EVALUATION OF NONLINEAR STATIC PROCEDURES FOR

SEISMIC DESIGN OF BUILDINGS

By

H.S. Lew1 and Sashi K. Kunnath2

Presented at the 33rd Joint Meeting of the UJNR Panel on Wind and Seismic Effects

ABSTRACT The American Society of Civil Engineers

(ASCE) is in the process of producing an
This paper examined the effectiveness of U.S. standard for seismic rehabilitation
nonlinear static procedures for seismic existing buildings. It is based on Guidelines
response analysis of buildings. Nonlinear for Seismic Rehabilitation of Buildings
static procedures are recommended by (FEMA 273) which was published in 1997
FEMA 273 document in assessing the by the U.S. Federal Emergency
seismic performance of buildings for a given Management Agency. FEMA 273 consists
earthquake hazard representation. Three of three basic parts: (a) definition of
nonlinear static procedures specified in performance objectives; (b) demand
FEMA 273 are evaluated for their ability to prediction using four alternative analysis
predict deformation demands in terms of procedures; and (c) acceptance criteria using
inter-story drifts and potential failure force and/or deformation limits which are
mechanisms. Two steel and two reinforced meant to satisfy the desired performance
concrete buildings were used to evaluate the objective.
procedures. Strong-motion records during
the Northridge earthquake are available for FEMA-273 suggests four different analytical
these buildings. The study has shown that methods to estimate seismic demands: (i)
nonlinear static procedures are not effective linear static procedure (LSP); (ii) linear
in predicting inter-story drift demands dynamic procedure (LDP); (iii) nonlinear
compared to nonlinear dynamic procedures. static procedure (NSP); and (iv) nonlinear
Nonlinear static procedures were not able to dynamic procedure (NDP). Given the
capture yielding of columns in the upper limitations of linear methods and the
levels of a building. This inability can be a complexity of nonlinear time-history
significant source of concern in identifying analyses, engineers favor NSP as the
local upper story failure mechanisms. preferred method of analysis.

KEYWORDS: dynamic analysis; Following the analysis of a building, the

earthquake engineering; nonlinear static safety and integrity of the structural system
procedures; performance design; story drift. is assessed using acceptance criteria. For
linear procedures acceptance criteria are
1.0 INTRODUCTION based on demand-to-capacity ratios, and for
nonlinear procedures, they are based on
deformation demands.
1
Building and Fire Research Laboratory, National Institute of Standards and Technology, Gaithersburg, MD
2
University of Central Florida, Orlando, FL

1
 participates in the fundamental mode of
This paper examined the ability of the vibration.
FEMA 273 nonlinear static procedures to
predict deformation demands in terms of ii) Uniform Load Pattern (FEMA-2)
inter-story drift and potential failure A uniform load pattern based on lateral
mechanisms in the system. forces that are proportional to the total
mass at each floor level.

1.1 Nonlinear Static Procedures for wx

Seismic Demand Estimation Fx = n
V
∑w
i =1
i
There are several procedures that can be
adopted for conducting a nonlinear static
analysis. While the fundamental procedure This pattern is expected to simulate story
for the step-by-step analysis is essentially shears.
the same, the different procedures vary
mostly in the form of lateral force iii) Modal Load Pattern (FEMA-3)
distribution to be applied to the structural A lateral load pattern proportional to the
model in each step of the analysis. FEMA- story inertia forces consistent with the
273 recommends the following three story shear distribution calculated by a
procedures: combination of modal responses.

i) Inverted Triangular Pattern (FEMA-1) Each of the above procedures will be

A lateral load pattern represented by the evaluated using four sample structures: 6-
following FEMA-273 equation: story steel frame building; 13-story steel
frame building, 7-story concrete building,
and 20-story concrete building. These
wx h kx buildings have strong-motion records from
Fx = n
V
k the Northridge earthquake. The strong-
∑ wi hi motion records at the roof level were used to
i =1
calibrate the building models.
where: Fx = lateral load at floor level x
wx,i = weight at floor level x,i
1.2 Benchmark Response
hx,i = height from base to floor
level x,i
Time-history analyses are performed on the
k = 1.0 for T < 0.5 seconds;
four buildings, and the results are used as
k = 2.0 for T > 2.5 seconds, with
benchmark values to which the results of
linear
each of the nonlinear static analyses are
interpolation for intermediate
compared. The target displacements are
values
based on either the actual measured building
V = total lateral load (base shear)
response or a displacement consistent with
to be applied to the building
the guidelines of FEMA-273. The following
This load pattern results in an inverted
aspects of the building response behavior are
triangular distribution across the height
examined:
of the building and is normally valid
when more than 75% of the mass

2
 1) The ability of each nonlinear static 1973 UBC and built in 1975. The layout of
procedure to predict the interstory the perimeter frames and a typical elevation
drift demands. of one of these perimeter frames are shown
2) The ability of each nonlinear static in Figure 3. The overall building plan
procedure to identify the distribution dimensions are 160 x 160 ft (48.7 x 48.7 m).
of yielding (plastic hinge formation) The typical floor system consists of about
in the structure leading to a local or 2.5 in (64 mm) of concrete fill over 3-inch
global failure mechanism. (76-mm) steel decking.

The instrumentation was located in the

2.0 MODELING AND basement and on the sixth and twelfth floors.
CALIBRATION OF BUILDINGS Recorded accelerations at the basement
indicate that the building experienced a PGA
2.1 6-Story Steel Building of 0.41g in the N-S direction and 0.32g in
the E-W direction. Approximately 12% of
This building, located in the Los Angeles the connections of the west perimeter (N-S
area, was designed in 1976 based on the direction) frame fractured during the
1973 UBC requirements. The primary earthquake. Connections of the remaining
lateral load resisting system is moment three sides fractured less than 6%.
frames around the perimeter of the building.
Since the structural system is symmetrical, Since the building was symmetrical in both
only one of the perimeter frames was directions, only one frame was analyzed.
considered in the evaluation. The interior Gross section properties were used to
frames were designed as gravity frames and compute the initial stiffness values. Each
consist of simple shear connections only. perimeter frame was assumed to carry 50%
The plan view of the building and the of the building mass in each direction.
elevation of a typical frame used in the
analysis are shown in Figure 1. The building response in the East-West
direction is shown in Figure 4. This
The building was instrumented with a total indicates that the building stiffness and mass
of 13 strong-motion sensors at the ground, modeling are reasonably accurate.
2nd, 3rd and roof levels.
2.3 7-Story RC Building
The building performed well in the
Northridge earthquake with no visible signs A seven-story nonductile reinforced
of damage. Recorded data indicate an concrete building was designed and
essentially elastic response. Minimal constructed during the mid-1960s in
calibration was required to obtain a good accordance with the 1964 Los Angeles City
match of the computed response with the building code. The building sustained
observed roof response (Figure 2). damage in both the 1971 San Fernando and
the 1994 Northridge earthquake. The
2.2 13-Story Steel Building building suffered relatively minor damage in
the 1971 event but was heavily damaged
A 13-story steel moment-resisting frame during the 1994 event.
building (13 floors above ground plus one-
story basement) was designed based on the

3
This building has a rectangular plan (Figure There are twenty stories above ground and a
5) with overall dimensions of approximately basement. The typical story height is 8.75
63 ft (19 m) in the north-south (transverse) ft (2.7 m). The bottom two and top two
direction and 150 ft (46 m) in the east-west levels have different heights as indicated in
(longitudinal) direction. The total height of the figure. This building experienced a PGA
the building is 66 ft (20 m). The floor of 0.33g at the basement level during the
system consists of reinforced concrete flat 1994 Northridge earthquake.
slabs. Perimeter spandrel beam-column
frames resist primarily the lateral load in The gravity load resisting system of the
each direction. The interior slab-column building consists of concrete slab supported
frames are also expected to carry a on concrete beams and columns. The
significant portion of the lateral load. primary lateral force resisting system
consists of moment-resisting frames with
The building was repaired after the 1971 strong shear walls in the basement only.
event. During the 1994 event, damage was Beams are provided between all column
primarily confined to the longitudinal lines except between lines 2 and 3 for
perimeter frames with the most severe intermediate frames (also shown in the
damage between the fourth and fifth floors figure). Post-earthquake survey of the
of the south perimeter frame in the form of building indicated that the building suffered
shear failure of the columns immediately significant non-structural and content
below the fifth floor spandrel beam. Many damage. However, there was no evidence of
beam-column joints suffered minor to any significant structural damage at any
moderate shear cracks below the fifth floor level.
level. Concrete spalling and hairline
flexural cracks were observed in several Two typical frames were considered in the
spandrel beams. In the transverse direction, modeling of the building. Eight frames with
the damage was limited to minor flexural three continuous beams across all three
cracks in the beams in the end bay. spans and seven frames with the
discontinuous beams were modeled
The two interior frames and the two exterior separately. The interior slab between the
frames were considered to be identical for discontinuous beams was assumed to remain
purposes of the modeling. The response of elastic. A total of 19 column types and 19
the calibrated building model to the input beam types were used to construct the
ground motion is shown in Figure 6. The building model. A fictitious stiff column
comparison indicates that the resulting was added at the basement level to simulate
model is adequate for the analytical studies. the effect of the rigid walls at this level.
Since the recorded ground motion was at the
basement level, it was considered essential
2.4 20-Story RC Building to model the basement level.

The building was designed in 1967 and Using gross section inertia and uncracked
constructed in 1968 in accordance with the concrete stiffness, a fundamental building
Los Angeles City building code. As shown period of 2.46 seconds was obtained. Since
in Figure 7, the building has a rectangular this value is in good agreement to the
plan with dimensions of 184 x 58 feet (56 x recorded data, no additional tuning of the
18 m) in the longitudinal (EW) and model was required. Figure 8 shows a plot
transverse (NS) directions, respectively.

4
of the time-history response of the roof, nonlinear static procedures but under-
which indicates that the building model is estimated at all other levels except the
adequate. second level. It should be noted that the
maximum inter-story drift in the case of the
time-history response could occur at
3.0 EVALUATION OF NONLINEAR different times.
STATIC PROCEDURES FOR
SEISMIC DEMAND A comparison of the hinge patterns at the
ESTIMATION target displacement is shown in Figure 11.
The plots show the distribution of beam and
3.1 Analysis of 6-Story Steel Building column yielding for each of the analyses
using different shape of lateral load
As indicated previously, this building did distribution. The yielding patterns are
not sustain damage during the Northridge consistent with the story drift profiles.
earthquake. Instrument records and the
calibrated response indicate that the 3.2 Analysis of 13-Story Steel Building
structural response was essentially elastic.
To examine the effectiveness of nonlinear This building experienced widespread
static procedures to estimate seismic connection damage during the Northridge
demands, particularly in the inelastic phase earthquake. Since this implies inelastic
of the response, the recorded base motion action, it was reasonable to select the
was scaled uniformly to achieve a roof recorded roof displacement as the target
displacement in the yielded state of the displacement for the comparative study.
system (the “target” displacement). This
approach allowed a better comparison of the Figure 12 compares the displacement profile
different methods of analyses. of the building at the target displacement for
the different nonlinear static procedures.
Figure 13 compares the maximum inter-
The target roof displacement was story drift along the building height.
determined according to FEMA-273. The
computed target roof displacement was 18 While the displacement profiles for all
inches (450 mm). The scale factor for the methods look very similar, the predicted
recorded base motion was established maximum inter-story drift varies
through trial-and-error until the computed considerably for the nonlinear static
roof displacement matched the target procedures when compared to the dynamic
displacement of 18 inches. response indicating the presence of higher
modes in the response that cannot be
The displacement profile at maximum roof captured by the static methods.
displacement is shown in Figure 9. It is
observed that all pushover procedures The locations of plastic hinges predicted by
predict a similar profile. In comparison with the different analytical methods are shown
nonlinear dynamic procedure, the in Figure 14. Since FEMA-1, FEMA-2 and
displacements are over-predicted. FEMA-3 resulted in almost the same hinge
configuration, only one of three patterns is
The drift profiles over the height of the shown.
building are shown in Figure 10. The drift
at the first level is over-estimated by all

5
3.3 Analysis of 7-Story Concrete third floor. Figure 20 presents a summary of
Building the plastic hinge patterns resulting from each
of the different analytical procedures. The
This building experienced inelastic behavior locations of plastic hinges are consistent
and significant structural damage during the with the inter-story drift profiles.
Northridge earthquake. Therefore, it was
not necessary to establish a projected target
displacement for the response analyses. The 4.0 CONCLUSIONS
actual displacement obtained from the
nonlinear time-history analysis was used as This paper examined the effectiveness of
the target displacement. The correlation of nonlinear static procedures for analysis of
the time-history procedure to the observed inelastic response of buildings. Specifically,
response was presented in the earlier the FEMA 273 procedures are evaluated to
section. see whether nonlinear static procedures can
predict deformation demands in terms of
The displacement profiles using the different inter-story drift and potential failure
analytical approaches are shown in Figure mechanisms in the system.
15. The corresponding drift values are
shown in Figure16. The overall trends in The evaluation was carried out using four
both displacement and drift are similar. The buildings for which instrumented data were
nonlinear static procedures tend to available:
underestimate drift in the upper levels. As
seen in Figure 17, the plastic hinge 1) Six-Story Steel Moment-Frame
formation in the upper levels is not Building,
estimated adequately by the nonlinear static 2) Thirteen-Story Steel Moment-
procedures. Resisting Frame Building,
3) Seven-Story Reinforced Concrete
3.4 Analysis of 20-Story Concrete Moment Frame Building, and
Building 4) Twenty-Story Concrete Moment
Frame Building.
This building did not experience damage
during the Northridge earthquake. A frame model of each of the above
Instrument records and the calibrated buildings was first calibrated against
response indicate that the structural response observed instrument data. Then, each of the
was essentially elastic. As was in the 6- building models was analyzed using a
story steel building (Section 3.1), the detailed nonlinear time-history analysis
recorded base motion was scaled uniformly followed by a series of nonlinear static
to achieve a roof displacement in the yielded pushover procedures. They were:
state of the system. The displacement
profiles at the target roof displacement are 1) A lateral load pattern represented by an
shown in Figure 18. inverted triangular load (FEMA-1).
The drift profiles predicted by different 2) A uniform load pattern based on lateral
procedures along the height of the building forces that are proportional to the total
are compared in Figure 19. It is seen that all mass at each floor level (FEMA-2).
nonlinear static procedures underestimate 3) A lateral load pattern proportional to the
the drift at almost all story level above the story inertia forces consistent with the

6
 story shear distribution calculated by a REFERENCE
combination of modal responses using a
response spectrum analysis (FEMA-3) Federal Emergency Management Agency
(FEMA), (1979), NEHRP Guidelines for the
The following conclusions are drawn: Seismic Rehabilitation of Buildings,
Washington, D. C.
1. Nonlinear static procedures are generally
not effective in predicting inter-story
drift demands compared to nonlinear
dynamic procedures. Drifts are
generally under-estimated at upper levels
and sometimes over-estimated at lower
levels.

2. The peak displacement profiles predicted

by both nonlinear static and nonlinear
dynamic procedures are in agreements.
This suggests that the estimation of the
displacement profile at the peak roof
displacement by nonlinear static
procedures is reasonable so long as inter-
story drifts at the lower levels are
reasonably estimated.

3. Nonlinear static methods did not capture

yielding of columns at the upper levels.
This inability can be a significant source
of concern in identifying local upper
story mechanisms.

7
 6 BAYS @ 20 ft = 120 ft

30 ft 30 ft 30 ft 30 ft

30 ft
20 ft

6 BAYS @ 20 ft = 120 ft
20 ft
20 ft
30 ft
TYPICAL FRAMING FLOOR PLAN

20.0 ft 20.0 ft 20.0 ft 20.0 ft 20.0 ft 20.0 ft

W24X68 ROOF
(TYP.)
W14X95

5 FLOOR @ 13.0 ft = 65.0 ft

W24X84 6TH FLOOR

(TYP.)
W24X84 5TH FLOOR
(TYP.)
W14X136

W27X84 4TH FLOOR

82.0 ft

(TYP.)
W27X102 3RD FLOOR
(TYP.)
W14X184

W30X116 2ND FLOOR

(TYP.)
17.5 ft

Figure 1 Plan and Elevation of 6-Story Steel Building

8
 4 Rec orded
3 IDA S S

ROOF DISPL (IN)

2
1
0
-1
-2
-3
-4
-5
0 5 10 15 20 25 30 35 40
TIM E (SEC)

Figure 2 Recorded vs. Computed Response of 6-Story Steel Building

W27X84
2238"

W14X167
4 5 6 7 8 9 W33X118

W33X118

W14X246
B W33X130

W33X130

W14X287
C W33X141

N W33X141

TRUE NORTH

W14X314
D W33X152
12 @ 158"
5 @ 32'

W33X152
W14X398

E W33X152

W33X152
W14X426

F W33x152

W36X230
366"
W14X500

G W33X194
174"

5 @ 32 '
Ground
Elev
PLAN VIEW OF PERIMETER FRAMES
ELEVATION

Figure 3 Plan and Elevation of Perimeter Frame of 13-Story Steel Building

9
 30
R e c o rd e d
20
S im u la t e d

ROOF DISPL (IN)

10

-1 0

-2 0

-3 0
0 5 10 15 20

T IM E (S E C )

Figure 4 Recorded vs. Computed Response of 13 Story Steel Building

1 2 3 4 5 6 7 8 9

241"
8" brick wall

C
N

250"
B

241"
A
8 bays @ 225"

8'8"
6 @ 8'8.5"

13'6"

Figure 5 Typical Floor Plan and Elevation of 7-Story RC Building

10
 10
8

ROOF DISPLACEMENT (IN)

6
4
2
0
-2
-4
-6
-8 Recorded
IDASS
-10
-12
0 5 10 15 20
TIME (SEC)

Figure 6 Recorded vs. Computed Response of 7-Story RC Building

2496"
2385"
1 2 3 4

16
@
10
14 5"
@ N
13 '

A A

552"

348"

162"

Elev
238" 211" 225"

PLAN Section A-A

Figure 7 Plan and Elevation of 20-Story RC Building

11
 10
8 Recorded
6 Computed

ROOF DISPL (IN)

4
2
0
-2
-4
-6
-8
0 5 10 15 20
TIME (SEC)

Figure 8 Recorded vs. Computed Response of 20-Story RC Building

6 6
NDP NDP
FEMA-1 FEMA-1
5 5
FEMA-2 FEMA-2
FEMA-3
STORY LEVEL

STORY LEVEL

FEMA-3
4 4

3 3

2 2

1 1

0 0
0.0 10.0 20.0 0 2 4 6 8
DISPLACEMENT (IN) STORY DRIFT (%)

Figure 9 Displacement Profiles of 6 Figure 10 Inter-story Drift Profiles of 6

Story Steel Building Story Steel Building

12
 (a) NDP (b) FEMA-1

(c) FEMA-2 (d) FEMA-3

Figure 11 Hinging Patterns of 6-Story Steel Building

at Target Displacement

14 14

12 12

10 10
STORY LEVEL

STORY LEVEL

8 8

6
6

4
4
NDP
NDP
2 FEMA-1
2
FEMA-2 FEMA-3
0
0
0 5 10 15 20
0 5 10 15 20
DISPLACEMENT (IN)
DISPLACEMENT (IN)

Figure 12 Displacement Profiles of Different NSPs vs.NDP of 13-

Story Steel Building

13
 14 14

12 12

STORY LEVEL 10 10

STORY LEVEL
8 8

6 6

4 4

2 2

0 0
0 0.5 1 1.5 2 0 0.5 1 1.5 2

S T O R Y D R IF T (% ) S T O R Y D R IF T (% )

Figure 13 Inter-story Drift Profiles of 13-Story Steel Building

(a) NDP (b) NSP

Figure 14 Hinging Patterns of 13-Story Steel Building

14
 7 7
NDP
NDP FEMA -1
6 6
FEMA -1 FEMA -2

FEMA -2 FEMA -3
5 5
FEMA -3
STORY LEVEL

STORY LEVEL
4 4

3 3

2 2

1 1

0 0
0 5 10 0.00 1.00 2.00 3.00 4.00
D ISPL AC EMEN T (IN ) S TO RY DRIF T (% )

Figure 15 Displacement Profiles of 7- Figure 16 Inter-story Drift Profiles of

Story Concrete Building 7-Story Cconcrete Building

(a)NDP

(b) FEMA-1and -3 (c) FEMA-2

Figure 17 Hinging Patterns of 7-Story Concrete Building

15
 21 21

NDP NDP
18 FEMA-1 18
FEMA-2 FEMA-3

15 15

STORY LEVEL
STORY LEVEL

12 12

9 9

6 6

3 3

0 0
0 10 20 30 40 50 0 10 20 30 40 50

DISPLACEMENT (IN) DISPLACEMENT (IN)

Figure 18 Displacement Profiles of 20-Story Concrete Building

21 21

18 18

15 15
STORY LEVEL
STORY LEVEL

12 12
NDP
N DP
FE M A -1
F E M A -3
FE M A -2
9 9

6 6

3 3

0
0
0 .0 0 2 .0 0 4 .0 0 6 .0 0 8 .0 0
0.00 2.00 4.00 6.00 8.00
S T O R Y D R IFT (% ) S TO R Y D R IF T (% )

Figure 19 Inter-story Drift Profiles of 20-Story Concrete Building

16
 Frame 1 Frame 2

(a) Nonlinear Dynamic Procedure

Figure 20 Hinging Patterns of 20-Story Concrete Building

17
 Frame 1 Frame 2

(b) Nonlinear Static Procedure

FEMA 1, FEMA 2, FEMA 3

Figure 20 Hinging Patterns of 20-Story Concrete Building

18