Compression Members

A.K.Mandal

A structural member which is subjected to compressive forces along its axis is called a
compressive member. Different terms are used to designate a compressive member depending
upon its position in structures. Column, stanchion or post is a vertical compression member
supporting floors or girders in a building. These compression members are subjected to heavy loads.
Strut is a compression member used in roof truss and bracings. It is of small span and lightly loaded
compression member. The principal rafter is a top chord member in a roof truss and boom is the
principal compression member in a crane.
The behaviour of an initially straight column, compressed by gradually increasing axial forces will
depend on its length. If the column is “short”, the applied forces will cause a compressive strain,
which results in the shortening of the column in the direction of the applied forces. If the loading is
gradually increased, the shortening continues until the column yields or “squashes”. However, if the
column is “long”, similar axial shortening is observed only at the initial stages of loading. Thereafter,
as the applied forces are increased, the column becomes unstable and develops a deformation in a
direction normal to the loading axis and its axis is no longer straight. The column is said to have
“buckled”.
Structural steel has high yield strength and ultimate strength compared with other construction
materials. So the compression members made of steel tend to be slender compared with RCC or
prestressed concrete compression members. Buckling is of particular interest while using slender
steel members.
Traditionally, the design of compression members was based on Euler analysis of ideal columns
which gives an upper band to the buckling load. However, practical columns are far from ideal and
buckle at much lower loads. The first significant step in the design of such columns was the use of
Perry Robertsons curves. Modern codes advocate the use of multiple-column curves for design.
Although these design procedures are more accurate in predicting the buckling load of practical
columns, Euler’s theory helps in the understanding of behaviour of slender columns and so it is
reviewed first.
In Euler’s analysis, the load at which the initially straight long column, subjected to axial load just
buckles or bends is called Euler’s load, or Euler’s critical load, or simply critical load. For a column of
length L with uniform cross-section and hinged (or pinned) at both ends, this critical load, Pcr is
given by
π 2 EI
Pcr =
L2

It is often convenient to study the onset of elastic buckling in terms of average compressive
stress (rather than force). The average compressive stress at buckling fcr is given by
P cr π 2 EI π2E π2E π2E
fcr = = AL 2
= A = L 2
= λ2
A L2 � �
I r

where, A = cross-sectional area, r = radius of gyration of the cross-section (I = Ar2) and

λ = slenderness ratio of column = L/r.
2

The equation fcr = (π2 E)/λ2 implies that the critical stress of a column is inversely proportional
to the square of the slenderness ratio of the column as shown in Fig 1.

fcr f
Elastic buckling stress
(fcr) defined by π2E/λ2 Yield plateau
fy

λ = L/r εy ε

Fig 1 Euler buckling relation between fcr & λ Fig 2 Idealised elasto-plastic relation for steel

From the idealized elasto-plastic stress-strain curve (Fig 2), it is evident that the column can
resist only a maximum force given by fy.A, when plastic squashing failure would occur by plastic
yielding of the entire cross section. This means that the stress at failure of a column can never
exceed fy, shown by A-A’ in Fig 3(a).
From Fig 1, it is obvious that the column would fail by buckling at a stress given by π2E/λ2. This is
indicated by B-B’ in Fig 3(a), which combines the two types of behaviour just described. The two
curves intersect at C. Obviously the column will fail when the axial compressive stress equals or
exceeds the values defined by ACB. In the region AC, where the slenderness values are low, the
column fails by yielding. In the region CB, the failure will be triggered by buckling.

fcr Fcr/fy
Plastic yield defined Plastic yield
by fcr = fy
fy 1.0
Elastic buckling Elastic buckling
defined by π2E/λ2

λc λ = L/r 1.0 ½
𝜆𝜆̅ = �fy /fcr �

Fig 3(a) Strength curve for axially loaded Fig 3(b) Strength curve in a non-dimensional
pin-ended initially straight column form

The changeover from yielding to buckling failure occurs at the point C [Fig 3(a)], defined by a
slenderness ratio given λc and is evaluated from
π2E
fy =
λc 2
 3

E
or λc = π �
f y

Plots of the type Fig 3(a) is sometimes presented in a non-dimensional form as illustrated in
Fig 3(b). Here fcr/fy is plotted against a generalized slenderness given by

λ fy
λ� = = �
λc f cr

The single plot [shown in Fig 3(b)] is used as design curve as per Euler analysis to define the
strength of all axially loaded initially straight columns with both ends hinged. A close look on Euler’s
formula shows that critical load for buckling depends not on the strength of material, but only on
dimensions and modulus of elasticity.
The highly idealized straight form assumed in Euler’s analysis cannot be achieved in practice.
Members are never perfectly straight and they can never be loaded exactly at the centroid of the
cross-section. Deviation from ideal elastic plastic behaviour defined in Fig 2 are encountered due to
strain hardening at high strain and absence of clearly defined yield point in some steel. Moreover,
residual stresses locked-in during the process of rolling provide an added complexity.
Thus the three components which contribute to a reduction in the actual strength of columns,
compared with the values from ideal design curve are
(a) Initial imperfection or initial bow
(b) Eccentricity in application of loads
(c) Residual stresses locked into the cross section.
In practice, a loaded column may experience
fcr
most, if not all, of the effects listed above i.e. out of
Data from collapse tests
straightness, eccentricity of loading, residual stresses fy
and lack of clearly defined yield point and strain
hardening occurring simultaneously. When all the Theoretical
effects are put together, the resulting column elastic buckling
strength curve is generally of the form shown in Fig 4.
All design codes provide column curves based on this
philosophy. Thus a lower bound curve or a family of
Lower bound curve
such curves is used in design. IS code adopted
multiple design curves for four buckling classes a, b, c λc λ
and d based on various cross-sections and are shown Fig 4 Column strength curve
in Fig 8 (page 35) of IS 800: 2007.

Concept of Effective Lengths

So far, our discussion has been centered around pin-ended (or hinged at both ends) columns.
The boundary conditions of a column, however, may be idealized in one of the following ways
1. Both ends pin jointed (i.e. the case considered so far)
2. Both ends fixed
3. One end fixed and the other end pinned
4. One end fixed and the other end free.
4

Euler’s expression for critical loads for all the four boundary conditions are given below and the
corresponding buckled shapes are shown in Fig 5.
π 2 EI
Both ends pinned: Pcr =
L2
4π 2 EI π 2 EI
Both ends fixed: Pcr = =
L2 (L/2)2
2π 2 EI π 2 EI
One end fixed and other end pinned: Pcr = = 2
L2 �L/√2�
π 2 EI π 2 EI
One end fixed and other end free: Pcr = =
4L 2 (2L)2

Hinged Fixed Hinged Free

Point of
L/√2
inflection
L L/2

Hinged Fixed Fixed Fixed

2L

Fig 5 Buckled mode for different

end conditions

Using the column, pinned at both ends as the basis of comparison, the critical load in all the
above cases can be obtained by using the concept of “effective length” KL, where L is the centre to
centre distance of the compression member between the restrained ends and K is a factor. So
effective length for various end conditions are as follows –
Boundary condition Theory Code value
Both ends pin jointed 1.0 L 1.0 L
Both ends fixed 0.5 L 0.65 L
One end fixed, and other end pinned 0.707 L 0.8 L
One end fixed, and other end free 2.0 L 2.0 L
Note: Refer Table No 11, page 45 of IS 800: 2007 for further details.
It can be seen that the effective length corresponds to the distance between the point of
inflection (point of contraflexure) in the buckled mode. The effective length of a given column with
given end conditions can be defined as the length of an equivalent column of same material and
cross-section with hinged ends, having the value of critical load equal to that of the given column.
The smaller the effective length of a column, the smaller is the danger of its lateral buckling and the
greater is its load carrying capacity.
Effective lengths in different planes
The concept of effective length, as explained above, is for compression members supported only
at their ends. However, in practice columns never exist in isolation. If there is a possibility of
providing additional lateral restraint (i.e. by bracing) to strengthen the weaker axis of the
compression member, effective length of the member can be reduced and consequently buckling
load will increase. If a compression member is supported differently with respect to each of its
 5

principal axes, the effective length will be different for two directions. For the pin-ended column,
shown in Fig 6, the column is supported differently with respect to each of its principal axes as a
brace has been provided at the mid-height of the column preventing translation perpendicular to
the weak axis of the column cross-section. Consequently, the weak axis is strengthened. So if the
member is to buckle about the major axis, the effective length will be L, while for buckling about
minor axis the effective length will be L/2.

Fig 6 Column supported differently with respect to its principal axes

Effective Length as per IS 800: 2007

(a) If end conditions can be assessed:
When the boundary conditions in the plane of buckling can be assessed, the effective length KL
can be calculated on the basis of Table 11 of IS 800: 2007.
(b) Compression members in trusses:
(i) In case of bolted, riveted or welded trusses and braced frames, the effective length KL
shall be taken as 0.7 to 1.0 times the actual length, depending upon the degree of end
restraints provided. For bucking in the plane perpendicular to the plane of truss, the
effective length may be taken as the actual length.
(ii) Design of angle struts shall be done as specified in cl. 7.5 of IS 800.
(c) In frames:
In the frame analysis, if the deformed shape is not considered (second order or advanced
analysis is not used), the effective length depends upon stiffnesses of the members meeting at the
joint. The method of finding the effective length factor K is shown in Annex D of IS 800.
(d) In case of stepped columns:
Expressions for finding the effective length factor for various stepped columns are presented in
Annex D2 & D3 of IS 800.

Appropriate radius of gyration

Appropriate radius of gyration means the radius of gyration of compression member about the
axis of buckling. For example, in case of column shown in Fig 6, when length of column is taken as L,
the radius of gyration about z-z axis should be considered. For buckling about y-y axis, the length of
column is L/2 and radius of gyration about y-y axis is to be considered. The maximum slenderness
6

ratio governs the design strength. If the length of the column to be considered is the same for
buckling about both the axes, naturally the governing slenderness rato is KL/rmin.

Design compressive stress and strength

IS 800: 2007 provides two methods to calculate the design compressive stress fcd of axially
loaded compression members. Any one method can be used.

(a) Method-I: Equation given in cl. no. 7.1.2.1 of IS 800 can be used to calculate fcd.
(b) Method-II: The value of fcd may be directly obtained from Table 9 of IS 800.

Since use of equation to calculate fcd is quite cumbersome, it is preferable to use Method-II.
Steps in Method-II
1. Decide the buckling class of the cross-section as per Table 10 of IS Code.
2. Calculate effective slenderness ratio λ = KL/r.
3. Obtain the value of fcd from Table 9 of IS Code based on buckling class, effective slenderness
ratio and yield stress of column material.
NOTE: Angle sections are commonly used in trusses, bracings, towers etc. When single angle,
connected by one leg either directly or through gusset plate, becomes eccentrically loaded and
design in such case shall be done as stipulated in cl. 7.5.1.2 of code. Single angle used as a leg of a
tower may be concentrically loaded and can be designed by Method-II. When two angles are used
back to back and connected on either side of gusset plate, or on same side of gusset plate, the
member may be considered concentrically loaded and Method-II can be used. Remember, for
unequal angles, the design becomes economical when longer leg is connected.

P-1: Calculate the design compressive load for a stanchion 350 @ 710.2 N/m, 3.5 m high. The
column is restrained in direction and position at both ends. Use steel grade of Fe 410.
Ans:
2
For steel grade of Fe 410, fy = 250 N/mm
The column is restrained in direction & position at both ends. Hence K = 0.65
Properties of ISHB 350 @ 710.2 N/m are
H = 350 mm, bf = 250 mm, tf = 11.6 mm, tw = 10.1 mm
2
A = 9221 mm , rz = 146.5 mm, ry = 52.2 mm
h 350
= = 1.4 > 1.2
bf 250
Tf = 11.6 mm ≤ 40 mm
Hence as per Table 10 of IS 800, buckling class will be class “a” for buckling about zz-axis and class
“b” for buckling about yy-axis.
KL 0.65 x 3.5 x 10 3
Effective slenderness ratio, λy = = = 43.58
ry 52.2

KL 0.65 x 3.5 x 10 3
λz = = = 15.53
rz 146.5
So λy governs the strength of column.

For KL/ry = 43.58, fy = 250 N/mm2 & buckling class b, design compressive stress from Table 9(b) of
code is given by
 7

206−194
Fcd = 206 ̶ x (43.58 ̶ 40) = 201.704 N/mm2
10

Design compressive strength Pd = Ae fcd = 9221 x 201.704 x 10-3 = 1859.87 kN.

P-2: Calculate the compressive resistance of the leg of a transmission tower consisting of ISA 200 x
200 x 20 of height 3 m. Assume that the conditions at both ends of zz and yy planes are such as to
provide simple support. Grade of steel is Fe410. Assume the load is concentrically applied to the
angle.

Ans:

Properties of ISA 200 x 200 x 20

A = 7640 mm2, rzz = ryy = 61.4 mm
rvv = 39.3 mm [minimum radius of gyration]
2
For steel of Fe 410, fy = 240 N/mm [note: fy is reduced as per Table 1 of code as thickness of section
= 20 mm)
Considering pin-ended supports, effective length KL = 1 x 3000 = 3000 mm
Effective slenderness ratio λ = KL/rmin = 3000/39.3 = 76.34
For single angle section, buckling curve as per Table 10 of code is class c
From Table 9(c) of the code,
for λ = 76.34 and fy = 240 N/mm2, fcd = 138.49 N/mm2
-3
Hence design strength = 138.49 x 7640 x 10 = 1058 kN.

Design of Compression Members

Design procedure is a trial and error method. The following are the steps in design of
compression members-
1. Design stress (fcd) in compression is to be assumed.
2
For rolled beam section, assume fcd = 135 N/mm
2
For angle struts, assume fcd = 90 N/mm
2
For compression members carrying heavy loads, assume fcd = 200 N/mm
2. Calculate effective sectional area required, Ag = Pu/fcd, where Ag is the tentative cross-
sectional area required and Pu is the factored load on the column.
3. Select a suitable section with the required effective sectional area and calculate rmin.
4. Calculate effective length of column for the given end conditions and calculate effective
slenderness ratio λ.
5. Calculate the design compressive stress fcd from table 9 of the code.
6. Calculate load carrying capacity.
7. Revise the section, if calculated load carrying capacity differs considerably from given
factored load.

P-3: Design a column to support a factored load of 1050 kN. The column has an effective length of
7.0 m with respect to z-axis and 5 m with respect to y-axis. Use steel grade of Fe 410.
Ans:

For steel of grade Fe 410, fy = 250 MPa

8

Let us provide an I-section.

Assuming fcd = 135 N/mm2,
3 2
Required A = 1050 x 10 / 135 = 7778 mm
Try ISHB 300 @ 63.0 kg/m
2
A = 8025 mm , h = 300 mm, bf = 250 mm, tf = 10.6 mm
rz = 127.0 mm, ry = 52.9 mm
h/bf = 300/250 = 1.2 & tf = 10.6 mm, hence as per Table 10 of code, the design buckling curve
about z-axis is b and about y-axis is c.
Buckling about z-axis
Effective length KL = 7 m = 7000 mm
Effective slenderness ratio λz = KL/rz = 7000/127 = 55.12 < 180 o.k.
2
For λz = 55.12, fy =250 N/mm & buckling curve b, as per Table 9(b) of code, the design
compressive stress is
194−181
fcd = 194 ̶ x 5.12 = 187.34 N/mm2
10
Design compressive load Pd = Ae x fcd = 8025 x 187.34 x 10-3 = 1503 kN > 1050 kN. o.K.
Buckling about y-axis
Effective length KL = 5 m = 5000 mm
Effective slenderness ratio λy = KL/ry = 5000/52.9 = 94.52 < 180 o.k.
2
For λy = 94.52, fy =250 N/mm & buckling curve c, as per Table 9(c) of code, the design
compressive stress is
121−107 2
fcd = 121 ̶ x 4.52 = 114.67 N/mm
10
Design compressive load Pd = Ae x fcd = 8025 x 114.67 x 10-3 = 920 kN < 1050 kN. Unsafe.
Hence we have to redesign.
Try ISHB 350 @ 67.4 kg/m
A = 8591 mm2, h = 300 mm, bf = 250 mm, tf = 11.6 mm
rz = 149.3 mm, ry = 53.4 mm
h/bf = 350/250 = 1.4 > 1.2
& tf = 11.6 mm < 40 mm.
Hence as per Table 10 of code, the design buckling curve about z-axis is a and about y-axis is b.
Buckling about z-axis
Effective slenderness ratio λz = KL/rz = 7000/149.3 = 46.89 < 180 o.k.
2
For λz = 46.89, fy =250 N/mm & buckling curve a, as per Table 9(a) of code, the design
compressive stress is
213−205
fcd = 213 ̶ x 6.89 = 207.49 N/mm2
10
Design compressive load Pd = Ae x fcd = 8591 x 207.49 x 10-3 = 1803 kN > 1050 kN. o.K.
Buckling about y-axis
Effective slenderness ratio λy = KL/ry = 5000/53.4 = 93.63 < 180 o.k.
2
For λy = 93.63, fy =250 N/mm & buckling curve b, as per Table 9(b) of code, the design
compressive stress is
134−118
fcd = 134 ̶ x 3.63 = 128.19 N/mm2
10
-3
Design compressive load Pd = Ae x fcd = 8591 x 128.19 x 10 = 1101 kN > 1050 kN. o.k.
Hence, provide ISHB 350 @ 67.4 kg/m.
 9

P-4: An ISA 100 x 100 x 6 (fy = 250 N/mm2) is used as a strut in a truss. The length of the strut
between the intersections at each end is 3 m. Calculate the strength of the strut if,
(a) it is connected by two bolts at each end
(b) it is connected by one bolt at each end
(c) it is welded at each end.
Ans:

Given fy = 250 N/mm2

Properties of angle section
2
A = 1167 mm , rmin = rvv = 19.5 mm
(a) Each end connected by two bolts
IS 800: 2007 is silent about effective length of single angle strut.
Let effective length = actual length = 3 m = 3000 mm
Considering the ends to be fixed,
k1 = 0.20, k2 = 0.35 and k3 =20 [from Table 12 of the code]
ε = (250/fy)0.5 = 1.0
π2E π 2 x 2 x 10 5
λvv = (l/rvv)/(ε� ) = (3000/19.5)/ � = 1.731
250 250

π2E π 2 x 2 x 10 5
λφ = [(b1 + b2)/2t]/ (ε� ) = [(100 + 100)/(2 x 6)]/ � = 0.1875
250 250

λe = ��k1 + k 2 λvv 2 + k 3 λϕ 2 � = �(0.20 + 0.35 x 1.7312 + 20 x 0.18752 ) = 1.397

Imperfection factor α = 0.49 [from Table 7 of the code]
2 2
𝜙𝜙 = 0.5 [ 1 + α (λe ̶ 0.2) + λe ] = 0.5 [1 + 0.49 x (1.397 ̶ 0.2) + 1.397 ] = 1.77
f y /γ m 0 250/1.1
Design compressive stress fcd = 0.5 = = 79.55 N/mm2
∅ + �∅2 − λ e 2 � 1.77 + (1.77 2 − 1.397 2 )0.5

Design compressive strength Pd = Ae fcd = 1167 x 79.55 x 10-3 = 92.83 kN

(b) Each end connected by one bolt
Considering the ends to be hinged,
k1 = 1.25, k2 = 0.5 and k3 =60 [from Table 12 of the code]
ε = (250/fy)0.5 = 1.0
λvv = 1.731 [as obtained in part (a)]
λφ = 0.1875 [ do ]

λe = ��k1 + k 2 λvv 2 + k 3 λϕ 2 � = �(1.25 + 0.50 x 1.7312 + 60 x 0.18752 ) = 2.203

Imperfection factor α = 0.49 [from Table 7 of the code]
2 2
𝜙𝜙 = 0.5 [ 1 + α (λe ̶ 0.2) + λe ] = 0.5 [1 + 0.49 x (2.203 ̶ 0.2) + 2.203 ] = 3.41
f y /γ m 0 250/1.1
Design compressive stress fcd = 0.5 = = 37.79 N/mm2
∅ + �∅2 − λ e 2 � 3.41 + (3.412 − 2.203 2 )0.5
-3
Design compressive strength Pd = Ae fcd = 1167 x 37.79 x 10 = 44.11 kN
(c) Each end is welded
The strength of the welded strut will be the same as that of the single strut connected with two
bolts at each end with ends considered fixed.
Hence Pd = 92.83 kN.
10

P-5: Design a double angle discontinuous strut to carry a factored load of 135 kN, resulting from
combination with wind load. The length of the strut is 3.0 m between intersections. The two angles
are placed back-to-back and tack bolted. Use steel of grade Fe 410.
(a) Angles are placed on opposite sides of 10 mm gusset plate
(b) Angles are placed on same side of 10 mm gusset plate.

Ans:

For steel grade of Fe 410, fy =250 MPa

2 2
Assuming fcd = 90 N/mm , required cross-sectional area A = 135 x 103/90 = 1500 mm
Try 2 ISA 70 x 70 x 6 @ 123.6 N/m with following properties of single angle
2 4 4
A =806 mm , Iz = Iy = 36.8 x 10 mm , rz = ry = 21.4 mm, Cx = Cy = 19.4 mm

(a) Angles are placed on opposite sides of gusset

With 2 ISA 70 x 70 x 6 @ 123.6 N/m,
Area of double angles A’ = 2 x 806 = 1612 mm2
2 4 2
Iy’ for double angles = 2 x [Iy + A x (Cy + tg/2) ] = 2 x [36.8 x 10 + 806 x (19.4 + 10/2) ]
4 4
= 169.57 x 10 mm
169.57 x 10 4
ry’ for double angle = � 1612
= 32.43 mm
rz’ for double angle = rz for single angle = 21.4 mm [as z-axis for both angles is same, resulting
into doubling of Iz & A]
So minimum radius of gyration = rz’ = rz = 21.4 mm
Effective length l = KL = 0.85 x 3000 = 2550 mm
Effective slenderness ratio λ = l/r = 2550/21.4 = 119.16 < 250 o.k.
2
For λ = 119.16, fy = 250 N/mm and buckling curve c as per Table 9(c) of code, the design
compressive stress is
94.6 − 83.7
fcd = 94.6 ̶ x 9.16 = 84.62 N/mm2
10
Design compressive load Pd = Ae x fcd = 1612 x 84.62 x 10-3 = 136.41 kN > 135 kN. o.k.
(b) Angles are placed on the same side of gusset
With 2 ISA 70 x 70 x 6 @ 123.6 N/m,
Area of double angles A’ = 2 x 806 = 1612 mm2
2 4 2
Iz’ for double angles = 2 x [Iz + A x Cz ] = 2 x [36.8 x 10 + 806 x 19.4 ]
4 4
= 134.269 x 10 mm
134.269 x 10 4
rz’ for double angle = � 1612
= 28.86 mm
 11

ry’ for double angle = ry for single angle = 21.4 mm [as y-axis for both angles is same, resulting
into doubling of Iy & A]
So minimum radius of gyration = ry’ = ry = 21.4 mm
Effective length l = KL = 1 x 3000 = 3000 mm [Note: K = 1 in this case.]
Effective slenderness ratio λ = l/r = 3000/21.4 = 140.19 < 250 o.k.
For λ = 140.19, fy = 250 N/mm2 and buckling curve c as per Table 9(c) of code, the design
compressive stress is
66.2 – 59.2 2
fcd = 66.2 ̶ x 0.19 = 66.067 N/mm
10
-3
Design compressive load Pd = Ae x fcd = 1612 x 66.067 x 10 = 106.5 kN < 135 kN.
So we have to revise the design.
Let us try with 2 ISA 80 x 80 x 6 mm. Properties of single angle
A =929 mm2, Iz = Iy = 56.0 x 104 mm4, rz = ry = 24.6 mm, Cx = Cy = 21.8 mm
Minimum radius of gyration = ry’ = ry = 24.6 mm
Effective length l = KL = 1 x 3000 = 3000 mm [Note: K = 1 in this case.]
Effective slenderness ratio λ = l/r = 3000/24.6 = 121.95 < 250 o.k.
For λ = 121.95, fy = 250 N/mm2 and buckling curve c as per Table 9(c) of code, the design
compressive stress is
83.7 – 74.3
fcd = 83.7 ̶ x 1.95 = 81.867 N/mm2
10
Design compressive load Pd = Ae x fcd = 2 x 929 x 81.867 x 10-3 = 152.11 kN > 135 kN.
Hence o.k.

Built-up Columns (Latticed column)

For economical design of heavily loaded long columns, the least radius of gyration of the column
section is required to be increased to maximum (ry ≥ rz). To achieve this condition the rolled or
fabricated sections are kept away from the centroidal axis of the column and are connected by some
connecting system known as lattice system. The commonly used lattice are
(a) Lacing
(b) Battens.

Design of Laced Column

Steel flats and angles are commonly used for lacing. One can use single lacing or double lacing
system [Fig 7]. IS 800: 2007 specifies the following rules for the design of latticed column:
1. As far as possible, the latticing system shall be uniform throughout.
2. In single laced system the direction of lattices on the opposite faces should be shadow of the
other.
3. In bolted/riveted, the minimum width of lacing bars shall be three times the nominal
diameter of bolt/rivet.
4. The thickness of flat lacing bars shall not be less than one-fortieth of its effective length for
single lacing and one-sixtieth of the effective length for double lacing.
5. Lacing bars shall be inclined at 400 to 700 to the axis of built-up column.
6. The maximum spacing of lacing bars shall be such that the maximum slenderness ratio of the
components of the main member (a1/r1), between consecutive lacing connections is not
greater than 50 or 0.7 times the most unfavourable slenderness ratio of the member as a
12

whole, whichever is less, where a1 is the unsupported length of the individual member
between lacing points, and r1 is the minimum radius of gyration of the individual member
being laced together.
7. The lacing shall be designed to resist transverse shear Vt = 2.5% of axial force in the column
and shall be divided equally among all transverse lacing systems in parallel planes.
𝐕𝐕𝐭𝐭 𝐭𝐭𝐕𝐕
8. The compressive force in the lacing bar is equal to 𝐍𝐍 𝐬𝐬𝐬𝐬𝐬𝐬 𝛉𝛉
for single lacing system and 𝟐𝟐 𝐍𝐍 𝐬𝐬𝐬𝐬𝐬𝐬 𝛉𝛉
for double lacing system, where N is the number of parallel planes of lacing systems and θ is
the inclination of lacing bar with the axis of built-up column.
9. The slenderness ratio KL/r of the lacing bars shall not exceed 145. In bolted/riveted
construction, the effective length of the lacing bar is the length between the inner end
rivets/bolts for a single lacing system and 0.7 times of this distance for a double lacing
system. In welded constructions, the effective length of the lacing bar is 0.7 times the
distance between the inner ends of the welds connecting the lacing bars to the member.
10. Laced compression members shall be provided with end tie plates.
11. The effective slenderness ratio of laced columns shall be taken as 1.05 times the actual
maximum slenderness ratio, in order to account for shear deformation effects.
12. In case of two lacing flats are connected at the same point by rivet/bolts, the load on
𝟐𝟐 𝐕𝐕𝐭𝐭
rivet/bolt will be 𝐍𝐍 𝐭𝐭𝐭𝐭𝐭𝐭 𝛉𝛉
and for lacing flats connected at different points the load on
𝐕𝐕𝐭𝐭
rivet/bolt will be . In case of welded construction, the welding of lacing bars to the
𝐍𝐍 𝐬𝐬𝐬𝐬𝐬𝐬 𝛉𝛉
main member should be sufficient to transmit the load in the bar. Welding should be
provided along each side of the lacing bar for the full length of the lap. The overlap of lacing
bar should not be less than four times the thickness of the bar or the member, whichever is
less.

Fig 7 Laced Column

 13

P-6: A laced column with two channels back to back of length 10 m is to carry an axial factored load
of 1080 kN. The column is restrained in position but not in direction at both the ends. Grade of steel
is Fe 410. Design the column if
(a) the lacing system is with bolted connection (grade of bolt is 4.6)
(b) the lacing system is with site welded connection.
Ans:
For steel of grade Fe 410, fu = 400MPa, fy = 250 MPa
For bolts of grade 4.6, fub = 400 MPa
Partial safety factors for material: γm0 = 1.10, γmb = 1.25, γmw = 1.5 [for site weld]
Design of column:
Assuming fcd = 135 N/mm2,
1080 x 10 3 2
Required cross-sectional area = 135
= 8000 mm
Let us try 2 ISMC 300 @ 351.2 N/m. Properties of individual members are
A = 4564 mm2, rz = 118.1 mm, ry = 26.1 mm, Cy = 23.6 mm, bf = 90 mm,
4 4 4 4
Iz = 6362.6 x 10 mm , Iy = 310.8 x 10 mm , g = 50 mm & tf = 13.6 mm
2
Area provided = 2 x 4564 = 9128 mm
Distance between channels will be kept so as to get ry > rz
So, minimum radius of gyration, r = rz = 118.1
KL 1 x 10000
Slenderness ratio = = = 84.674 [K = 1, as column is hinged at both ends]
r 118.1
Since it is a laced column, effective slenderness ratio, λe = 1.05 x 84.674 = 88.90
For λe = 88.9, fy = 250 MPa and bulking curve c, from Table 9(c) of the code,
2
fcd = 122.65 N/mm
-3
Design compressive strength, Pd = 9128 x 122.65 x 10 = 1119.5 kN > 1080 kN. O.K.
Spacing between the channels, placed back to back
Let clear distance = S
Now, Iz = 2 x 6362.6 x 104 mm4
2 4 2
Iy = 2 x [Iy + A x (S/2 + Cy) ] = 2 x [310.8 x 10 + 4564 x (S/2 + 23.6) ]
Equating Iy to Iz, we get
4 2 4
2 x [310.8 x 10 + 4564 x (S/2 + 23.6) ] = 2 x 6362.6 x 10
or, S = 183.10 mm
Provide S = 183.5 mm
(a) Design of lacing system with bolted connection
Let us provide single lacing system at 450 with vertical.
Spacing of lacings, a1 = 2 x (183.50 + 50 + 50) x cot 450 = 567 mm
[Note gauge distance = 50 mm]
a1 567
ry
= = 21.724 < 50
26.1

Also a1/ry should be < 0.7 x 88.9 = 62.23. So, it is o.k.

2.5
Transverse shear to be resisted by lacing systems, Vt = x 1080 x 103 = 27,000 N
100
Vt
Compressive force in lacing bar = = 27000/( 2 sin 450 ) x 10-3 = 19.09 kN
N sin θ
Let us provide 16 mm diameter bolt of grade 4.6 for connection of lacings.
Minimum width of lacing flats = 3 x 16 = 48 mm [cl. 7.6.2 of the code]
14

Let us provide 50 mm flats.

1
Minimum thickness of lacing flats = x length of flat between inner end bolts
40
1
= x (183.5 + 50 + 50) cosec 450
40
= 10.02 mm
Provide 12 mm thick flat for lacings.
t 12
Minimum radius of gyration, r = = = 3.464 mm
√12 √12
283.5 x cosec 45 0
l/r of lacing bar = 3.464
= 115.72 < 145 o.k.
For l/r = 115.72, fy = 250 MPa and buckling curve c, from Table 9(c) of the code,
2
fcd = 88.365 N/mm
-3
Design compressive stress of lacing, Pd = (50 x 12) x 88.365 x 10 = 53.02 kN > 19.09 kN
Tensile strength of lacing bar is minimum of
-3
(i) 0.9 x (B ̶ dh) t fu / γm1 = 0.9 x (50 ̶ 18) x 12 x 410 x 10 / 1.25 = 113.35 kN
-3
(ii) Ag fy / γm0 = 50 x 12 x 250 x 10 / 1.1 = 136.363 kN
Hence tensile strength = 113.35 kN > 19.09 kN
So chosen section of flat is safe.

Design of connection with bolts

Assuming two lacing flats will be connected at different points as shown in Fig 8 (a), using one 16
mm diameter bolt for each lacing flat, the bolts will be in single shear.
Strength of bolt in single shear, considering shear plane in threaded area
f ub 𝜋𝜋 400
= nn Anb = 1 x (0.78 x x 162) x = 28,974 N = 28.974 kN
√3 γ mb 4 √3 x 1.25
Strength of bolt in bearing = 2.5 kb d t fu / γmb = 2.5 x 1 x 16 x 12 x 410 / 1.25 [assuming kb = 1]
= 157,440 N = 157.44 kN
So strength of 16 mm dia bolt = 28.974 kN
Vt 27
Force on bolt = = = 19.09 kN
N sin θ 2 𝑥𝑥 si n 45 0
19.09
No of bolts required = = 0.658
28.974
Provide 1 bolt, 16 mm diameter of grade 4.6 to connect each end of the flat with channel.

Tie Plates
Tie plates are provided at the ends of the laced column.
Effective depth = 183.5 + 2 Cyy = 183.5 + 2 x 23.6 = 230.7 mm > 2 bf (= 2 x 90 = 180) O.K.
Overall depth of tie plate = 230.7 + 2 x 30 = 290.7 mm ≈ 300 mm
[as minimum edge distance for 16 dia bolt = 1.5 x 18 = 27 mm. Provide e = 30 mm]
Length of tie plate = 183.5 + 2 bf = 183.5 + 2 x 90 = 363.5 mm ≈ 364 mm
1 1
Thickness of tie plate = x (183.5 + 2g) = x (183.5 + 2 x 50) = 5.67 mm [g = gauge distance]
50 50
So provide tie plates of 364 x 300 x 6 mm at both ends with 6 nos 16 mm diameter bolts.

(a) Design of lacing system with welded connection

Spacing of channels = 183.5 mm
Compressive force in lacing flats = 19.09 kN
Effective length of welded lacing bar = 0.7 x 183.5 x cosec 450 = 181.65 mm
 15

1
Minimum thickness of lacing flat = x 183.5 x cosec 450 = 6.48 mm ≈ 8 mm
40
Let us try flat section 50 ISF 8 mm.
t 8
Minimum radius of gyration, r = 12
= = 2.31 mm
√ √12
l/r = 181.65/2.31 = 78.63 < 145 O.K.
For l/r = 78.63, fy = 250 MPa and buckling curve c, from Table 9(c) of the code,
2
fcd = 138.19 N/mm
-3
Design compressive stress of lacing, Pd = (50 x 8) x 138.19 x 10 = 55.27 kN > 19.09 kN O.K.
Overlap of lacing flat = 50 mm > 4 x 8 (=32)
Connection
Flange thickness = 13.6 mm
Minimum size of fillet weld for 13.6 mm thick member = 5mm
Strength of weld per unit length = 0.7 x 5 x 1 x 410/(√3 x 1.5) = 552 N/mm
3
Length of weld required = (19.09 x 10 )/552 = 34.58 mm
Adjust the weld length as shown.
Weld length provided = 2 x 50 + 2 x50 = 200mm
Tie plate
Overall depth = effective depth = 183.5 + 2 x 23.6 = 230.7 mm > 2 x 90 mm O.K.
Provide depth of tie plate = 231 mm
Length of tie plate = 183.5 + 2 x 50 = 283.5 mm
1
Thickness of tie plate = x 283.5 = 5.67 mm ≈ 8 mm [to accommodate 5 mm weld)
50
Provide a tie plate 283.5 x 231 x 8 mm and connect it with 5 mm weld as shown in Fig 8(b).

(a) (b)
Fig 8 Problem P-6
16

Column Bases
Column bases transmit the column load to the concrete or masonry foundation blocks. The
column base spreads the load on wider area so that the intensity of bearing pressure on the
foundation block is within the bearing strength. There are two types of column bases commonly
used in practice.
1. Slab Base
2. Gusseted Base.
Slab Base: These are used in columns carrying small loads. In this type, the column is directly
connected to the base plate through cleat angle as shown in Fig 9. The load is transferred to the base
plate through bearing.
Gusseted Base: For column carrying heavy loads gusseted bases are used. In gusseted base, the
column is connected to the base plate through gussets. The load is transferred to the base partly
through bearing and partly through gussets. Fig 10 shows a typical gusseted base connection.

Fig 10 Gusseted Base

Fig 9 Slab Base

Design of Slab Base

The design of slab base consists of finding the size and thickness of base plate. The following are
the design steps to be followed:
1. Based on grade of concrete, find the bearing strength of concrete, which is = 0.45 fck
Pu
2. Area of base plate required = 0.45 f ck
, where Pu is the factored load.
3. Select the size of base plate. For economy, as far as possible keep projections a and b
equal [see Fig 9]
4. Calculate intensity of pressure w below base plate using
𝑃𝑃𝑢𝑢
w = 𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴 𝑜𝑜𝑜𝑜 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝
5. Minimum thickness of base plate required is given by cl. 7.4.3.1 of the code as

ts = ��2.5 w (a2 − 0.3 b 2 )γm0 /fy � > tf

where, ts = thickness of base plate

tf = thickness of flange
 17

Connections:
1. Connect base plate to foundation concrete using two or four 20 mm diameter and 300 mm
long holding-down bolts.
2. If bolted connection is to be used for connecting column to base plate, use 2 ISA
65x65x6mm with 20 mm bolts.
3. If weld is to be used for connecting column to base plate, check the weld length of fillet
welds.

P-7: Design a slab base for a column ISHB 350 @ 710.2 N/m subjected to a factored load of 1500 kN,
if the load is transferred to the base plate by direct bearing of column flanges. Take fu = 410 MPa and
fy = 250 MPa.
Ans:
2
Assuming grade of concrete as M20, bearing strength of concrete = 0.45 fck = 0.45 x 20 = 9.0 N/mm
Factored load Pu = 1500 kN
Properties of ISHB 350 @ 710.2 N/m:
tf = 11.16 mm
tw = 10.1 mm
bf = 250 mm
D = 350 mm
1500 x 10 3
∴ Area of base plate required, A = = 166,666.6 mm2
9.0
Let us provide rectangular base plate with projections beyond column edges as a & b. Let a = b
Area of base plate = (D + 2 a) x (bf + 2 b) = (350 + 2 a) x (250 + 2 a)
Then, (350 + 2a) x (250 + 2a) = 166,666.6
or, a = 55.65 mm
Let us take a = 60 mm
Length of base plate, L = 350 + 2 x 60 = 470 mm
Width of base plate B = 250 + 2 x 60 = 370 mm
1500 x 10 3
Bearing pressure = 470 x 370
= 8.62 N/mm2 < 9.00 N/mm2 O.K.

Thickness of base plate ts = ��2.5 w (a2 − 0.3 b 2 )γm0 /fy �

= �(2.5 x 8.62 x (602 − 0.3 x 602 )x 1.10 /250)

= 15.458 ≈ 16 mm > 11.6 mm (tf)
Hence provide a base plate 470 x 370 x 16 mm.
Since transfer of load to base plate is by direct
bearing, connection of column with base plate is not
essential. However, to keep the column in position,
two cleat angles with two holding down bolts may be
provided as shown.
Note: When transfer of load to the base plate is by direct bearing, the column end and the base
plate have to be machined for perfect bearing. In this case, connection of column with base
plate need not be designed and provide two cleat angles with holding-down bolts to keep
the column in position. But, when column end and base plate are not machined for bearing,
the connection is to be provided using fillet weld as shown in the next problem.
18

P-8 : Redo the problem P-7, if the column end and base plate are not machined for bearing.
Ans:
As obtained in P-6, size of base plate is 470 x 370 x 16 mm.
We shall provide welded connection for transfer of load from column to base plate.
Length available for welding around the column profile
La = 2 x 250 + 2 x (250 ̶ 10.1) + 2 x (350 ̶ 2 x 11.6) = 1633.4 mm
Let us provide 8 mm fillet weld. Since welding is not possible at toes and fillets of column
section, end returns (2S) will have to be subtracted at the end of each fillet weld to get the effective
length of weld.
Number of total end returns = 12
Effective length of weld = 1633.4 ̶ 12 x (2 x 8) = 1441.4 mm
Throat thickness, tf = 0.7 S = 0.7 x 8 = 5.6 mm
fu 410
Strength of weld per mm length = 1 x tf x =1 x 5.6 x = 1060.47 N/mm
√3 γ mw √3 x 1.25
1500 x 10 3
Required length of weld = 1060 .46
= 1414.46 mm < 1441.4 mm O.K.
So provide 8 mm fillet weld all round the column section. Also provide two numbers 20 mm
diameter holding-down bolts to keep the base in position.
Design of Gusseted base
The design steps are as follows:
Factored Load
1. Area of base plate =
0.45 f cd
2. Assume various thickness of gusset base
(a) Thickness of gusset plate is assumed as 16 mm.
(b) Size of gusset angle is assumed such that its vertical leg can accommodate two bolts in
one vertical line. Other leg is assumed in which one line of bolts can be provided.
(c) Thickness of angle is kept approximately equal to thickness of gusset plate.
3. Width of gusset base is kept such that it will just project outside the gusset angle and then
Area of plate
length =
width
4. When end of column is machined for complete bearing on base plate, 50% of load is
assumed to be transferred by bearing and 50% by fastenings.
When column end and base plate are not machined for bearing, fastenings shall take all the
loads.
5. The thickness of base plate is computed by flexural strength at the critical sections.

P-9: A column ISHB 350 @ 661.2 N/m carries an axial compressive factored load of 1700 kN. Design
a suitable bolted gusset base. The base rests on M15 grade concrete pedestal. Use 24 mm diameter
bolts of grade 4.6 for making the connection.
Ans:
For Fe 410 grade of steel, fu = 410 MPa, fy = 250 MPa
2
For M15 grade of concrete, bearing strength = 0.45 fck = 0.45 x 15 = 6.75 N/mm
Partial safety factor, γm0 = 1.1
Properties of ISHB 350 @ 661.2 N/m:
Depth of section, D = 350 mm
Width of flange, bf = 250 mm
 19

Thickness of flange, tf = 11.6 mm

Thickness of web, tw = 8.3 mm
For bolts of grade 4.6, fub = 400 MPa γmb = 1.25
For 24 mm diameter bolts,
π
Anb = 0.78 x x 242 = 353 mm2
4
do = 24 + 2 = 26 mm
e = 1.5 do = 1.5 x 26 = 39 mm
p = 2.5 d = 2.5 x 24 = 60 mm
1700 x 10 3
Required area of base plate = = 251.85 x 103 mm2
6.75
Assume two 16 mm thick gusset plates, one on each flange of column section. Also assume two
gusset angles ISA 150 x 115 x 15 mm as shown in the Fig. 11.

Fig 11 Problem 9

Minimum width required = 350 + 2 x 16 + 2 x 115 = 612 mm ≈ 620 mm

Projection of angle beyond flange angle toe = (620 ̶ 612)/2 = 4 mm
251.85 x 10 3
Length of base plate = = 406.20 mm ≈ 410 mm
620
So provide base plate of 620 x 410 mm.
1700 x 10 3
Bearing pressure of concrete, w = 620 x 410
= 6.68 N/mm2 < 6.75 N/mm2. O.K.
Critical section of base plate for bending moment is x-x section (see Fig 11).
Projection of base plate beyond x-x section, C1 = 4 + 115 ̶ 15 = 104 mm
w C12 6.68 x 104 2
Moment at this section, Mx = 2
= 2
= 36125.44 Nmm
Moment capacity of base plate & angle leg combined,
fy Ze 250 t2
Md = 1.2 γm 0
= 1.2 x x (1 x ) = 45.45 t2
1.1 6
Equating Mx & Md, 45.45 t2 = 36125.44
∴ t = 28.19 mm
Then thickness of base plate, tb = 28.19 ̶ 15 = 13.19 mm ≈ 16 mm > 11.6 mm (tf) O.K.
Provide a base plate 620 x 410 x 16 mm.
Bolted Connection:
Bolts will be in single shear and bearing.
Strength of bolt in single shear
f ub 400 -3
Vdsb = ( nn Anb + ns Asb ) = ( 1 x 353 + 0 ) x 10 = 65.21 kN
√3 γ mb √3 x 1.25
20

Strength of bolt in bearing

fu
Vdpb = 2.5 kb d t γ mb
kb is least of
e 39
a) = = 0.50
3 d0 3 x 26
p 60
b) − 0.25 = − 0.25 = 0.52
3 d0 3 x 26
f ub 400
c) = = 0.9756
fu 410
d) 1.0
∴ kb = 0.50
Design strength of a single bolt in bearing
fu 410
Vdpb = 2.5 kb d t = 2.5 x 0.50 x 24 x 16 x x 10-3 = 157.44 kN
γ mb 1.25
Hence strength of bolt = 65.21 kN.
Assuming column end and gusset plate have complete bearing, bolt connections is designed for
50% load.
0.5 x 1700
So number of bolts required = = 13.03 ≈ 16
65.21
Provide 8 nos 24 mm dia bolts on each flange in two rows as shown in Fig 11. The number of bolts
required to connect the gusset angle with gusset plate will be same.
Height of gusset plate = 150 + 2 x 39 + 1 x 60 = 288 mm
Length of gusset plate = length of base plate = 410 mm
Provide gusset plate 410 x 288 x 16 mm as shown in Fig 11.
Check for buckling of compression edge of gusset plate
250 250
ε=� = � =1
f yg 250

Gusset outstand from column, S0 ≯ 13.66 ε tg i.e. 13.6 x 1 x 16 = 217.6 mm

Now, gusset outstand = (410 ̶ 250)/2 = 80 mm
Slat edge outstand, S0 = �802 + (60 + 2 x 39)2 = 159.51 mm ≈ 160 mm < 217.6 mm
Hence it is O.K.

*****