) table = (KL ) eq. =  ry = = 11.52'
 r x 12

Thus: fcPn = 1084.4 kips > 1000

kips:

The selected W10x100 section is

safe and economic
 Effective Length of Columns in
Framed Structures
• The idealized end conditions
given earlier for columns are
most of the time not realistic,
especially for real framed
structures (i.e., having rigid
connections between columns
and beams).
• Therefore, LRFD provides the
following procedure to
determine the approximate
values of (K) in order to
calculate the effective buckling
length (KL).
 Effective Length of Columns in
Framed Structures (Cont’d)
Effective Buckling Length of Column A-B:
• Step (1)
Calculate:

 Ic 
 EI c 
     
GA =  c  A
L  Lc A
 EI g 
GA =
  L   Ig 
 g A   
 Lg A

GA = Relative rigidity Factor at end A

 Effective Length of Columns in
Framed Structures (Cont’d)
where:
Ic = moment of inertia of any column
connected to joint A about the
buckling axis.
Lc = the physical length of that column.
Ig = moment of inertia of any girder
connected to joint A about the
buckling axis.
Lg = the physical length of that girder.
 = indicates a summation of all members rigidly connected to joint
A and lying in the plane of buckling.
 Effective Length of Columns in
Framed Structures (Cont’d)
• Step (2):
Compute GB in a way similar to calculation of GA.

 Ic 
 EI c 
     
GB =  c  B
L  Lc B
 EI g 
GB =
  L   Ig 
 g B   
 Lg B

GB = Relative rigidity Factor at end B

 Effective Length of Columns in
Framed Structures (Cont’d)
• Step (3):
You can use the following equations to find “K”:
1. For columns in braced frames:
3G A GB + 1.4(G A + GB ) + 0.64
K=
3G A GB + 2(G A + GB ) + 1.28

2. For columns in unbraced frames:

1.6G A GB + 4(G A + GB ) + 7.5

K=
G A + GB + 7.5
Effective Length of Columns in Frames Structures (Cont’d.)
Alignment Charts (AISC manual pp. 16.1-241 and 16.1-242)

Inhibited means restrained / prevented (BRACED)

 Effective Length of Columns in
Framed Structures (Cont’d)
Notes:
 Ic 
  L 
 c
• In GA =
 Ig 
, the subscripts A and B refer to the joints at the
  L 
 g
two ends of the column section being considered.
• For pin-supported columns, G is theoretically infinity, but
practically it is not friction-free pin and therefore G is taken as
“10” (as per LRFD – p. 16.1-241).
• If the column end is fixed to a properly designed footing, G
may be taken as “1.0” (as per LRFD – p. 16.1-241).
• For braced columns 0.5 ≤ K ≤ 1.0 whereas for unbraced
columns K > 1.0
 Example 4
The figure below shows an unbraced plane frame made of A572
Grade 50 steel. All columns are rolled W12x120 and placed with
their cross-section x-axis (strong axis) perpendicular to the plane
of the frame. At the support joints, use the recommended G values
given in the last paragraph in LRFD p. 16.1-241. Assuming that
(KL)y is not controlling the design, find (KL)x, and the design
strength (fcPn) for columns 1 to 5.

W30x173 8 W30x173
5 11
3 5 15 ft
W30x116 W30x173 W30x173 W30x116
2
4 7 10 13
1 2 4 20 ft
1 3 9 12
6

30 ft 40 ft 40 ft 30 ft
Solution

For W12x120: Ag = 35.3 in2 , Ix = 1070 in4

rx = 5.51 in, ry = 3.13 in

For W30x173: Ix = 8200 in4

For W30x116: Ix = 4930 in4

From the LRFD Manual p. 16.1-241:

G1 = 10, G3 = 10, G6 = 1
At the interior (rigid) joints of the plane
frame:
I
(
  L  c 1070 20 )
G2 =
I
  L  g
=
(4930
30
)
= 0.326

I
(
  L  c 1070 20 + 107015 )
G4 =
I
  L  g
=
(4930 + 8200
30 40
)
= 0.338

W30x173 8 W30x173
5 11
3 5 15 ft
W30x116 W30x173 W30x173 W30x116
2
4 7 10 13
1 2 4 20 ft
1 3 9 12
6
 I
(
  L  c 107015 )
G5 =
I
  L  g (
=
8200
40
)= 0.348
5
W30x173 8 W30x173 11
3 5
W30x116 W30x173 W30x173 W30x116
2
4 7 10 13
1 2 4
1 3 9 12
6

I
(
  L  c 1070 20 + 107015 )
30 ft 40 ft 40 ft 30 ft

G7 =
I
  L g
 
=
(8200
40
+ 8200
40
)
= 0.380

I
  L  c (1070
15
)
G8 =
I
  L  g (
=
8200 + 8200
40 40
)
= 0.174
Using the LRFD alignment char, we get:

G1 = 10

K1 = 1.74

G2 = 0.326

K1 =1.741, K2 =1.744, K3 =1.129, K4 =1.239, K5 =1.104

Now we can calculate (KL)x for each column as
follows:
(KL)x = Kx L

For example, for column (1)

(KL)x-1 = 1.741 (20)

(KL)eq-1 = [(KL)x1/(rx)] x ry
=[1.741(20)/5.51] x 3.13 = 19.78`

Using LRFD Design Tables for A572 Grade 50,

and W12x120 and the calculated (KL)x for
each column we can get fc Pn.
 Controlling
(KL)eq=
(KL)max
fc Pn
Col Sec. (KL)y
[(KL)x / rx ] (ry) (kips)

[(1.741x20)/5.51]x3.13
1 19.780' 1041.0
= 19.780’
[(1.744x20)/5.51]x3.13
2 Not 19.814' 1039.3
= 19.814’
Controlling
the Design [(1.129x15)/5.51]x3.13
3 W12x120
(as given 9.620’ 1441.4
= 9.620’
in the
question) [(1.239x20)/5.51]x3.13 1286.9
4 14.076' 6
= 14.076’
5 [(1.104x15)/5.51]x3.13 1447.7
9.407' 9
= 9.407’
 SUMMARY
Design Procedure of Compression Members
(Manual Calculations)
(1) Check that local buckling does not control the
design using:
b
l =  lr
t
where lr is as given in Table B4.1, LRFD, p. 16.1-
16 to 16.1-18.
Note that this step should be done for every element
of the cross section (i.e., web and flange).
Design Procedure of Compression Members

(2) Calculate the effective buckling length factor (K) in

both x and y directions. The K-factor is to be found
as follows:

(i) from Table C-C2.2 (p. 16.1-240) for

compression truss members or isolated columns.
(ii) from Alignment Charts (p. 16.1-241 and 16.1-
242) for framed columns.
 Design Procedure of Compression
Members (Cont’d)
(3) Find the maximum slenderness ratios about the two
principle axes as follows:

 KL  ( KL) x
  =  200
 r x rx

 KL  ( KL) y
  =  200
 r y ry
 Design Procedure of Compression
Members (Cont’d)
(4) Based on (kL/r)max:
Elastic Buckling

a) when KL
 4.71
E
(or Fe  0.44Fy )
r Fy

AISC - LRFD Eq. (E3-3),

Fcr = 0.877Fe p. 16.1-33

where:
 E2
(AISC - LRFD Eq. (E3-4),
Fe =
(K L r ) p. 16.1-33)
2
 Design Procedure of Compression
Members (Cont’d)
Or: Inelastic Buckling

b) when KL
 4.71
E
(or Fe  0.44Fy )
r Fy

 Fy
 AISC - LRFD Eq. (E3-2),
Fcr = 0.658 Fe
 Fy p. 16.1-33
 
 Design Procedure of Compression
Members (Cont’d)

(5) Finally: LRFD design requirement:

fc Pn > Pu where fc = 0.9; and

Pn = Fcr . Ag AISC - LRFD Eq. (E3-1), p. 16.1-33

hence: fc Pn = 0.9 Fcr . Ag

Design Procedure of Compression Members
(Design Tables)
(1) Find the maximum slenderness ratios about the two
principle axes as follows:

 KL  ( KL) x
  =  200
 r x rx

 KL  ( KL) y
  =  200
 r y ry
 Calculations Using LRFD Design Tables
(Cont’d)
To compute the design strength using the Table, one
should follow the procedure provided below:

(2) If
(KL )x  (KL )y , then:
rx ry

 KL 
➢   governs the design (as assumed by LRFD Manual)
 r y

➢ User should enter the table with (KL)y

➢ Get fc Pn based on (KL)y

 Calculations Using LRFD Design Tables
(Cont’d)

(3) Else if
(KL )x  (KL )y , then:
rx ry

 KL 
➢   governs the design
 r x

 KL  r
➢ User should enter the table with (KL)eq.=   y
 r x

➢ Get fc Pn based on (KL)eq.

 This section “Design of Double-Angle Compression Members ” is NOT included in
the final exam

Design of Double-Angle
Compression Members
• The two angles that make up the
compression member must be connected
by spacer plates (or connectors) placed
at some suitable distance “a” such that:

a   KL 
   0.75  
 r z one  r max for the column
angle
Design of Double-Angle Compression Members (Cont’d.)
P/2 P/2 P NOT included
in the final
exam

P/2 P/2 P

Two Columns - Single Angle Each Double AnglehColumn

Two Column – Single Angle Each
(Angles are not Connected) Doubleare
(Angles Angle Column
Connected)
(Angles are not Connected) (Angles are Connected)
Design of Double-Angle Compression Members NOT included
(Cont’d.) in the final
exam

• From the above equation, compute the max.

allowed spacing “a”, and then the min. number
of connectors, n, to be used is n = [(L/a) - 1]
rounded off to the next integer.

• The buckling about the y-y axis, shown above,

relative shear deformation between the two
angles reduces their design strength fcPny;
therefore, (KL/r)y should be replaced by
(KL/r)m according to LRFD Equations (E6-1)
and (E6-2) p.16.1-37 of the AISC - LRFD.
Design of Double-Angle Compression Members (Cont’d.)
NOT included
2
 KL   a 
2 in the final
 KL  (Eq. E6-1),
 =   + 
exam
 p. 16.1-37
 r m  r o  r i 
for snug-tight bolted connectors
2
0.82  a 
2 2
 KL   KL  (Eq. E6-2),
  =   +  
 r m  r o 1 +   r ib 
2
( ) p. 16.1-37

for welded and fully-tensioned bolted connectors where a = (h / rz)

connectors.
A "snug-tight" condition is defined as the tightness attained by the full
effort of a person using a spud wrench.
Design of Double-Angle Compression Members (Cont’d.)
NOT included in
the final exam

• The strength of the member (i.e., Fcr) will be

based on the maximum of (KL/r)x or (KL/r)m.

• Note that (KL/r)m is always equal or bigger

than (KL/r)y.

Notes:
1. LRFD requires a min. of two connectors (i.e.
3 spaces) to be used. The number of
connectors is based on the buckling about y-
axis only.
Design of Double-Angle Compression Members (Cont’d.)
2. Tabulated column load values (fcPn) for double
angle sections satisfies the condition of
(a/rz)for one angle < 0.75 (KL/r)m, while it
requires an extra check for (a/rz)for one angle <
0.75 (KL/r)x. NOT included in
the final exam
3. When back-to-back distance between the
two angles is not 3/8”, we modify (KL)y by
using:
 r y for 3/8" spacing 
( KL y )equiv . = ( KL )y 
 r y for new spacing 

 
and enter the tables with (KLy)equiv (instead of
KLy) to find fcPny.
 NOT included in
Example 5 the final exam

Find the lightest double angle section (i.e.,

  ) for Pu = 70 kips, A36 steel, 3/8" thick.
welded connectors, and (KL)x = (KL)y = 8’.

Solution

From tables in pages 4-118 to 4-156

look for fc Pn for each of (KL)x = 8' and
(KL)y = 8' so that fcPn ≥ Pu=70 kips and
Ag is minimum:
1) For Equal Leg Double Angle NOT included in
the final exam
Sections
For   4 x 4 x 1/4 (Ag = 3.88 in2,
fPnx = 86.2 kips, fPny = 81.4 kips)
(w / 3 connectors)

For   3.5 x 3.5 x 5/16 (Ag= 4.18 in2,

fPnx = 89.2 kips, fPny = 103 kips)
(w / 3 connectors)
 NOT included in the
final exam

For   3 x 3 x 3/8 (Ag = 4.22 in2,

fPnx = 76 kips, fPny = 102 kips)
(w / 3 connectors)

Choose   4 x 4 x 1/4 with 3 connectors

(Ag = 3.88 in2, rx = 1.25 in, rz = 0.795

in)
Check for (a/rz)for one angle < 0.75 (KL/r)x

a = (KL)x / 4 = 8*12/4 = 24"

NOT included in the
final exam

(a/rz)for one angle = 24/0.795 = 30.2

0.75 (KL/r)x = 0.75 (8*12/ 1.25) = 57.6

Therefore (a/rz)for one angle < 0.75 (KL/r)x

OK.
 2) For Equal Leg Double Angle Sections
This part is left for the student to perform during the class

NOT
included in
the final
exam
NOT included in the final exam