8.

Earthquake-Resistant Engineering
of Steel Structures

Hitoshi Kuwamura

8.1 Introduction

Steel structures are predominant in building construction in Japan.

As shown in Fig. 8-1, steel structures increased rapidly after World War
II and at present account for about 40% of the total floor areas of newly
constructed buildings in a year. This ratio is comparable with wooden struc-
tures, while concrete structures are limited to about 20%. Such a high share
of steel in building construction is unique in the world. The reason why
steel buildings are so popular in Japan may be found in the strong support
the government gave to developing the steel industry soon after the Meiji
Restoration. The policy invited an extraordinary advancement of the
technology of applying steel materials to various industrial products. At the
same time, however, another important factor cannot be neglected: Japan is
an earthquake-prone country located in one of the most quake-hazardous
regions on Earth. Thus, the basic idea of structural design has been estab-
lished: that strong and tough materials should be used in building structures
to resist the impact of severe earthquakes. Obviously, steel is the best material
to fit such requirements. Although steel is more expensive than concrete
and timber, the practice of employing it in buildings has been sustained by
the national wealth of Japan.
The origin of the basic idea of earthquake resistance is found in the
construction of Mitsui Honkan (Fig. 8-2), which was built on the basis of
lessons derived from the 1923 Kanto Earthquake. This building was designed
such that earthquake forces are totally sustained by steel frames and post-
earthquake fire is fully insulated by brick walls. The phrase“Buildings must
be absolutely earthquake-resistant and fire-resistant”described the design
concept at that time, one that has been inherited by today’s structural engi-
neers (Kuwamura (2002) ).

Y. Fujino, T. Noguchi (eds.) Stock Management for Sustainable Urban Regeneration, 133
© 2009 to the complete printed work by Springer, except as noted. Individual authors
or their assignees retain rights to their respective contributions; reproduced by permission.
134 Hitoshi Kuwamura

110
steel
100
wooden
90

80
Floor area (×106 m2)

70

60

50 RC

40

30
steel-RC
20

10
masonry
others
0
50 55 60 65 70 75 80 85 90 95 00
year

Fig. 8-1. Building construction in the latter half of the 20th century of Japan

Fig. 8-2. Mitsui-Honkan (Picture at 1929 completion, heritage of 1998 assignment.

With permission from MITSUI FUDOSAN CO., LTD.)

Even in the field of bridge construction, steel piers are often seen supporting
urban highways on short-span girders, while reinforced concrete piers are
common practices in advanced foreign countries. This obviously comes from
the earthquake-resistant concepts prevalent in Japan (Usami (2006) ).
 8. Earthquake-Resistant Engineering of Steel Structures 135

8.2 Damage to Steel Structures Caused by Earthquakes

Earthquakes are still serious natural hazards of unpredictable intensity

that defy human understanding. Even steel structures cannot be totally
free from seismic risk. The largest scale of damage in steel structures was
experienced during the 1995 Hyogoken-Nambu Earthquake (Architectural
Institute of Japan (1997) ). This earthquake was a devastating near-field
earthquake that devastated a big city of Japan, Kobe, that had been enjoying
continuous prosperity after World War II. Many buildings as well as
infrastructures were seriously damaged by the earthquake, and building
collapses, mostly of wooden houses, killed more than 6,000 citizens. The
damage to steel building structures can be categorized into two groups: one
whose causes and outcomes can be explained from current knowledge, and
another that cannot.
Introduced here are three buildings, the causes of whose failure can
be explained (Kuwamura (2002); (Kuwamura Lab (1995) ). Figure 8-3(a)
shows the collapse of a four-storey residential building of steel construction.
The first and second stories have been totally crushed without any original
shape. The double faults in the bracings and the columns, as sketched in Fig.
8-3(b), account for the collapse. The faults in terms of connection fracture
of the bracing and column are evidenced in Fig. 8-3(c) and Fig. 8-3(d),
respectively. The cause of the brace-end fracture is that fracture strength Pu
sustained by the net sectional area of the bolted connection is smaller than
the yield strength Py of the gross sectional area of the bracing. Thus, the
brace could not provide a sufficient amount of plastic elongation to resist
the seismic attack. If the condition of seismic connection is satisfied, i.e.,
Pu > Py, such a fatal failure can be prevented. The fracture of the column
base is due to the insufficient size of the fillet weld, since the weld deposit
is too small to transfer the column stress to the foundation. If the fillet size
is large enough, or a full-penetration weld is deposited, such a crucial failure
can be avoided.
Figure 8-4(a) shows an overturned four-storey office building. The failure
process as shown in Fig. 8-4(b) was triggered by the fracture of the welded
connection of the column as evidenced in Fig. 8-4(c). The connection was
too weak to sustain the earthquake force introduced against the column.
At present, it is common sense to employ full-penetration weld to such an
important connection of a main frame.
The third example as shown in Figure 8-5(a) is a parking-lot building
seriously distorted due to the fracture of bracing-to-column connections.
This is a typical example of poor detailing of steel connections. As shown
in Figure 8-5(b), the gusset plate joining the bracing and the column was
136 Hitoshi Kuwamura

Fig. 8-3. Collapse of a four-story steel residential building

welded to the web plate of the column, and consequently the web plate
was pulled out in the perpendicular direction of its plane and ruptured. An
adequate stiffening of the column web was necessary so that the tensile force
in the bracing is smoothly transferred into the column.
The cause of failure due to brittle fracture is not well understood given
the current state of knowledge. Figure 8-6(a) shows high-rise residential build-
ings constructed in a mega-structural system with the use of truss-type steel
columns. The main columns underwent through-section cracking by the
earthquake as shown in Fig. 8-6(b) and Fig. 8-6(c). Figure 8-7(a) shows the
same mode of fracture in the lower flange of a beam in an ordinary moment
frame for office use. This type of fracture was observed in more than 100
steel building frames in and around Kobe City and was a controversial
issue among structural engineers in those days. Experts guessed the cause
differently such as the lack of material toughness, cold temperature, strain
concentration at the vicinity of welded joints, poor welding workmanship,
large building drift caused by the epicentral earthquake, and large strain
rate due to vertical shock waves. A scientist picked up a sample from the
damaged building, shown in Fig. 8-7(b),, and observed the fracture surface
with a magnification of 1,000 by means of a scanning election microscope.
 8. Earthquake-Resistant Engineering of Steel Structures 137

Fig. 8-4. Overturn of a four-storey steel office building

Fig. 8-5. Distortion of a parking lot building

138 Hitoshi Kuwamura

Fig. 8-6. Brittle fracture of columns in a mega-structure

He identified a dimple pattern of Figure 8-7(c) at the origin of fracture and

a river pattern of Fig. 8-7(d) over the area of fracture propagation. The same
patterns of brittle fracture were found in a steel specimen he tested in his
laboratory. From this observation, it was clarified that the brittle fracture
is triggered by a micro-ductile crack generated during seismic vibration. A
constitutional equation for crack initiation has been calculated from stress-
strain analysis, while the condition of transition from ductile cracking to brittle
fracture has not yet been quantified. At present, a tentative provision is pro-
posed to prevent brittle fracture, one in which material, welding, and detail-
ing are ranked from poor to excellent quality, respectively (Architectural
Institute of Japan (2000) ); (Kuwamura (2003) ). It was fortunate that none
 8. Earthquake-Resistant Engineering of Steel Structures 139

Fig. 8-7. Brittle fracture of a beam in a moment frame

of the steel building frames suffering from such brittle fracture collapsed.
However, the reasons for this have not been clarified. The fail-safe mecha-
nism that fracture of some structure elements does not lead to system failure
is an important issue in seismic design.
In the field of bridge construction, the same type of damage also occurred
during the 1995 Kobe Earthquake. Figure 8-8 shows the buckling of plat
plates in a pier of rectangular hollow section, and Figure 8-9 shows the
so-called elephant foot bucking of a pier of circular hollow section. The dis-
tortion took place in the thinner plates at the place where the thickness alter-
nates. Since steel bridge structures are of thin-walled construction, they are
most susceptible to plate instability during the overload of earthquake. Figure
8-10 shows the fatal collapse of a box-section pier due to progressive tearing
of corner welds. Figure 8-11(a) and Fig. 8-11(b) show the brittle fracture of
columns of centrifugally cast steel pipe that supported a railroad.
140 Hitoshi Kuwamura

Fig. 8-8. Buckling of steel pier of rectangular hollow section

8.3 Fundamentals of Steel Structural Safety

The safety of steel structural buildings against external disturbances, espe-

cially earthquakes, is governed by the total coordination of three factors:
adequate selection of steel material, quality control of fabrication (espe-
cially welding), and technology of structural design. Japan, the most earth-
quake-prone country among highly industrialized nations, has advanced the
holistic research and development for the above three phrases, which are
described in the following sections.

8.3.1 Steel Material Selection

Most of the structural steels employed in building construction are speci-

fied in JIS (Japan Industrial Standards). A new steel named“Rolled Steel
for Building Structure”entered JIS in 1994. This steel was designed with
 8. Earthquake-Resistant Engineering of Steel Structures 141

Fig. 8-9. Buckling of steel pier of circular hollow section

Fig. 8-10. Collapse of steel pier of rectangular hollow section

142 Hitoshi Kuwamura

Fig. 8-11. Buckling of steel pier of centrifugally cast steel pipe

unique properties for earthquake resistance. This steel, as shown in Table

8-1, is designated SN followed by a number such as 400 and 490 that
indicates tensile strength. The final symbols A, B, and C designate more
detailed performance grades. The specific characteristics in the steel are seen
in grades B and C, in that the yield ratios are limited to 80% and the yield
points have upper as well as lower bounds. These properties are known to be
effective for steel structures to resist a strong earthquake in terms of energy
absorption after yielding.
A lower yield ratio means a larger extra-strength capability after yielding.
Thus, a steel member made of a lower yield ratio exhibits a larger plastic
deformation during the post-yielding strength, as shown in Figure 8-12. Thus,
the upper limit of yield ratio (YR) is essential to develop member ductility.
In a member made of the steel with an upper bound to its yield point, a
plastic hinge is certainly formed at the prescribed level of bending moment.
This means that the ultimate mechanism of the designer’s intention is real-
ized in his structure and then such intended ductility is surely provided to
the whole structural system. This is concretely demonstrated in Figure 8-13.
The plots in Fig. 8-13(a) show that the yield points of conventional steels
are scattered in a considerably larger range for smaller thickness. However,
when the range of dispersion becomes narrower like Range C (COV = 10%)
to Range A (COV = 2.5%) of Fig. 8-13(a), the system ductility is signifi-
cantly improved as shown in Fig. 8-13(b) and Fig. 8-13(c) (Kuwamura and
Sasaki (1990) ). In Fig. 8-13(b), three types of ultimate behavior are shown.
Types 1, 2, and 3 have horizontal sway of one story, three stories, and the
whole six stories, respectively, under earthquake forces. Type 3 exhibits the
highest strength in terms of load factor, but also the largest deformation in
terms of top displacement. This suggests that system ductility is increased
with the number of plastic sway stories involved in the ultimate behavior.
The result of statistical simulation of Fig. 8-13(c) shows that the reduction
Table 8-1. JIS specification of “Rolled Steels for Building Structures (SN steel)”

Yield point (N/mm2) Yield ratio (%) Elongation (%) Reduction Charpy
of area (%) impact
No. 1A No. 1A No. 4 energy (J)

Plate
thickness,
Plate thickness, t (mm) Plate thickness, t (mm) Plate thickness, t (mm)
t (mm)
Specimen
Tensile no. 4
6≤t 12 ≤ t 16 16 < t 40 < t strength 6 ≤ t 12 ≤ t 16 16 < t 40 < t 6 ≤ t 16 < t 40 < t 16 ≤ t rolling
2
Symbol < 12 < 16 ≤ 40 ≤ 100 (N/mm ) < 12 < 16 ≤ 40 ≤ 100 ≤ 16 ≤ 50 ≤ 100 ≤ 100 direction
SN400A 235 or 215 or 400– No 17 or 21 or 23 or No limit No limit
greater greater 510 limit greater greater greater
SN400B 235 or 235– 215– No 80 or 18 or 22 or 24 or 27 J or
greater 355 335 limit less greater greater greater greater
SN400C – – 235– – – 80 or Average
355 less of three 25
or greater
(each 15
or greater)
SN490B 325 or 325– 295– 490– No 80 or 17 or 21 or 23 or No limit
greater 445 415 610 limit less greater greater greater
SN490C – – 325– – – 80 or Average
445 less of three 25
8. Earthquake-Resistant Engineering of Steel Structures

or greater
(each 15
or greater)
143
144 Hitoshi Kuwamura

M
M/MP P
P
1.6 q

Py
1.4
0.0 Axial force ratio
0.2
1.2
0.0
1.0 0.2

0.8
0.4
0.4
0.6

0.4 Steel A (Low YR)

Steel B (High YR)
0.2
0.6 0.6
0
2 4 6 8 10 12 14 16
q /q P

Fig. 8-12. Yield ratio vs plastic deformation capacity of steel member

Fig. 8-13. Range of yield point vs Plastic deformation capacity of steel frame
 8. Earthquake-Resistant Engineering of Steel Structures 145

of yield point dispersion in terms of COV (coefficient of variation) brings

the statistical shift of larger number of sway stories involved in the ultimate
behavior. Thus, the upper limit of yield point (YP) is very important for
developing system ductility. The earthquake- resistant design on the basis
of weak beam concept, which is mentioned in the latter discussion, should
be associated with the use of the YP-control steel.

8.3.2 Welding

Needless to say, strength is fundamental for welded joints. In addition to

strength, alterations of material due to welding heat input as well as con-
figurations of welded joints are taken care of in the steel construction of
medium- to high-rise buildings in Japan. Such thoughtful quality control
may not be seen in other countries, because Japanese bitter experience from
past earthquakes impelled the careful fabrication.
Figure 8-14 shows the relationships between welding heat input and tough-
ness in terms of Charpy impact energy at heat-affected zones (Kuwamura
(2002) ). The welding heat input tends to increase in this order: manual
welding, CO2 gas shield arc welding, submerged arc welding, electro-gas
welding, and electro-slag welding; from that point the toughness tends to
decrease in the same order, since the welded connections from larger heat
input are more susceptible to brittle fracture. Welding with a large heat input
enhances the productivity of steel fabrication, but it may cause problems
in seismic performance of steel structures. Thus, the amount of heat input
must be limited and time-consuming workmanship practiced with care
about material properties.
Figure 8-14 shows various configurations at beam-to-column welded
connections subjected to severe stresses during earthquake motion. The
detailings in Fig. 8-15(a) are the conventional ones, but now it is known that
these connections are susceptible to fracture, because sharp notches at the
toes of fillet weld can become a cause of brittle fracture. Thus, the improved
connections in Fig. 8-15(b) are now in place.

8.3.3 Design

The safety of steel structures against a big earthquake is largely dependent on

the sophistication of structural design. Structural control, or the technique of
absorbing earthquake input energy is decisive. Recently, structural designers
have prescribed the ultimate behavior of their buildings and identified the
146 Hitoshi Kuwamura

Fig. 8-14. Welding heat input vs fracture toughness

Fig. 8-15. Improvement of connection detailing

 8. Earthquake-Resistant Engineering of Steel Structures 147

structural members forced into plastic strain. Such members are designed to
have enough ductility to dissipate earthquake energy. Figure 8-16 shows an
example of such structural control, which is called a weak-beam structure.
This type of seismic structure is practiced in reinforced concrete as well as
in steel construction. The columns are proportioned to be stronger than the
beams. When the building is attacked by a severe earthquake, plastic hinges
are formed at beam ends and horizontal drift is sustained by the cooperation
of all stories. This means that the total building system exhibits a very tough
resistance. If the columns are weak, a local inter-story failure may immediately
occur. Actually, several reinforced concrete frames collapsed in the mode of
local inter-story failure due to weak columns during the 1995 Earthquake.
Recently, base-isolated structures as schematically shown in Figure 8-17(a)
have become popular. However, it costs more than ordinary construction due
to double layer foundations for installing the isolation, and the resonance to
long-period seismic wave is apprehended. In order to avoid the double foun-
dations, a new type of isolation as shown in Fig. 8-17(b) has been invented,
in that the first story is provided for isolation. Figure 8-17(c) shows a typical
vibration-controlled system, one that uses dampers for energy dissipation
in steel bracings. Such advanced technology in seismic design is indebted
to newly developed damper materials and also computer-aided earthquake
response analysis.

8.4 Future Seismic Design

The concept of seismic design of buildings constructed of steel as well as

other materials is now in a revolutionary transition from purposes of life-
saving only to one that also includes property reservation, which is called

Fig. 8-16. Earthquake-resistant structure of weak-beam concept

148 Hitoshi Kuwamura

Fig. 8-17. Base-isolated structure and vibration-control structure

Performance-Based Design. The original idea for this concept is found in the
activities of the Vision 2000 Committee organized in 1992 by the Structural
Engineers Association of California (SEAOC), based on the lessons from
the 1989 Loma Prieta earthquake that attacked the coastal area around San
Francisco. Since this earthquake was not so severe (magnitude 7.1), the
safety of buildings designed to meet prevailing codes was assured, though
it caused economic losses of US$7 billion. The study of performance-based
design originated from the necessity of including the criterion of property
preservation in addition to life safety in the framework of seismic design.
The Committee submitted a final recommendation in 1995, in which holis-
tic seismic engineering of construction and maintenance as well as design is
schemed (Structural Engineers Association of California (1995) ).
The 1994 Northridge Earthquake (M6.7) caused a larger economic loss
(US$20 billion) than the Loma Prieta Earthquake and enhanced the activi-
ties of the Structural Engineers Association of California (SEAOC). In addi-
tion, the earthquake revealed the brittle fracture problems of steel building
frames. It had been believed that steel moment frames were ductile enough
to resist a big earthquake, but a lot of beam-to-column connections under-
went brittle fracture causing an unacceptable level of losses of building
properties despite little loss of human life. In order to establish a provision for
preventing brittle fracture in steel moment frames, the SAC Joint Venture
 8. Earthquake-Resistant Engineering of Steel Structures 149

of SEAOC, ATC, and CUREe was organized in 1994 on a research fund

from the Federal Emergency Management Administration (FEMA). The
SAC Joint Venture submitted a final report in 2000, in which a framework
of performance-based design was proposed with an emphasis on seismic
performance of steel moment frames (SAC Joint Venture (2000) ).
In Japan, in accordance with the American movement of performance-
based design, a national project named “Development of New Building
Structural Systems” was started in 1995 by the initiative of Ministry of
Construction, and the study on performance-based design was promoted
for three years. The necessity of performance-based design in Japan was
compelled by two events: the 1995 Hyogo-ken Nambu Earthquake (M7.2)
which caused huge economic losses of more than ¥10-Cho (about US$100
billion) as well as more than 6,000 human deaths; and the 1996 US-Japan
top-level conference about the business of building construction being open
to the free market. The final report of the national project was published in
2000 (Ministry of Construction (2000) ). The basic concept in the report is
similar to that of the Vision 2000 Committee, and is reflected in the amend-
ment of the Building Standard Law issued in 1998 and enforced in 2000.
Building design is now moving from being specification-based to being
performance-based.
As mentioned above, the aim of performance-based design is to preserve
building properties together with human lives. However, the essence of
performance-based design lies in mutual recognition between designers
and customers about building performance. In order to preserve building
properties against earthquake, an adequate extra amount of money should
be invested to the construction. Thus the designers should be responsible
for the accountability on any cost vs. effect issues. In such a situation, the
interface between designer and customer is the performance matrix. Several
forms of the performance matrix were proposed by SEAOC Vision 2000,
SAC-FEMA, and the Japan National Project, and are not yet unified, but the
concept is a shared one, as shown in Table 8-2 (Kuwamura et al. (2002) ).
The row (horizontal axis) in the matrix indicates the level of external load
in terms of the probability of occurrence, and the line (vertical axis) indi-
cates the level of performance in terms of the degree of damage. The target
performance of the building structure of concern is determined by placing
marks in the matrix, like single-hollow circles. Since a larger intensity of
lead obviously causes heavier damage, a set of marks is placed downward
to the right-hand side. When the set of single-hollow circles represents a
standard level of performance, the sets of double-hollow circles and star
marks indicate higher levels of performance.
150 Hitoshi Kuwamura

Table 8-2. Performance matrix

The performance level, which constitutes an axis of the performance

matrix, is correspondent to the degree of damage resulting from the appli-
cation of external loads. Thus, the performance level ranges widely from
healthy state to fatal collapse. The lower bound of a performance level, i.e.,
the limit beyond which the performance cannot be kept, is called the limit
state. The definition of performance levels is now subject to various proposals,
but the following three are considered to be essential: (1) the classified
performance should be meaningful to the customers, (2) the associated
damage should be identifiable, and (3) the limit state should be quantifi-
able by means of engineering mathematics. The performance levels shown
in Table 8-3 are proposed for steel structural buildings to meet the above-
mentioned three requirements (Kuwamura et al. (2002) ). The performance
is ranked into four levels: Level 1, Functional; Level 2, Undamaged; Level
3, Reparable; and Level 4, Life Safe. Level 5, Collapse is not permissible
except for special cases such as warehouses without human accommodation.
 8. Earthquake-Resistant Engineering of Steel Structures 151

The limit states associated with Performance Levels 1–4 are called function
limit, damage limit, repair limit, and safety limit, respectively.
From the viewpoint of stock preservation, the performance of reparabil-
ity has recently been noticed as important. If a building damaged by an
earthquake can be repaired and reused without demolishing, the economic
loss will be significantly reduced and furthermore environmental impact
due to constructional wastes will be diminished. The reparability design,
which has not been considered in the conventional design philosophy, first
appeared in the framework of performance-based design. From the analysis
of damage and repair costs to steel-framed buildings that suffered moderate
damage from the 1995 Kobe Earthquake, the reparability limit state was
quantified. The reparability for reuse is found to be judged from the residual
drift after an earthquake. When overall residual drift and maximum inter-
story residual drift are not more than 1/200 and 1/90, respectively, the building
can be repaired with acceptable cost and current technology (Iwata et al.
(2005) ). The problem that remains is to establish a method of predicting the
residual drift as exactly as possible in the design procedure.

8.5 Conclusive Remarks with Seismic Rehabilitation

of Existing Structures

There exist a lot of inadequate structures in terms of earthquake resistance.

They were constructed on the old design code of low demands and have
possibly deteriorated during long service periods. Many types of rehabilita-
tion techniques for old steel structures are going to be developed, such as
strengthening with supplemental structural elements and adding of base
isolation or energy absorbing devices.
A part of the 2004 supplementary government budget was assigned to
earthquake rehabilitation of Building No.11 of the School of Engineering
at the University of Tokyo (Fig. 8-18). Construction for strengthening was
started in May 2005 and was completed at the end of December 2005.
In earthquake-prone countries like Japan, the technology of seismic retrofit-
ting of existing buildings is substantial for the sustainability of urban soci-
ety. The current advanced seismic technology of Japan was applied to the
retrofitting of Building No.11, from which a model of seismic rehabilitation
was proposed in the COE program Creation of New Society and Engineering
for Sustainable Urban Regeneration (Department of Architecture, School of
Engineering, the University of Tokyo (2005) ).
Table 8-3. Performance description for steel buildings
152

Damage state of building parts

Building as a Structural Non-structural
Performance level whole skeleton elements Facilities Furniture Remarks
Level 1 Functional Continuous No damage occurs No damage. Operational Perfectly No damage and
service is in structure. as usual. protected. no
retained without Deflection and inconvenience
Hitoshi Kuwamura

any discomfort vibration are minor for users.

and disturbance. to keep comfort.
(Function limit)
Level 2 Undamaged Slight damage is Negligible damage Little crack or Emergency Slide and Temporary
observed, but such as micro minor peeling lock may overturning inconvenience
basic function is yielding at stress may occur, but work, but of light and little
sustained for concentration any malfunction immediately furniture may economical
normal service. may takes place. such as leak of restart. No occur, but no losses may
Deformation is rain water does malfunction major damage. occur. Normal
within elastic limit not occur. of safety Hazardous services can be
with no residual Repair may be equipment materials are resumed
deflection. done at users such as protected. immediately
Inspection and convenience. sprinklers. (green tag).
repair are
unnecessary.
(Damage limit)
Level 3 Repairable Apparent damage Elastic limit is Moderate to Many Many of Continuous
is observed, but is exceeded, and severe damage facilities are fixtures are occupation
technically and some extent of occurs, but are damaged damaged due after damage
economically yielding, buckling reparable or and non- to slide and may be inad-
reparable. Original and fracture are replaceable. operational, overturning, equate, which
performance is partly observed. Large falling but can be but valuable is judged only
recovered and The structural hazards are not repaired by or hazardous by experts
building is reusable capacity is created. experts. materials are (yellow tag).
as before. Minor moderately protected.
injury may occur. impaired.
(Repair limit)
Level 4 Life safe Major damage Extensive yielding, Extensive Emergency Many are Off limits
occurs, but build- buckling and damage occurs, facilities seriously (red tag).
ing sustains its own fracture occur, but but fatal falling for human damaged, but
weight. Floor and the frame can hazards are not lives are dangerous
roof do not fail. withstand gravity created. Egress operational, materials do
Human lives are load pertinent to may be but others not spill.
protected, while building. obstructed, but are non-
some are wounded. self-escape and operational.
rescue activities
are possible.
(Safety limit)
(continued)
8. Earthquake-Resistant Engineering of Steel Structures
153
Table 8-3. (continued)
154

Damage state of building parts

Building as a Structural Non-structural
Performance level whole skeleton elements Facilities Furniture Remarks
Level 5 Collapse Major parts or Serious damage Falling hazards Almost all Almost all are This state is
whole of building causes major loss are fatal. Egress facilities are substantially not socially
collapse, and of structural may be seriously fatally damaged. permitted except
Hitoshi Kuwamura

human lives are capacity for obstructed. damaged Hazardous for

exposed to critical sustaining building and non- materials may special build-
state. own weight. operational. spill out. ings in which
no humans and
no hazardous
materials are
contained.
 8. Earthquake-Resistant Engineering of Steel Structures 155

Fig. 8-18. Building No.11 of school of engineering (before seismic rehabilitation)

References

Architectural Institute of Japan (1997) Report on the Hanshin-Awaji Earthquake

Disaster, Building Series Vol 3, Structural Damage to Steel Buildings,
Maruzen
Architectural Institute of Japan (2000) Tentative Provisions for Brittle Fracture of
Steel Buildings, Proceedings of Panel Discussion: 9.
Department of Architecture, School of Engineering, the University of Tokyo (2005)
Report on Seismic Retrofitting of Building No 11 of School of Engineering: 3
Iwata, Y., Sugimoto, H. and Kuwamura, H. (2005) Reparability Limit of Steel
Structural Buildings. Journal of Structural and Construction Engineering
(Transactions of AIJ) 588(2): 165–172
Kuwamura, H. (2002) Performance and Design of Steel Structures. 1.2 History of
Iron and Steel Structures, 11.10 Welding and Earthquake Damage, Kyoritsu-
Shuppan
Kuwamura, H. (2003) Classification of Material and Welding in Fracture
Consideration of Seismic Steel Frames, Engineering Structures. Journal of
Earthquake, Wind and Ocean Engineering 25(5): 547–563
Kuwamura, H. and Sasaki, M. (1990) Control of Random Yield-Strength for
Mechanism-Based Seismic Design. Journal of Structural Engineering, ASCE
116(1): 98–110
Kuwamura Lab (1995) Field Survey Report on 1995 Hyogoken-Nambu Earthquake
Disaster
156 Hitoshi Kuwamura

Kuwamura, H., Tanaka, N., Sugimoto, H. and Kouno, T. (2002) Performance

Description of Steel Structures. Journal of Structural and Construction
Engineering 562(12): 175–182.
Ministry of Construction (2002) Concept of Performance-Based Design of Building
Structures, Gihodo-Shuppan
SAC Joint Venture (2000) Recommended Seismic Design Criteria for New Steel
Moment-Frame Buildings, FEMA-350, Chap 2 General Requirements, Chap 4
Performance Evaluation
Structural Engineers Association of California (SEAOC) (1995) Sacramento,
CA: Vision 2000 Committee: Performance Based Seismic Engineering of
Buildings, Sacramento, CA
Usami, B. (2006) Guideline for Seismic and Damage Control Design of Steel
Bridges, Gihodo-Shuppan