Design of Compression

Members
Chapter 03 C

5/9/16

Example 3.1
Design a truss compression member
using the following three given crosssectional shapes:
1. 2Ls with 10 mm thick gusset plate and
snug tight bolted stay plates.
2. Single angle connected through longer
leg.
3. W-section.
The other data is as under:
PD = 110 kN ; PL = 140 kN
5/9/16
2
L = 3.4 m ; A36 Steel

Solution:
For truss members, K=1
Lu=L as there is no bracing between
the joints
Pu= 1.2PD +1.6PL
Pu= 1.2 x 110 +1.6 x 140 = 356 kN
1- Double Angles

5/9/16

5/9/16

A. The section will be economical if it has nearly

same slenderness ratio in both directions. This

means that rx and ry should be nearly same
(as the length for both directions is equal).
B. The radius of gyration, rx, for the double
section will be same as that for a single
section.
C. The radius of gyration, ry, for the double
section will be considerably higher than for a
single section.
D. From the above facts it is clear that, to make r
same in both directions, unequal leg angles
are to be used with their longer legs
connected together.
5/9/16
5

E. The projecting leg should be relatively

smaller.
Pu = 356 kN

KLu = 3.4 m

Rass = 90
(cFcr)ass = 146.46 MPa
Areq
6

Pu
356 1000
2
for 2 Ls

2431 mm
c Fcr ass 146.46
5/9/16

Areq for single Angle = 1216 mm2

Asel Areq
Select section Minimum weight section
b bmin
bmin = 3.25d +26
= 75 mm (assuming d = 15 mm)
L/40= 85 mm (not fully applicable here)
7

5/9/16

Different options available are:

1. 2L 89 x 89 x 7.9 A = 2700 mm2
2. 2L 102 x 76 x 7.9 A = 2700 mm2
3. 2L 102 x 76 x 9.5 A = 3200 mm2
Note: Anyone out of the first two options
may
be
tried
first.
However,
calculations show that these sections
are not sufficient with only the third
section closer to the requirements. The
number of trials may be reduced by
using the column selection tables. 5/9/16
8

Using the double angles selection

tables of Design Aids for Pu= 356 kN,
KxLx = KyLy = 3.4 m and stay plates
with fully tensioned bolts, following
options are
available:
Section
P
KL (m)
Area
c

2Ls 102 x 76
x 9.5

(mm2)

(kN)
452.8

KxLx = 3.0

382.7

KxLx = 3.5

396.7
448.8

KxLx = 3.4 by
interpolation
KyLy = 3.0

377.8

KyLy = 3.5

392.0

K L = 3.4 by

3206

5/9/16

Using the double angles selection

tables of Design Aids for Pu= 356 kN,
KxLx = KyLy = 3.4 m and stay plates
with fully tensioned bolts, following
options are
available:
Section
P
KL (m)
Area
c

2Ls 102 x 89
x 9.5

10

(mm2)

(kN)
482.9

KxLx = 3.0

407.0

KxLx = 3.5

422.2
555.5

KxLx = 3.4 by
interpolation
KyLy = 3.0

492.2

KyLy = 3.5

504.9

K L = 3.4 by

3445

5/9/16

11

5/9/16

12

5/9/16

Hence, keeping in view the above values

for fully tight stay plates, trial section for
snug tight stay plates may be 2Ls 102 x 76
x 9.5. However, considering snug tight
stay plates, 2Ls 102 x 89 x 9.5 may
ultimately be the final result.
Trial Section1: 2L 102 x 76 x 9.5 A= 2 x
1600 mm2
Long legs back-to-back, connected to the
gusset plate.

13

b = 102 mm > bmin OK

Longer leg is connected.

5/9/16

b 102

10.7 r
t 9.50

OK r 12.7

rx = 32.0 mm as for single angle

Using Design Aids:
b
= 2 x 76 + 10 = 162 mm
(ry)appr = b(0.19 +s/1270)
= 162 x (0.19 + 10/1270) = 32.06
mm

14

5/9/16

K x Lux 1 3400

107
rx
32.0
K y Luy
ry

1 3400

106.05
32.06

Although rx is lesser than ry, the column

may buckle about y-axis if modified
slenderness ratio in y-direction exceeds
KxLux/rx.

15

5/9/16

Assume that snug tight bolts connect

stay plates and gusset plates.
Using 75 mm long stay plates and
treating a as the clear distance
between the stay plates and/or gusset
plates, the c/c spacing of stay plates
will be a+75 mm.

16

5/9/16

Selection of Stay Plates Spacing

a/ri = 0.75 x larger of KxLux/rx & KyLuy/ry
= 0.75 x (107) = 80.25
ri
= rz = 16.4 mm
amax = 80.25 x 16.4/1000
= 1.316 m
For 75 mm long stay plates
Number of spaces between stay plates =
3400/(1316+75)
= 2.45 (say 3 spaces)
17

5/9/16

Number of stay plates = 2

a = (3400/3)-75 = 1058 mm
Note: Number of stay plates may not
generally exceed three.
a/ri = 1058/16.4 = 64.53
For snug tight plates:
K y Luy

18

ry

106.052 64.532 125

m

5/9/16

R = larger of KxLux/rx & modified KyLuy/ry

= 125
Maximum Slenderness Ratio Check
R 200
OK
Capacity Check
From table of Design Aids: cFcr=98.28
MPa
19

5/9/16

c
Fcr
Asel
=98.28
1600/1000=314.50 kN
< Pu=356 kN

NG

Trial Section: 2L 102 x 89 x 9.5 A= 2 x

1720 mm2
Long legs back-to-back, connected to the
gusset plate.
b = 102 mm > bmin
Longer leg is connected.
20

OK
5/9/16

b 102

10.7 r
t 9.50

OK r 12.7

rx = 31.8 mm as for single angle

Using Design Aids:
b
= 2 x 89 + 10 = 188 mm
(ry)appr = b(0.19 +s/1270)
= 188 x (0.19 + 10/1270) = 37.2
mm
Exact calculation for ry (to be made only
for the last trial):
21

5/9/16

Iy

2 123 10

y exact

1720 (24.3 5) 541.3 10 mm

2

541.3 10 4

39.7 mm
2 1720

K x Lux 1 3400

107
rx
31.8
K y Luy
ry
22

1 3400

85.64
39.7
5/9/16

Using 75 mm long stay plates and

treating a as the clear distance
between the stay plates and/or gusset
plates, the c/c spacing of stay plates
will be a+75 mm.

23

5/9/16

Selection of Stay Plates Spacing

a/ri = 0.75 x larger of KxLux/rx & KyLuy/ry
= 0.75 x (107) = 80.25
ri
= rz = 18.5 mm
amax = 80.25 x 18.5/1000
= 1.485 m
For 75 mm long stay plates
Number of spaces between stay plates =
3400/(1485+75)
= 2.18 (say 3 spaces)
24

5/9/16

Number of stay plates = 2

a = (3400/3)-75 = 1058 mm
a/ri = 1058/18.5 = 57.19
For snug tight plates:
K y Luy

25

ry

85.64 57.19 103

m
5/9/16

For fully tightened bolts or welds,

appropriate formula for the modified
slenderness ratio must be used.
R = larger of KxLux/rx & modified KyLuy/ry =
107
Maximum Slenderness Ratio Check
R 200 OK
Capacity Check
From table of Design Aids: cFcr=122.63 MPa
26

5/9/16

c
Fcr
Asel
=122.63
1720/1000=421.63 kN
> Pu=356 kN

OK

Final Selection: 2Ls 102 x 89 x 9.5

27

5/9/16

2- Single Angle Connected Through Longer

Leg
It is almost impossible to load such members
concentrically due to the fact that only one
leg is generally connected with the gusset
plate.
Theoretically, concentric loading could only be
applied by milling the ends of an angle and
loading it through bearing plates (Figure
3.20).
In practice, the actual eccentricity of loading
is relatively large and neglecting it in design
may lead to a dangerously under-designed
5/9/16
28 member.

 Further single angle can undergo lateral

29

torsional buckling.
Separate AISC provisions are applicable
for such design.
The procedure used in this example is
applicable
for
single
angles
with
eccentricity expected in a planar truss.
Equal leg angles are generally preferred
for single angle columns in case the two
effective lengths KxLx and KyLy are equal.
However, unequal leg angles may also be
used if the resulting section comes out to
be economical.
5/9/16

 The selection can also be made using

the single angle

selection tables.

30

section

column

5/9/16

Pu = 356 kN
Rass = 125
(cFcr)ass = 98.28 MPa

Areq
31

356000

3623 mm 2
98.28
5/9/16

Asel Areq
Select section Minimum weight section
b bmin = 75 mm
Equal legs will be preferred as KxLx and
KyLy are equal.
Trial Section:
L 152 x 152 x 12.70, rz = 30 mm, ry =
47.2 mm
A = 3710 mm2
b = 152 > bmin
Ok
32

5/9/16

b 152

11.98
t 12.70

r 12.7

Ok

KL 3400

114
rz
30
L 3400

72.03 80
ry 47.2
KL
L
72 0.75 72 0.75 72.03 126
r
ry
33

R 126

200

Ok

5/9/16

c Fcr 96.98 MPa

c Fcr Asel 96.98 3710 1000 359.80 kN 356 kN Ok

Loading cycles are lesser than 20,000.

Final Selection: L 152 x 152 x 12.7

34

5/9/16

3 W-Section
Pu = 356 kN
Rass = 90
(cFcr)ass = 146.46 MPa

Areq
35

356000

2431 mm 2
146.46
5/9/16

Asel Areq
Select
section
Minimum
section
b bmin = 75 mm
d 360mm

36

weight

Using the W-shape selection table of

Design Aids for Pu = 356 kN, KyLy = 3.4
m, following option is obtained, giving
cPn 412 kN:
5/9/16

Trial Section: W 150 x 22.5

(KyLy)eq = 3.4/1.75 = 1.94 m (not critical)
rx = 65; ry = 37.1 mm; A = 2860 mm2
bf

152

76 mm
2
2

bmin

Ok

11.5

r 15.8

Ok

h
21.6
tw

r 42.1

Ok

37

bf
2t f

5/9/16

KL 3400
R

92
rmin 37.1

200

Ok

c Fcr 143.66 MPa

c Fcr Asel 143.66 2860 1000 411 kN 356 kN Ok

Final Section: W 150 x 22.5

38

5/9/16

Example 3.2: Members under Stress

Reversal
Design a bridge truss compression
member (K=1) that is 3.5m long.
PD= 80 kN (C),
PL= 270 kN (C),
TL= 270 kN (T).
Select following two types of sections:
i) 2Ls with 15 mm thick gusset plate,
welded connections but snug bolted
stay plates.
5/9/16
39 ii) W-section, welded connections.

Solution:
Pu= 1.2PD + 1.6PL = 1.2 x 80 + 1.6 x
270 = 528 kN
Tu=-1.2PD + 1.6TL =-1.2 x 80 + 1.6 x
270 = 336 kN
Pu > T u
Design as compression
member for Pu and then check as
tension member for Tu.

40

5/9/16

Complete the Example

Yourself

41

5/9/16

EQUAL STRENGTH COLUMN

Equal strength column is defined as a
column having same critical buckling
strength along both strong and weak
axes.
In other words, for an equal strength
column, KL/r ratio along both the axes
should approximately be the same.
The material of the column is used
more efficiently if this criterion is
satisfied which results in the more
economical column design.
5/9/16
42

K x Lx K y L y

rx
ry
rx K x Lx

ry K y Ly
For an equal strength column, ratio of
radii of gyration in the two directions
should be equal to the ratio of effective
length in the two directions.
Bracing is provided in such a way that the
above equation is approximately satisfied.
The ratio rx/ry is the main criterion for
design of bracing.
43

5/9/16

Knowing KxLx and rx/ry ratio, the required

effective un-braced length along y-axis
may be determined as follows:
K x Lx
K y Ly
rx ry

Based on the above information, the

LRFD Specifications provides a method
with which a W-section can be selected
from the column selection tables with
little trial and error when the un-braced
lengths in the two directions are
known/decided earlier.
44

5/9/16

The designer enters the appropriate

table with KyLy, selects a section, takes
the rx/ry value given at the bottom of
the table for that shape, and calculates
(KyLy)eq as follows: K x Lx

K L
y

45

y eq

rx ry

If (KyLy)eq is greater than KyLy, buckling

is expected about x-axis and the
designer will re-enter the table with
(KyLy)eq and will select the final
selection.
5/9/16

Example 3.3
A 9 m long column in a frame structure
is
to
resist
an
axial
factored
compressive load of 1600 kN. Assume
the bottom end of the column to be
pinned and the top end to be fixed in
both directions. The column is braced
at its third points along weak axis.
Assume that there is no sidesway
between the top and bottom points.
Design a wide flange section using A36
steel and AISC Specifications.
46

5/9/16

Solution:
Buckling of the column with
braces results in rotation of
the main member at the
brace points and lateral
bracing does not prevent this
rotation.
Hence, brace point may be
considered just like a hinge
for that direction (Figure
3.22).
However, lateral translation
or sway of the brace point is
almost zero.
47

Figure 3.22 Buckling of

Column along weak axis
5/9/16

L/3

(a) Top Half Column Buckles in Weak Direction

48

5/9/16

L/3

(a) Top Half Column Buckles in Weak Direction

L/3

a) Middle & Bottom Column Buckling in Weak Direction

49

5/9/16

Lu= 9 m

(c) Column Buckling in Major Axis Direction

50

Effective Lengths for Buckling of Column

5/9/16

Pu = 1600 kN
Rass = 90
cFcr = 146.46 Mpa

Areq

Pu
1600 1000

10,925 mm 2
c Fcr
146.46

Ky = 0.8 for top portion (Figure 3.23 (a))

Ky = 1.0 for bottom portion (Figure 3.23 (b))
Critical
51

5/9/16

Further,
AISC
Specifications
recommends K = 1.0 for braced
columns.
Kx = 0.8 (figure 3.23c)
Use Kx = 1.0
KyLuy = 1.0 x 9/3 = 3.0 m
KxLx = 1.0 x 9
= 9.0 m
K x Lx 9
3
K y Ly 3
Hence rx/ry of the selected section
closer to 3.0 will be preferred.
52

5/9/16

Following options are available:

1. W 360x72 A = 9,100 mm2 rx/ry = 3.06
2. W 360x91 A = 11,500 mm2 rx/ry = 2.44
3. W 310x86 A = 11,000 mm2 rx/ry = 2.10
4. W 310x97 A = 12,323 mm2 rx/ry = 1.75
5. W 200x86 A = 11,000 mm2 rx/ry = 1.74
The first option has area less than the
assumed required value but rx/ry is close
to 3.0.
This may be tried as first option. The
other options, although have greater
5/9/16
53 area, may fail.

Trial Section: W 360x72

A = 9,100 mm2
rx= 149 mm ry= 48.5
mm

bf

6.7

r 15.8

Ok

h
33.5
tw

r 42.1

Ok

2t f

K y luy
ry
54

3000

62
48.5

K x lux 9000

61
rx
149

Critical

5/9/16

Maximum Slenderness Check:

R = 62 < 200
Ok
Using Design Aids:
MPa

(cFcr)ass =183.52

c Pn 183.52 9100 1000 1670 kN 1600 kN Ok

Final Selection: W 360 x 72

55

5/9/16

Note: In this example, area of qualifying example

section is lesser than the assumed required
area. This is only because of selecting the
equal strength column. Further, by using the
column selection tables, the criterion of equal
strength is automatically satisfied.
Example 3.4:
Design a W-section column in a building frame
of 4m story height for an axial factored
compressive load of 975 kN. The frame is
braced against sidesway along its weak axis
but sidesway is allowed along the major axis
(K=2). Assume the ends of the column to be
pinned at the base and fixed at the top.
56

5/9/16

Solution:
Pu = 975 kN
Rass = 90 :

Areq

57

cFcr = 146.46 MPa

Pu
975 1000

6657 mm 2
c Fcr
146.46

5/9/16

Kx = 2.0
Ky = 1.0
Lux = Luy = 4m : KxLux = 8.0 m :
KyLuy = 4.0m
K x lux
rx
2 .0
close to 2.0 will be preferred
K y luy
ry
For Pu= 975 kN and KyLy = 4.0m, the
selected section using the column
selection tables is W 250 x 49.1.
58

5/9/16

rx

ry

2.16

K l

y y eq

K xlx

8.0 2.16 3.7 m

rx ry

Hence (KyLy)eq
respect to KyLy.

is

not

critical

with

Trial Section: W250x49.1

59

5/9/16

For Pu = 975 kN and KyLy= 4.0m:

Trial Section: W250x49.1
A = 6260 mm2 ; rx/ry = 2.15 ; d = 247
mm
bf = 202 mm ; tf = 11.0 mm ; tw =
7.4 mm
rx = 106 mm ; ry = 49.3 mm

bf

9.1

r 15.8

Ok

h
27.1
tw

r 42.1

Ok

60

2t f

5/9/16

K x Lux
76
rx
K y Luy
ry

82

Critical

Maximum Slenderness Check:

R= 82 < 200
Ok
(cFcr) =157.54 MPa
61

5/9/16

c Pn 157.54 6260 1000 986 kN 975 kN Ok

Final Selection: W 250 x 49.1

62

5/9/16