Proceedings of the 8th U.S.

National Conference on Earthquake Engineering

April 18-22, 2006, San Francisco, California, USA
Paper No. 1076

A SEISMIC DESIGN LATERAL FORCE DISTRIBUTION BASED ON INELASTIC

RESPONSE

S. H. Chao 1 , S. C. Goel2, and S. S. Lee3

ABSTRACT

It is well recognized that structures designed by current codes undergo large

inelastic deformations during major earthquakes. However, lateral force
distributions given in the seismic design codes are typically based on results of
elastic response studies. In this paper, current code lateral force distributions are
reviewed in view of the results obtained from nonlinear dynamic analyses of
several example structures. It is concluded that code lateral force distributions do
not represent the maximum force distributions that may be induced during
nonlinear response. That may lead to inaccurate predictions of deformation and
force demands, causing structures to behave in a rather unpredictable and
undesirable manner. A new lateral force distribution based on study of inelastic
behavior is developed by using relative distribution of maximum story shears of a
number of example structures subjected to a variety of earthquake ground motions.
The results show that the proposed lateral force distribution is more rational and
gives a much better prediction of inelastic seismic demands at global as well as at
element levels.

Introduction

It is well-known that the current lateral seismic force distributions in building codes are
based on first-mode dynamic solution of lumped MDOF elastic system (Clough and Penzien
1993; Chopra 2000; NEHRP 2001), which can be expressed as:

⎛ wiφi1 ⎞
fi1 = V1 ⎜ n ⎟
⎜⎜ ∑ w jφ j1 ⎟⎟
(1)
⎝ j =1 ⎠
where fi1 is the lateral force at level i ; V1 is the contribution of the first-mode to the total base
shear force; w j is lumped seismic weight at jth level; φ j is the amplitude of the first-mode at jth
level. The code expression assumes that the lateral force distribution can be expressed using the
first-mode deflected shape of an elastic lumped-mass system subjected to dynamic loading,
which can be written as:

1
Graduate Research Assistant, Dept. of Civil and Environmental Engineering, University of Michigan, Ann Arbor,
MI 48109-2125.
2
Professor, Dept. of Civil and Environmental Engineering, University of Michigan, Ann Arbor, MI 48109-2125
3
Senior Bridge Engineer, URS Corporation, Roseville, CA.
 φi1 = hik L (2)

where L is the total height of the structure; hik is the height of level i above the base with
exponent k being related to the building’s fundamental period (T). Observations of the elastic
response of tall buildings suggest that the first-mode shape is close to a straight line (k =1) when
the fundamental period is 0.5 sec or less; and is close to a parabola (k = 2) when the fundamental
period is 2.5 sec or more. This leads to the following code lateral force distribution (ICC 2003):

⎛ wi hik ⎞
Fi = V ⎜ n ⎟ (3)
⎜ ∑ w j h kj ⎟
⎜ ⎟
⎝ j =1 ⎠

where V is the total design base shear, which replaces V1 in Eq. 1 since V1 is a dominant part of
the total force V.

As can be seen, equivalent static design lateral forces in the current codes are obtained
from simplified models assuming that the structures behave elastically and primarily in the first
mode of vibration. However, building structures designed according to current code procedures
are expected to undergo large deformations in the inelastic range when subjected to major
earthquakes, thereby leading to lateral force distributions which can be quite different from those
given by code formulas such as Eq. 3. In order to achieve the main goal of performance-based
seismic design, i.e. a desirable and predictable structural response, it is necessary to account for
inelastic behavior of structures directly in the design process. The commonly used elastic
analysis and design procedures in current practice, together with elastic design lateral force
distributions, do not fulfill this goal in a realistic manner.

One of the essential elements of performance-based seismic design of structures should

be to use more realistic design lateral force distribution, which represents peak lateral force
distribution in a structure in the inelastic state, as well as the higher mode effects. This paper
introduces a new lateral force distribution based on the study of inelastic responses of various
types of steel frame systems, using extensive nonlinear dynamic analysis results.

New Lateral Force Distribution

The format for this new design lateral force distribution based on inelastic state of a
structure was originally proposed by Lee and Goel (2001) using shear proportioning factor
derived from the relative distribution of maximum story shears of a large number of moment
frames (MFs) subjected to selected earthquake records:

Fi = (βi − βi+1 ) Fn when i = n, βn+1 = 0 (4-1)

 αT −0.2
⎛ wn h n ⎞⎟
⎜⎜ ⎟⎟
Fn = V ⎜⎜ n ⎟ (4-2)
⎜⎜ ∑ j j ⎟⎟⎟
w h
⎝ j=1 ⎠
αT −0.2
⎛ n ⎞
⎜⎜ w h ⎟⎟
V ⎜ ∑ i i ⎟⎟
βi = i = ⎜⎜⎜ i ⎟⎟⎟ (4-3)
Vn ⎝⎜ wn h n ⎠

where βi is the shear distribution factor at level i; Vi and Vn , respectively, are the story shear
forces at level i and at the top (nth) level; wi and w j are the seismic weights at levels i and j,
respectively; hi and h j are the heights of levels i and j from the ground, respectively; wn is the
weight at top level; hn is the height of roof level from ground; T is the fundamental
period; Fi and Fn are the lateral forces at level i and top level n, respectively; V is the total design
base shear. The value of parameter α was originally proposed as 0.5 by Lee and Goel (2001),
which was revised to 0.75 based on more extensive nonlinear dynamic analyses on eccentrically
braced frames (EBFs) and special truss moment frames (STMFs) by Chao and Goel (2005 and
2006) and briefly presented in this paper.

Justification of the New Lateral Force Distribution

Nonlinear dynamic analyses were carried out to evaluate the validity of the proposed
design lateral force distribution. Its merits over the code distributions are discussed as follows:

Relative Story Shear Distributions

The proposed new lateral force distribution was justified by investigating the relative
story shear distributions (defined as the ratio of peak story shear force at level i to that at top
level n; i.e. Vi / Vn ) as obtained from nonlinear dynamic analyses. Note that story shear
distribution and lateral force distribution have direct relationship. As can be seen in Fig. 1, code
lateral force distributions generally result in relatively smaller design story shear forces at upper
levels while nonlinear dynamic analyses show that the upper floors are subjected to larger forces
than those given by the code formulas. The relatively smaller lateral design forces at upper floors
generally lead to smaller member sizes, and therefore, larger story drifts and member
deformations at upper levels. This trend was noticed in all types of study frames.

On the other hand, the proposed lateral force distribution (thus, the relative story shear
distribution) is closer to the results obtained from nonlinear dynamic analyses. It is noted, as
shown in Fig. 1, that relative story shear distribution using α = 0.5 generally represents a lower
bound of the nonlinear dynamic analysis results. This would normally lead to larger design
forces at upper floors, which may result in concentration of inelastic deformation at the lower
levels. Further analyses by Chao and Goel (2005 and 2006) showed that relative story shear
distribution using α = 0.75 represents an upper bound of the nonlinear dynamic analysis results
(Fig. 1) and generally leads to more uniform deformations of elements as well as stories over the
height of the structure.

Fig. 2 shows the relative story shear distributions for a 10-story EBF, as obtained from
elastic and inelastic dynamic response analyses. As can be seen, the elastic relative story shear
distributions are not unique and can vary quite a bit with the ground motion used. On the other
hand, the inelastic relative story shear distributions were closer to the values given by the
proposed expression (Eq. 4; α = 0.75 ). This is quite understandable because the story shears in
the inelastic state are limited by the strength of the system. Similar trend was also observed by
Goel (1967) in his early study of inelastic seismic behavior of multistory steel moment frames,
as shown in Fig. 3.

Maximum Interstory Drift Distributions

Since the proposed lateral force distribution is based on inelastic response, the structures
designed by using such distribution tend to be better proportioned. In other words, the possibility
of over-design or under-design in certain regions is greatly reduced. Fig. 4 shows that, an EBF
designed by using the proposed lateral force distribution (labeled as PPD frame in the figure)
showed more uniform distribution of maximum interstory drifts than EBF designed by IBC
lateral force distribution (labeled as IBC frame in the figure). A uniform interstory drift
distribution also implies that intended “fuse” elements (such as beams in moment frames, shear
links in EBFs, and chord members in STMFs) at all levels can be utilized to dissipate earthquake
energy, minimizing the possibility of concentration of excessive inelastic deformation and
damage in some stories or localized regions.

Column Design Moments

It has been pointed out by several investigators in the past that when a structure is
subjected to seismic loading, especially in the inelastic state, large moments can occur in the
columns, which can be quite different from those calculated by elastic analysis. The
conventional capacity design approach used to design columns is not appropriate and usually
does not accurately predict the maximum moments and their location in the columns (Bondy
1996; Medina and Krawinkler 2005), because the columns are subjected to moments not only
from those delivered from the beams or other members framing into the columns (conventional
capacity design approach), but also from their own deformation.

In view of the above mentioned shortcoming in conventional design approach, an

alternative design method was proposed by Leelataviwat, Goel, and Stojadinović (1999) by
considering the equilibrium of an entire “column tree” in the extreme limit state. Fig. 5 shows
the free-body diagram of an exterior “column tree” of a moment frame for calculating the design
column moments in each story. In order to ensure the formation of strong-column weak-beam
mechanism, columns should be designed for maximum expected forces by considering a
reasonable extent of strain-hardening in the beam plastic hinges. The moment at a strain-
hardened beam plastic hinge can be obtained by multiplying its nominal plastic moment ( M pb )
by an overstrength factor ( ξ ), which accounts for the effect of strain-hardening and material
overstrength. The column moments and shear forces can be calculated by applying the expected
beam end moments and equivalent lateral forces applied at each level ( Fiu ) necessary to keep the
“column tree” in equilibrium. An appropriate moment at the column base needs also to be
applied. It is in the calculation of column design moments and shears that the importance of
using a realistic lateral force distribution becomes critical.

Fig. 6 shows comparison of the column moments calculated by the above mentioned
procedure with the new lateral force distribution (Eq. 4; α = 0.5 ), the UBC (ICBO 1997) lateral
force distribution, and maximum column moment envelopes from nonlinear dynamic analyses
for the exterior and interior columns of a 9-story moment frame from the study by Lee and Goel
(2001). The column design moments calculated by using the new lateral force distribution agree
very well with the maximum column moment envelopes at most levels. On the other hand, the
design column moments calculated by using UBC lateral force distribution significantly deviate
from the nonlinear dynamic analysis results.

Higher Mode Effects

The new lateral force distribution increases the forces in the upper stories, thereby better
representing the higher mode effects. Fig. 7 shows the shapes of IBC (ICC 2003) and the
proposed (Eq. 4; α = 0.75 ) lateral force distributions for the example 9-story moment frame. It
can be seen that the new lateral force distribution results in more force at the upper levels than
given by the IBC lateral force distribution. This implies that the additional design force at the top
level as given by the IBC (intended to reflect the higher mode effects) may not be adequate,
especially when the structures respond in the inelastic state.

Conclusion

This paper presents a brief study aimed at formulating a more realistic design lateral
force distribution, which accounts for inelastic behavior of structures when subjected to major
earthquakes. Use of a realistic force distribution based on inelastic response is one of the
important steps in a comprehensive seismic design methodology if accurate representation of
expected structural response is to be realized. It should be mentioned that the proposed lateral
force distribution can be further refined by using more ground motion time-histories than used in
this study. The study also needs to be extended to include other types of structural systems, such
as braced frames, RC frames and shear wall-frame systems in order to arrive at a design lateral
force distribution that would be suitable for more general application in practice.

Following conclusions can be drawn from the results of this study:

1. Frames (MF, EBF, and STMF) designed by using the proposed lateral force distribution
resulted in maximum story shears that agreed well with those obtained from nonlinear
dynamic analyses. Maximum story shear distributions as given in the codes, which are based
on first-mode elastic behavior, deviate significantly from time-history dynamic analysis
results no matter whether the structures respond in the elastic or inelastic range.

2. Frames designed by using the proposed lateral force distribution generally experienced more
uniform maximum interstory drifts along the height.
3. The proposed column design procedure along with the new design lateral force distribution
gives a very good estimation of maximum column moment demands when the structures are
responding to severe ground motions.

4. Higher mode effects are also well reflected in the proposed design force distribution.

9
El Centro
Eq. 4 (α = 0.75 ) Newhall
8
Sylmar
Synthetic
7
UBC 97
6
Story Level IBC 2003
5
Eq. 4 (α = 0.5 )
4

2
9-Story SMF
1
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

Relative Distribution of Story Shear Vi / Vn

(a) Moment Frame

10
la01
la02
9 la09
Eq. 4 (α = 0.75) la12
la13
8 la16
la17
7 la19
Story Level

6
IBC 2003
5
Eq. 4 (α = 0.5 )
4

2
10-Story EBF
1
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

Relative Distribution of Story Shear Vi / Vn

(b) Eccentrically Braced Frame

Figure 1. Relative story shear distributions from nonlinear dynamic analysis, code expressions,
and proposed expression for study MF, EBF, and STMF.
 9 LA02 (El Centro)

Eq. 4 (α = 0.75 )
LA04 (Array #05)
LA07 (Barstow)
8 LA09 (Yermo)
LA12 (Loma Prieta)
LA13 (Newhall)
LA17 (Sylmar)
7 LA18 (Sylmar-2)
LA19 (North Palm)
LA21(Kobe_2%in50yr)
LA23 (Loma Prieta_2%in50yr)
6

Story Level
LA26 (Rinaldi_2%in50yr)
LA27 (Sylmar_2%in50yr)
LA30 (Tabas_2%in50yr)

5 Eq. 4 (α = 0.5 )

4 NEHRP 2000
(IBC 2003)
3

2
9-Story STMF
1
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
Relative Distribution of Story Shear Vi / Vn

(c) Special Truss Moment Frame

Figure 1 (continued). Relative story shear distributions from nonlinear dynamic analysis, code
expressions, and proposed expression for study MF, EBF, and STMF.

10 10
LA01 Earthquake LA09 Earthquake
9 9

8 8
Eq. 4 (α = 0.75 ) Eq. 4 (α = 0.75 )
7 7
Story Level
Story Level

IBC 2003 IBC 2003

6 6

5 5

4 4

3 10-Story EBF 3 10-Story EBF

2 Elastic Analysis Elastic Analysis
2
Inelastic Analysis Inelastic Analysis
1 1
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

Relative Distribution of Story Shear Vi / Vn Relative Distribution of Story Shear Vi / Vn

Figure 2. Relative story shear distributions from elastic and inelastic dynamic analysis, code
expressions, and proposed expression for study EBF.
 10 10
10-Story SMF
9 9
(Goel, 1967)
Elastic Analysis 8
8 Inelastic Analysis
7
7
Story Level

Story Level
IBC 2003 6
6
5
Eq. 4 (α = 0.5 )
5
4
4
3
3 10-Story EBF
Eq. 4(α = 0.75) 2
Interstory Drift
2 1 IBC
Taft Event (1952) PPD LA16 Earthquake
1 0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0

Relative Distribution of Story Shear Vi / Vn Maximum Interstory Drift (%)

Figure 3. Elastic and inelastic relative story Figure 4. Comparison of maximum interstory drift
shear distributions of a 10-story moment between EBFs designed by using the proposed and
frame (Goel, 1967). code design lateral force distributions.

ξi M pbi M c (h)
Fiu
hi
M pc
M pc

Figure 5. Free body diagram of an exterior “column tree”.

 Design Moment
9 9
Design Moment
8 (UBC Distribution) 8
El Centro
7 7
6 New hall 6
5 Sylmar 5
Story

Story
4 Synthetic 4
3 3
2 2
1 1
0 0
0

0
00

00

00

00

00
0
0
00

00

00

00

40

60
20

40

20
-4

-2

-4

-2
Moment (kip-ft) Moment (kip-ft)

Figure 6. Design column moments and maximum column moment envelopes from nonlinear
dynamic analyses for a 9-Story MF. Left: Exterior column; Right: Interior column.

Eq. 4 (α = 0.75)
IBC 2003

Figure 7. The proposed new lateral force distribution and IBC 2003 lateral force distribution for
the 9-story MF.
 References

Bondy, K. D., 1996. A more Rational Approach to Capacity Design of Seismic Moment Frame Columns,
Earthquake Spectra, 12 (3), 395-406.

Chao, S. H. and Goel, S. C., 2005. Performance-Based Seismic Design of EBF Using Target Drift and
Yield Mechanism as Performance Criteria, Report No. UMCEE 05-05, Department of Civil and
Environmental Engineering, University of Michigan, Ann Arbor, MI.

Chao, S. H. and Goel, S. C., 2006. Performance-Based Seismic Design of STMF Using Target Drift and
Yield Mechanism as Performance Criteria, UM Report (in preparation), Department of Civil and
Environmental Engineering, University of Michigan, Ann Arbor, MI.

Clough, R.W., and Penzien, J., 1993. Dynamics of Structures, 2nd Ed., McGraw-Hill, Inc., New York.

Chopra, A. K., 2000. Dynamics of Structures—Theory and Applications to Earthquake Engineering, 2nd
Ed., Prentice Hall, Englewood Cliffs, New Jersey.

Goel, S. C., 1967. Inelastic Behavior of Multistory Building Frames Subjected to Earthquake Motion,
Ph.D. Thesis, University of Michigan, Ann Arbor, MI.

International Code Council (ICC), 2003. International Building Code, ICC, Birmingham, AL.

Lee, S. S. and Goel, S. C., 2001. Performance-Based Design of Steel Moment Frames Using Target Drift
and Yield Mechanism, Report No. UMCEE 01-17, Department of Civil and Environmental
Engineering, University of Michigan, Ann Arbor, MI.

Leelataviwat, S., Goel, S. C., and Stojadinović, B., 1999. Toward Performance-Based Seismic Design of
Structures, Earthquake Spectra, 15 (3), 435-461.

Medina, R. A., and Krawinkler, H., 2005. Strength Demand Issues Relevant for the Seismic Design of
Moment-Resisting Frames, Earthquake Spectra, 21 (2), 415-439.

NEHRP, 2001. Recommended Provisions for the Development of Seismic Regulations for New Buildings
(FEMA 368), Federal Emergency Management Agency, Washington, D. C.

NEHRP, 2001. Commentary on Recommended Provisions for the Development of Seismic Regulations for
New Buildings (FEMA 369), Federal Emergency Management Agency, Washington, D. C.

UBC, 1997. Uniform Building Code, International Conference of Building Officials, Whittier, CA.