STRUCTURE magazine August 2008

d
i
s
c
u
s
s
i
o
n
s

o
n

d
e
s
i
g
n

i
s
s
u
e
s

f
o
r

s
t
r
u
c
t
u
r
a
l

e
n
g
i
n
e
e
r
s
S
t
r
u
c
t
u
r
a
l

D
e
S
i
g
n
August 2008 STRUCTURE magazine 34
Optimum Beam-to-Column Stiffness Ratio of
Portal Frames under Lateral Loads
By Pedro Silva, Ph.D., P.E. and Sameh S. Badie, Ph.D., P.E.
An article in the March 2008 issue of
STRUCTURE

magazine discussed de-

sign optimization of portal frames, such
as those shown in Figure 1, under grav-
ity loads. This follow-up article discusses
similar design optimization under the
infuence of lateral loads. It presents,
both analytically and graphically, the
procedure that is used to establish the
design optimization. Since the drift ratio,
defned as the ratio of the frame lateral
defection to column height, , is one of
the controlling factors in the design of
portal frames subjected to lateral loads,
the design optimization is performed as a
function thereof. Design optimization is
achieved by minimizing the sum of nor-
malized column and beam moments of
inertia necessary to satisfy a prescribed
drift ratio limit. The authors performed
their analyses using fnite element subrou-
tines implemented in Matlab

. Following
the results, the authors present simple
graphical procedures that can be used to
illustrate the optimum beam-to-column
stiffness (I
b
/I
c
) ratios in the design of por-
tal frames under lateral loads.
Effect on the Frame
Behavior
Under lateral loads, frame design is highly
dependent on the beam-to-column stiff-
ness ratio, beam-span-to-column-height
ratio, and the column end supports.
Fixed Support Conditions
If axial deformations are ignored, us-
ing slope-defection methods it can be
demonstrated that for a fat frame with
fxed supports, as shown in Figure 1(a),
the lateral defection,
f
, is:

Equation (1)
In Equation (1), is the beam-to-
column moment of inertia ratio, given by
=I
b
/I
c
, and is the drift ratio designated
as the ratio of lateral defection to column
height. This equation also includes an
arbitrary beam-span-to-column-height
ratio, designated herein as =L
b
/L
c
.
The authors used this equation to de-
velop Table 1, depicting the frame
defection,
f
, under three beam-to-
column ratios, =I
b
/I
c
, and for a beam-
span-to-column-height ratio of =1.
For =0 (corresponding to infnitely
stiff columns), the frame defection,
f
,
is four times higher than the defection
for = (corresponding to infnitely stiff
beams). This is the case because for =0
the frame response reverts to that of two
cantilever columns, and for = it re-
verts to that of two propped cantilever
columns. The results for =0.74 are also
shown in Table 1.
In order to meet the drift ratio, , the
column stiffness is obtained by solving for

f
L
c
in Equation (1). Normalizing the
required column stiffness in terms of the
variable EI
c
/PL
2
c
gives the following result:
Equation (2)
In Table 1, the normalized column stiff-
ness is given in terms of the variable ,
showing once again that the required
c
for
=0 is 4 times higher than for =.
The authors developed expressions
similar to those presented in Table 1
for other values of roof pitch, beam-
to-column Is, , and . However, for
brevity and illustrative purposes, only
the results for =1.5 are presented in
Figure 2(a). In this fgure is shown the
ratio
c
as a function of for different
roof pitches. It is clear that as a function
of ,
c
is insensitive to variations in the
roof pitch. This fgure also shows that as
increases,
c
reduces, which matches
the results discussed in Table 1.
The design optimization rule can
be obtained in terms of summing the
normalized ratios +
c
. Once again, for
=1.5, the sum
c
+ as a function of
is shown in Figure 2(b). Curves in this
fgure clearly indicate that a minimum
point exists; this point corresponds to
the optimum beam-to-column stiffness
ratio, . Curves in this fgure also
show that the minimum point is nearly
insensitive to the roof pitch.
Consequently, the optimum beam-to-
column stiffness ratio, , can be expressed
in terms of the following minimization
rule:
A
B
C
D
E
A
c
,I
c
A
c
,I
c
A
b
,I
b
A
B
C
D
E
A
c
,I
c
A
c
,I
c
A
b
,I
b
L
b
L
c

L
b
L
c
P
P
(a) (b)
Figure 1: Portal frames under lateral loads (a) Flat frame, (b) Pitched frame.
=
3
6 4
6 24
c
f c
c
PL
L
EI
A K
D
A K
+

+
: Fixed Support
6 4 1
6 24
c
A K
L
A K D
+
=

+
: Fixed Support
$f
c
L
A
B
C
L
P
f

3
4
24
c
c
PL
EI
+
4
24D
+
(a) = I
b
/I
c
= zero
f

A
B
C
L
P
3
24
c
c
PL
EI
+
(c) = I
b
/I
c
= infinity ( )
Table 1: Deflected shape for varying values of = I
b
/I
c
for K = 1.00: Fixed frame
(b) = I
b
/I
c
= 0.74
3
1.5515
24
c
c
PL
EI

+
1.5515
24D
+
f

A
B
C
L
P
Deflected
shape
1
24D
+
d
Table 1: Defected shape for varying values of = I
b
/I
c
for = 1.00: Fixed frame.
| |
6
1
4
3
0
* *
|
|
.
|

\
|
=
(

= + = K
D
K
A A L
A
A
c
d
d
: Fixed Support
Equation (3)
S

T

R

U

C

T

U

R

E

m
a
g
a
z
i
n
e
C
o
p
y
r
i
g
h
t
S

T

R

U

C

T

U

R

E

m
a
g
a
z
i
n
e
C
o
p
y
r
i
g
h
t
August 2008 STRUCTURE magazine August 2008 35
A
D
V
E
R
T
I
S
E
M
E
N
T

-
F
o
r

A
d
v
e
r
t
i
s
e
r

I
n
f
o
r
m
a
t
i
o
n
,

v
i
s
i
t

w
w
w
.
S
T
R
U
C
T
U
R
E
m
a
g
.
o
r
g
Figure 2: Results for fxed supports.
0 0.4 0.8 1.2 1.6 2
Beam to Column I Ratio; =I
b
/I
c
0.00
2.00
4.00
6.00
8.00
S
u
m
o
f

c
+

A
B
C
D
E
Pitch = 0
Pitch = 10
Pitch = 20
Pitch = 30

0 0.4 0.8 1.2 1.6 2

Beam to Column I Ratio;=I
b
/I
c
0.00
2.00
4.00
6.00
8.00
N
o
r
m
a
l
i
z
e
d
C
o
l
u
m
n
S
t
i
f
f
n
e
s
s
,

c
(a) K (L
b
/L
c
) =1.5, Fixed support (b) K (L
b
/L
c
) =1.5, Fixed support
Thank you for reviewing this ad proof for the upcoming issue of STRUCTURE

Magazine.
To ensure that the proper advertisement for your company is run, please print out this
document, ll out the information below and fax it to us at: 608-524-4432.
Yes, the ad looks ne.
No, we require the following changes:
If we recieve no fax within 48 hours of this email, we will assume that there is no change
necessary and will run the ad as presented here. Thank you for your assistance.
DRAWING
N
E
E
D

T
O

I
N
C
R
E
A
S
E

P
R
O
F
I
T
S
?
G
I
V
E

G
.
E
.
T
.

A

C
A
L
L

O
R

V
I
S
I
T

O
U
R

W
E
B

S
I
T
E
!
L
O
O
M
I
N
G

D
E
A
D
L
I
N
E
S
?
U
N
D
E
R
S
T
A
F
F
E
D
?
P
O
O
R

Q
U
A
L
I
T
Y
?
Y
O
U

D
E
S
I
G
N

I
T
.

W
E

D
R
A
F
T

I
T
.
024 OF 102
W
W
W
.
G
L
O
B
A
L
E
N
G
T
E
C
H
.
C
O
M
F
O
R

Y
O
U
R

A
P
P
R
O
V
A
L
!
G
L
O
B
A
L

E
N
G
I
N
E
E
R
I
N
G

T
E
C
H
N
O
L
O
G
I
E
S
,

L
L
P
P
H
O
N
E
:

5
0
2
-
5
8
9
-
0
1
4
9
F
A
X
:

5
0
2
-
5
8
9
-
2
2
4
0
L
O
U
I
S
V
I
L
L
E
,

K
E
N
T
U
C
K
Y



4
0
2
0
2
3
0
4

W
E
S
T

L
I
B
E
R
T
Y

S
T
R
E
E
T
,

S
U
I
T
E

4
2
1
?
Inside_Cover_Outside_Cover_Perfe1 1 5/7/2008 2:17:31 PM
$
f
c
L
Deflected
shape
A
B
C
L
P
f

A
B
C
L
P
f

(b) I
b
/I
c
= 1.8257 (c) I
b
/I
c
= infinity
3
5.0955
24
c
c
PL
EI

+
5.0955
24
D
+
c
3
c
EI 24
PL 4
1
24
D
+
Table 2: Deflected shape for varying values of=I
b
/I
c
forK=1.00: Pinned frame
Table 2: Defected shape for varying values of =I
b
/I
c
for =1.00: Pinned frame.
Figure 3: Results for pinned supports.
0 1 2 3 4 5
Stiffness Ratio (I
b
/I
c
)
0.00
4.00
8.00
12.00
16.00
N
o
r
m
a
l
i
z
e
d
C
o
l
u
m
n
S
t
i
f
f
n
e
s
s
,

c
0 1 2 3 4 5
Beam to Column I Ratio; =I
b
/I
c
0.00
4.00
8.00
12.00
16.00
S
u
m
o
f

c
+

Pitch=0
Pitch=10
Pitch=20
Pitch=30
A
B
C
D
E
(a) K (L
b
/L
c
) =1.5, Pinned support (b) K (L
b
/L
c
) =1.5, Pinned support
continued on next page
S

T

R

U

C

T

U

R

E

m
a
g
a
z
i
n
e
C
o
p
y
r
i
g
h
t
S

T

R

U

C

T

U

R

E

m
a
g
a
z
i
n
e
C
o
p
y
r
i
g
h
t
STRUCTURE magazine August 2008 STRUCTURE magazine 36
Sameh S. Badie, Ph.D, P.E. is an Associate
Professor at the Civil & Environmental
Engineering Department of George
Washington University, Washington DC.
His research interests include analysis and
design of reinforced/prestressed concrete
structures. Any comments or insights relating
to this article are encouraged and can be
sent to badies@gwu.edu.
Pedro Silva, Ph.D,, P.E. is an Associate
Professor at the Civil & Environmental
Engineering Department of George
Washington University, Washington DC.
His research interests include analysis and
design of structures subject to seismic and
blast loading. Any comments or insights
relating to this article are encouraged and
can be sent to silvap@gwu.edu.
0 1 2 3
Beam Span-to-Column Height Ratio; =L
b
/L
c
0.00
0.10
0.20
0.30
0.40
0.50
0.60
N
o
r
m
a
l
i
z
e
d
O
p
t
i
m
u
m
B
e
a
m
t
o
C
o
l
u
m
n
I
r
a
t
i
o
1
.
P
i
n
n
e
d
S
u
p
p
o
r
t
;

2
.
F
i
x
e
d
S
u
p
p
o
r
t
;
(

/
6
+

Fixed-Fixed
Pinned-Pinned
0 1 2 3
Beam Span-to-Column Height Ratio,
0.00
2.00
4.00
6.00
8.00
10.00
N
o
r
m
a
l
i
z
e
d
C
o
l
u
m
n
I
c
,

c
Fixed w/ =0.025
Fixed w/ =0.050
Pinned w/ =0.025
Pinned w/ =0.050
(a) Normalized ( ) D
K A 6 /
*
+ and D A
*
(b) Normalized column stiffness,
c
Figure 4: Optimum ratio charts for lateral loads.
Table 1: Design example results for a roof pitch of 20

Supports
(1)

*
(3)

c
a)
(4)
s
2
c
c c
E
PL
I L =
cm
4
(in
4
) (5)
c
*
b
I I A =
mm
4
(in
4
)
(6)
$
f
b)
mm (in.)
(7)
Fixed ( ) D
K A 6 /
*
+ =0.177 0.868 2.785 1.66 (398) 1.44 (346)
92.02
(3.62)
Pinned D
A
*
= 0.354 2.236 8.903 5.30 (1,273) 11.85 (2,846)
a)
Using Figure 4 (b) Using Equations (1) and (4)
Using Figure 4(a)
(2)
b)
Table 3: Design example results for a roof pitch of 20

.
Pinned support conditions
Following the same methodology, the authors
developed expressions for frames with pinned
supports. As before, if axial deformations are
ignored, for a fat frame with pinned supports,
the lateral defection,
f
, is given by:
Equation (4)
The authors used Equation (4) to develop
Table 2, depicting the frame defection,
f
,
under =1.8257 and infnity, and for an
arbitrary . Clearly, in the condition of =0
the frame is unstable; consequently, this case
is not shown. In order to meet the required
drift ratio, , the required column stiffness is
obtained by solving for
f
L, as depicted
in Equation (4). Normalizing the required
column stiffness in terms of EI
c
/PL
2
c
results in
Equation (5).
As before, the authors developed expressions
similar to those presented in Table 2 for other
conditions and for brevity and illustrative
purposes, only the results for =1.5 are
presented in Figure 3(a).
Equation (5)
For a frame with pinned supports, Figure
3(b) shows once again that there is an
optimum value for design, corresponding
to the minimum point on these curves. The
optimization rule is mathematically expressed
by Equation (6).
Equation (6)
Design Optimization
Graphical approach
For frames with fxed and pinned supports,
the results of normalizing Equations (3)
and (6) in terms of (*+/6) and * ,
respectively, are presented in Figure 4(a).
The results of substituting Equation (3) and
Equation (6) into Equations (2) and (5) are
presented in Figure 4(b). Figure 4(a) and the
work developed in this article show that:
1) As =L
b
/L
c
increases, so does the optimum
stiffness ratio, *.
2) In frames with pinned supports, the opti-
mum stiffness ratio is signifcantly higher
than for a frame with fxed supports.
3) As the drift ratio increases, so does the
required stiffness ratio.
[ ]
D
K
A A L
A
A
3
6
1
0
* *
=

= + =
c
d
d
: Pinned Support
3
2
4
24
c
f c
c
PL
L
EI
K
D
A

: Pinned Support
2 1
4
24
c
K
L
A D
= +

: Pinned Support
Design example using optimization rules
The authors used the charts in Figure 4(b)
to develop the design example presented in
this section. Design optimization follows the
design calculations presented in Table 3. In
this section, the frame has a roof pitch of 20

and consists of Grade 50 steel with E
s
=200
GPa (29,000ksi), an applied lateral load of
889.6 kN (200 kips), and beam and column
lengths of L
b
=5.49 meters (18 feet) and L
c
=
3.66 meters (12 feet), providing an aspect
ratio =1.50. Using a design drift ratio
=0.025 (
1
/
40
), the limiting lateral defection
of the frame was
f
L
c
= 91.44 millimeters
(3.6 inches). The computed values, shown
in Table 3 column (7), only slightly exceed
this value, indicating that the design charts
presented in Figure 4 can be used reliably to
compute the optimum moments of inertia
I
c
and I
b
.
Future investigations
The design charts developed herein con-
sidered only frames under lateral loads.
The design optimizations based only on
drift ratios may not necessarily apply to
cases where gravity loads also contribute
signifcantly to the frame response. In fu-
ture investigations, design optimizations
will consider the frame response dominated
by both gravity and lateral loads.
D D
S

T

R

U

C

T

U

R

E

m
a
g
a
z
i
n
e
C
o
p
y
r
i
g
h
t
S

T

R

U

C

T

U

R

E

m
a
g
a
z
i
n
e
C
o
p
y
r
i
g
h
t